Sunteți pe pagina 1din 122

PUBLIC

SAP Predictive Analytics


2018-06-04

Predictive Analytics Installation Guide


Content

1 About this Guide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2 What's New in Predictive Analytics Installation Guide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 Installation Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.1 SAP Predictive Analytics Deployment Scenarios. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.2 Installing the Desktop Version of SAP Predictive Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.3 Installing the Client/Server Version of SAP Predictive Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.4 Installing SAP Predictive Analytics for SAP HANA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.5 Installing SAP Predictive Analytics for SAP HANA HEC or SAP Cloud Platform . . . . . . . . . . . . . . . . . 14

4 Installing Predictive Factory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16


4.1 Overview of Predictive Factory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.2 Installing Predictive Factory on Windows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3 Installing Predictive Factory on Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
4.4 Starting and Stopping Predictive Factory Services - On Windows. . . . . . . . . . . . . . . . . . . . . . . . . . .22
4.5 Starting and Stopping Predictive Factory Services - On Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.6 Upgrading Predictive Factory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.7 Uninstalling Predictive Factory - On Windows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
4.8 Uninstalling Predictive Factory - On Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

5 Installing SAP Predictive Analytics Server for Windows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25


5.1 SAP Predictive Analytics Enterprise Edition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Minimum Skills Required to Perform the Installation on Windows. . . . . . . . . . . . . . . . . . . . . . . . 26
Server Sizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Network Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Java Runtime Environment for Client. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Java Web Start Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.2 Server Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.3 Performing a Silent Installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.4 Checking Installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.5 Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Basic Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Configuration of Server Instances. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Port Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Communication Channel Encryption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Authenticated Server Based on the System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Authenticated Server Based on a Fixed Encrypted Password List. . . . . . . . . . . . . . . . . . . . . . . . 39

Predictive Analytics Installation Guide


2 PUBLIC Content
Authenticated Server Based on a Fixed Password List. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
No Authentication. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.6 Managing the Server on Windows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42
Launching a .bat File. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Using Windows Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43
5.7 Deploying the Client. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.8 Starting the Administration Console. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5.9 Logging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5.10 Troubleshooting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.11 Setting User Security for Windows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

6 Installing SAP Predictive Analytics Server for Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47


6.1 SAP Predictive Analytics Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
SAP Predictive Analytics Enterprise Edition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Reloading the Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Managing the Server on Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66
Deploying the Client. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
6.2 Monitoring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Logging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Starting the Administration Console. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.3 Troubleshooting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Displaying the Server Log File. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Error: ./tmp/KxAuthServer_.pid is not running. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69

7 Installing SAP Predictive Analytics Desktop. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71


7.1 SAP Predictive Analytics Desktop Edition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71
7.2 Installing Predictive Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Installation Prerequisites. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Installing SAP Predictive Analytics Using the Setup Program. . . . . . . . . . . . . . . . . . . . . . . . . . . 72
Performing a Silent Installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
7.3 Configuring Trace Logs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Configuring Trace Logs for Expert Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Configuring Trace Logs for Automated Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
7.4 Important Considerations for Using SAP HANA with Expert Analytics. . . . . . . . . . . . . . . . . . . . . . . 77
Configuring _SYS_REPO for the SAP Predictive Analytics User. . . . . . . . . . . . . . . . . . . . . . . . . . 78
Granting Object Privileges to Access HANA Online Data Source. . . . . . . . . . . . . . . . . . . . . . . . . 78
Granting Privileges to Run Demand Forecasting Functionality with AFL. . . . . . . . . . . . . . . . . . . . 79
Granting Privileges to Run PAL Algorithms with AFL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Granting Privileges to Run OFL Functionality with AFL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

Predictive Analytics Installation Guide


Content PUBLIC 3
Supported OLAP Measures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Using SAP APL Functions with Expert Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
7.5 Important Considerations for Using SAP BusinessObjects Universes with Expert Analytics. . . . . . . . 81
7.6 Connecting Automated Analytics to your Database Management System. . . . . . . . . . . . . . . . . . . . .81
7.7 Installing and Configuring R. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Installing R and the Required Packages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81
Configuring R. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Important Considerations for Using Expert Analytics with R Algorithms in SAP HANA Online
Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
7.8 Uninstalling SAP Predictive Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Performing a Standard Uninstallation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Performing a Silent Uninstallation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

8 Installing SAP Predictive Analytics Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86


8.1 Installing the SAP Predictive Analytics Client on Windows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .86
Performing a Silent Installation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
8.2 Installing the SAP Predictive Analytics Client on Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
8.3 Upgrading the Client. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
8.4 Checking the Connection between the Client and Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
8.5 Starting the Client for Linux Hosts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
8.6 Troubleshooting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Installation Error Message. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Invalid or Missing License. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
Full Client System Application Installation (Server and Client). . . . . . . . . . . . . . . . . . . . . . . . . . 90
Server and Remote Assistant Installed on Different Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . 90
Common Check Point. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

9 Installing Java Web Start for Windows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91


9.1 About Java Web Start. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
9.2 Installing the Web Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Installing Microsoft IIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Installing Apache Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92
9.3 Setting Up the Web Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92
Publishing with Microsoft IIS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Publishing with Apache HTTP Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
9.4 Configuring the Remote Assistant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
9.5 Setting the SSL Certificates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
9.6 Firewall Settings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
9.7 Starting the Remote Assistant and the Administration Console. . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
9.8 Allowing Client Access to SAP Predictive Analytics Java Web Start Interface. . . . . . . . . . . . . . . . . . .99

10 Installing Java Web Start for Linux. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

Predictive Analytics Installation Guide


4 PUBLIC Content
10.1 About Java Web Start. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Requirements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
10.2 Installing the Web Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .100
10.3 Publishing with Apache HTTP Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Publishing the Web Start Folders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Configuring the MIME Types. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
10.4 Configuring the Remote Assistant. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103
10.5 Configuring the Administration Console. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
10.6 Setting the SSL Certificates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
10.7 Firewall Settings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
10.8 Starting the Remote Assistant and the Administration Console. . . . . . . . . . . . . . . . . . . . . . . . . . . 105

11 Running Multiple Instances of Java Web Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106


11.1 Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
General Multiple Installation Architecture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
11.2 Installing SAP Predictive Analytics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Server Side. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Client Side. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .108
11.3 Configuring the Application Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Managing Your Software License(s). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Log Files Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
11.4 Microsoft Windows Application Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Configuring the Service Ports and Starting the Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
11.5 Microsoft Windows Service Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Configuring the Service Ports and Registering the Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
11.6 Linux Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
To Configure the Services. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
11.7 Connecting to the Server. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
On MS-Windows Platforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
On Linux Platforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
11.8 Setting Multiple Java Web Start Launchers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116
JNLP Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Web Server Mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119

Predictive Analytics Installation Guide


Content PUBLIC 5
1 About this Guide

This guide provides you with information and procedures for installing and upgrading SAP Predictive Analytics
applications. It is a collection of previously independent installation guides that have been grouped into a single
guide.

Predictive Analytics Installation Guide


6 PUBLIC About this Guide
2 What's New in Predictive Analytics
Installation Guide

Links to information about the new features and documentation changes in the installation guide for SAP
Predictive Analytics 3.3.

Section What's New Link to More Information

Installing Predictive Factory The Predictive Factory architecture has Overview of Predictive Factory [page
been updated. 16]

Installing Predictive Factory Predictive Factory is now supported on ● Installing Predictive Factory on Li­
Linux. nux [page 18]
● Uninstalling Predictive Factory - On
Linux [page 24]
● Starting and Stopping Predictive
Factory Services - On Linux [page
22]

Installing SAP Predictive Analytics Update of the prerequisites RedHat [page 51]
Server for Linux

Predictive Analytics Installation Guide


What's New in Predictive Analytics Installation Guide PUBLIC 7
3 Installation Overview

3.1 SAP Predictive Analytics Deployment Scenarios

The diagram below shows the four main deployment scenarios for SAP Predictive Analytics described in this
overview document.

For an overview of the installation steps for a deployment, click on the relevant part of the image below.

The SAP packages are available from the SAP download center (https://support.sap.com/swdc ). Refer to
the Product Availability Matrix for more information about supported software versions.

● Installing the Desktop Version of SAP Predictive Analytics [page 9]


● Installing the Client/Server Version of SAP Predictive Analytics [page 10]
● Installing SAP Predictive Analytics for SAP HANA HEC or SAP Cloud Platform [page 14]
● Installing SAP Predictive Analytics for SAP HANA [page 11]

Predictive Analytics Installation Guide


8 PUBLIC Installation Overview
3.2 Installing the Desktop Version of SAP Predictive
Analytics

How to install SAP Predictive AnalyticsDesktop, the on-premise standalone version of SAP Predictive Analytics.

For full information about how to install SAP Predictive Analytics Desktop in the default setup, or in your own
personalized setup, go to section SAP Predictive Analytics Desktop Edition [page 71].

Before installing SAP Predictive Analytics, make sure the following requirements are met:

● Check the SAP Product Availability Matrix http://service.sap.com/sap/support/pam for operating


system compatibility. SAP Predictive Analytics is supported on 64-bit machines.
● If you have already installed SAP Predictive Analysis on your machine, you need to uninstall it before
installing SAP Predictive Analytics.

Note
If you have already installed SAP Lumira on your machine, you do not need to uninstall it. SAP Predictive
Analytics is installed in addition in a separate folder.

If you have already installed SAP InfiniteInsight on your machine, you do not need to uninstall it.
However, SAP Predictive Analytics is installed in addition in a separate folder and will not replace any
installed version of SAP InfiniteInsight.

● You must have Administrator rights to install SAP Predictive Analytics on the computer.
● Sufficient disk space must be available on various resources. For a detailed list of supported environments
and hardware requirements, see the Product Availability Matrix at: http://service.sap.com/pam

Installing

The SAP Predictive Analytics Setup program is contained within the self-extracting archive -
SAPPredictiveAnalysisSetup.exe. The program automatically recognizes your computer's operating
system and checks for platform requirements. It updates files as required.

Note
To perform silent installation, go to Performing a Silent Installation [page 73].

1. Navigate to the SAP Predictive Analytics self-extracting archive and double-click it. The User Account
Control dialog box appears with a warning message.
2. Choose Yes in the confirmation prompt. The SAP Predictive Analytics Setup program is extracted from the
archive. The Installation Manager performs a verification check for all of the installation prerequisites. A
Prerequisites page opens only if the verification fails for any requirement. Close the wizard and correct any
missing prerequisite before relaunching PADESKTOP<version number>_<patch level>.EXE. If all of
the installation prerequisites are confirmed, the Define Properties page opens.
3. Select the setup language from the dropdown list.
4. Specify the destination folder for installing SAP Predictive Analytics. To accept the default installation
directory, choose Next. The License Agreement page appears.

Predictive Analytics Installation Guide


Installation Overview PUBLIC 9
5. Review the license agreement and select I accept the License Agreement.
6. Choose Next.
7. In the Ready to Install page, choose Next.
8. When the Finish Installation page opens, to automatically launch the program, select Launch SAP
Predictive Analytics after installation completes.

You can configure trace logs according to your own requirements as described in Configuring Trace Logs for
Expert Analytics [page 75].

3.3 Installing the Client/Server Version of SAP Predictive


Analytics

How to install the SAP Predictive Analytics Client/Server version for use with a non-SAP HANA database.

Note
You can install the Desktop version and the Client/Server version on the same machine if you want to use
the Desktop Expert Analytics module.

To install the Client/Server version of SAP Predictive Analytics, the reference sections that you need to refer to
depend on the environment where you want to install the client and the server. If you choose to deploy your
client machines from the Server host, you need to install the Java Web Start on the server host. Refer to the
relevant section as described below:

Installing the Client/Server version of SAP Predictive Analytics

Software to deploy Installing on Windows Installing on Linux

Installing the Server software To install the server on Windows, refer To install the server on Linux, refer to
to section Server Installation [page Installing the Server [page 51].
28].
To deploy your client software from the
To deploy your client software from the server host, install the Java Web Server
server host, install the Java Web Server on the Linux host machine. Refer to In­
on the Windows host machine. Refer to stalling the Web Server [page 100].
Installing the Web Server [page 91].

Installing the Client software To deploy your client software directly To deploy your client software directly
on the client machines, refer to Instal­ on the client machines, refer to Instal­
ling the SAP Predictive Analytics Client ling the SAP Predictive Analytics Client
on Windows [page 86]. on Linux [page 87].

Setup your database connectivity. Refer to the Predictive Analytics Data Refer to the Predictive Analytics Data
Access guide. Access guide.

Predictive Analytics Installation Guide


10 PUBLIC Installation Overview
Installing Predictive Factory

You can install Predictive Factory on a Windows environment or on a Linux environment. To install Predictive
Factory on windows, go to Installing Predictive Factory on Windows [page 17] and to install Predictive Factory
on Linux, go to Installing Predictive Factory on Linux [page 18].

3.4 Installing SAP Predictive Analytics for SAP HANA

How to install SAP Predictive Analytics for SAP HANA Runtime use, our on-premise solution where you can
install the Desktop version or the Client/Server version of SAP Predictive Analytics and Predictive Factory.

To install for use on SAP HANA, perform the following steps:

1. Install SAP HANA DB.


2. Install SAP HANA APL (Automated Predictive Library).
3. Install UDF AFL (which also installs PAL) if you want to use demand forecasting capabilities.
4. Install R Server (this is typically installed on a separate machine from the SAP HANA system) if you need it.
5. If required, install the Hadoop/Spark/Hive instance.
6. Install SAP Predictive Analytics Desktop or Client/Server and Predictive Factory.

Installation of SAP HANA

You should have the following already installed. If not, install these before installing R server, Hadoop, or SAP
Predictive Analytics. To install SAP HANA, refer to the installation chapter of the SAP HANA Installation Guide
downloadable from: http://help.sap.com/hana and refer to Product Availability Matrix to know which version
to install.

Install SAP HANA APL (Automated Predictive Library)

Install SAP HANA Automated Predictive Library, refer to Chapter 2 of the SAP HANA Automated Predictive
Library Reference Guide.

Install UDF AFL

To make use of the Demand Forecasting option, add UDF AFL for SAP HANA, the information on how to install
this can be found in SAP Note 2050229 (https://support.sap.com/)

Predictive Analytics Installation Guide


Installation Overview PUBLIC 11
Install R Server

Download the necessary Linux libraries, guides and packages from the R Server site. Go to Installing R and the
Required Packages [page 81].

Install the Hadoop/Spark/Hive instance (optional)

Set up your Hadoop/Spark/Hive instance and populate with datasets, guides and packages for these are
available from the appropriate Web site.

Installing SAP Predictive Analytics Desktop

For full information about how to install SAP Predictive Analytics Desktop in the default setup, or in your own
personalized setup, go to Installation Prerequisites [page 72].

Before installing SAP Predictive Analytics, make sure the following requirements are met:

● Check with the SAP Product Availability Matrixhttp://service.sap.com/sap/support/pam that the


operating system installed on your machine is supported by. SAP Predictive Analytics.
● If you have already installed SAP Predictive Analytics on your machine, you need to uninstall it before
installing SAP Predictive Analytics.

Note
If you have already installed SAP Lumira on your machine, you do not need to uninstall it. SAP Predictive
Analytics is installed in addition in a separate folder.

If you have already installed SAP InfiniteInsight on your machine, you do not need to uninstall it.
However, SAP Predictive Analytics is installed in addition in a separate folder and will not replace any
installed version of SAP InfiniteInsight.

● You must have Administrator rights to install SAP Predictive Analytics on the computer.
● Sufficient disk space must be available on various resources. For a detailed list of supported environments
and hardware requirements, see the Product Availability Matrix at: http://service.sap.com/pam

Installation
The SAP Predictive Analytics Setup program is contained within the self-extracting archive -
SAPPredictiveAnalysisSetup.exe. The program automatically recognizes your computer's operating
system and checks for platform requirements. It updates files as required.

Note
To perform silent installation, refer to Performing a Silent Installation [page 73].

1. Navigate to the SAP Predictive Analytics self-extracting archive - PADESKTOP<version


number>_<patch level>.EXE, for example PADESKTOP2002_0.EXE for version 2.2 patch level 0 - and
double-click it. The User Account Control dialog box appears with a warning message.

Predictive Analytics Installation Guide


12 PUBLIC Installation Overview
2. Choose Yes in the confirmation prompt. The SAP Predictive Analytics Setup program is extracted from the
archive. The Installation Manager performs a verification check for all of the installation prerequisites. A
Prerequisites page opens only if the verification fails for any requirement. Close the wizard and correct any
missing prerequisite before relaunching PADESKTOP<version number>_<patch level>.EXE. If all of
the installation prerequisites are confirmed, the Define Properties page opens.
3. Select the setup language from the dropdown list.
4. Specify the destination folder for installing SAP Predictive AnalyticsNext. The License Agreement page
appears.
5. Review the license agreement and select I accept the License Agreement. To accept the default installation
directory, choose .
6. Choose Next.
7. In the Ready to Install page, choose Next.
8. When the Finish Installation page opens, to automatically launch the program, select Launch SAP
Predictive Analytics after installation completes.

Configure trace logs as described in Configuring Trace Logs for Automated Analytics [page 76].

Installing SAP Predictive Analytics Client/Server software

To install the Client/Server version of SAP Predictive Analytics, the reference sections that you need to refer to
depend on the environment where you want to install the client and the server. If you choose to deploy your
client machines from the Server host, you need to install the Java Web Start on the server host. Refer to the
relevant section as described below:

Installation Scenarios

Installing the Client/Server version of SAP Predictive Analytics

Software to deploy Installing on Windows Installing on Linux

Installing the Server software To install the server on Windows, refer To install the server on Linux, refer to
to Server Installation [page 28]. Installing the Server [page 51].

To deploy your client software from the To deploy your client software from the
server host, install the Java Web Server server host, install the Java Web Server
on the Windows host machine. Refer to on the Linux host machine. Refer to In­
Installing the Web Server [page 91]. stalling the Web Server [page 100].

Installing the Client software To deploy your client software directly To deploy your client software directly
on the client machines, refer to Instal­ on the client machines, refer to Instal­
ling the SAP Predictive Analytics Client ling the SAP Predictive Analytics Client
on Windows [page 86] . on Linux [page 87].

Setup your database connectivity. Refer to the Predictive Analytics Data Refer to the Predictive Analytics Data
Access guide section Connecting to Access guide section Connecting to
your Database Management System on your Database Management System on
Windows . Linux .

Predictive Analytics Installation Guide


Installation Overview PUBLIC 13
Installing Predictive Factory
You install Predictive Factory on a Windows or on a Linux environment. To install Predictive Factory on
Windows, refer to Installing Predictive Factory on Windows [page 17]. To install Predictive Factory on Linux,
refer to Installing Predictive Factory on Linux [page 18].

3.5 Installing SAP Predictive Analytics for SAP HANA HEC


or SAP Cloud Platform

How to install SAP Predictive Analytics for SAP HANA full use with HEC (SAP HANA Enterprise Cloud) or SAP
Cloud Platform.

Installing SAP Predictive Analytics for SAP HANA Full Use on HEC (SAP
HANA Enterprise Cloud)

For further information about HEC, refer to HEC documentation on the SAP HANA Platform page at http://
help.sap.com/hana.

To install for use on HEC, perform the following steps:

1. Install SAP HANA Advanced Edition


2. Install SAP HANA APL (Automated Predictive Library)
3. Install UDF AFL (which also installs PAL) if you want to use demand forecasting capabilities.
4. Install R Server (this is typically installed on a separate machine from the SAP HANA system)
5. Install the Hadoop/Spark/Hive instance (optional, depending on your setup)
6. Install SAP Predictive Analytics Client/Server
7. Install Predictive Factory

Installing of SAP HANA Advanced Edition


You should have the following already installed. If not, install these before installing R server , Hadoop, or
Predictive Analytics. To install SAP HANA, refer to the installation chapter of the SAP HANA Installation Guide
downloadable from: http://help.sap.com/hana

Installing SAP HANA APL (Automated Predictive Library)


Install SAP HANA Automated Predictive Library, refer to Chapter 2 of the SAP HANA Automated Predictive
Library Reference Guide.

Installing UDF PAL


Add UDF AFL for SAP HANA, the information on how to install this can be found in SAP Note 2050229 (https://
support.sap.com/)

Installing R Server
Download the necessary Linux libraries, guides and packages from the R Server site. Refer to Installing R and
the Required Packages [page 81].

Predictive Analytics Installation Guide


14 PUBLIC Installation Overview
Installing the Hadoop/Spark/Hive instance

Set up your Hadoop/Spark/Hive instance and populate with datasets, guides and packages for these are
available from the appropriate Web site.

Installing SAP Predictive Analytics Client/Server software

To install the Client/Server version of SAP Predictive Analytics, the reference sections that you need to refer to
depend on the environment where you want to install the client and the server. If you choose to deploy your
client machines from the Server host, you need to install the Java Web Start on the server host. Refer to the
relevant section as described below:

Installing the Client/Server version of SAP Predictive Analytics

Software to deploy Installing on Windows Installing on Linux

Installing the Server software To install the server on Windows, refer To install the server on Linux, refer to
to Server Installation [page 28]. Installing the Server [page 51].

To deploy your client software from the To deploy your client software from the
server host, install the Java Web Server server host, install the Java Web Server
on the Windows host machine. Refer to on the Linux host machine. Refer to In­
Installing the Web Server [page 91]. stalling the Web Server [page 100].

Installing the Client software To deploy your client software directly To deploy your client software directly
on the client machines, refer to Instal­ on the client machines, refer to Instal­
ling the SAP Predictive Analytics Client ling the SAP Predictive Analytics Client
on Windows [page 86] . on Linux [page 87].

Setup your database connectivity. Refer to the Predictive Analytics Data Refer to the Predictive Analytics Data
Access guide section Connecting to Access guide section Connecting to
your Database Management System on your Database Management System on
Windows . Linux .

Installing Predictive Factory

You install Predictive Factory on a Windows or on a Linux environment. To install Predictive Factory on
Windows, refer to Installing Predictive Factory on Windows [page 17]. To install Predictive Factory on Linux,
refer to Installing Predictive Factory on Linux [page 18].

Using Predictive Analytics Functionality in SAP Cloud Platform

About SAP Cloud Platform

SAP Cloud Platform Predictive Services are a set of developer-level services running within an SAP Cloud
Platform Compute instance. It allows any application on SAP Cloud Platform to easily embed functionality from
an underlying Automated Predictive Library (APL) that executes directly on a tenant’s SAP HANA instance.
SAP Cloud Platform Predictive Services enables developers to utilize directly SAP Predictive Analytics
“Automated Analytics” capabilities without predictive modeling or the skills of a data scientist. It is available
only on top of SAP Cloud Platform HANA DB Service, Platform Edition.

Predictive Analytics Installation Guide


Installation Overview PUBLIC 15
4 Installing Predictive Factory

4.1 Overview of Predictive Factory

Predictive Factory is a thin-client, Web server-based application that lets you create and automate the
management of predictive models created in SAP Predictive Analytics Modeler. Users can automate much of
the production lifecycle, such as retraining a model, applying a model to a new dataset, and detecting model
deviations.

Predictive Factory is installed as part of an SAP Predictive Analytics client-server configuration. Once your
Automated Analytics server is deployed, a Predictive Factory server can be installed and configured.

The diagram shows Predictive Factory in the SAP Predictive Analytics client-server landscape.

● The Predictive Factory server contains a group of services handled by the Server Intelligence Agent (SIA).
The services can be stopped and started with the Start and Stop Predictive Factory commands.
● Communication between the Predictive Factory server and clients uses HTTP. You can configure encrypted
communication (HTTPS) using Secure Sockets Layer (SSL) or Transport Layer Security (TLS).
● In Predictive Factory, you declare one or more modeling servers. (A modeling server is the machine that
hosts the Automated Analytics server that will run the modeling tasks). Predictive Factory communicates
with the Automated Analytics server using CORBA. All connections to data and model sources are handled
via the modeling server.
● Users create models using the SAP Predictive Analytics client connected to the same Automated Analytics
server.
● If you have set up an SAP HANA server in Predictive Factory, you can import Predictive Pipelines created
with Expert Analytics. To apply or run predictive pipelines, Predictive Factory connect directly to SAP HANA
using a JDBC communication.

● If you have set up a PAi Application in Predictive Factory, you can publish model versions to an external
application.

Predictive Analytics Installation Guide


16 PUBLIC Installing Predictive Factory
4.2 Installing Predictive Factory on Windows

Before installing Predictive Factory, check the following requirements:

● The operating system is Microsoft Windows 64-bit.


● You have administrator rights.
● You have minimum of 3 GB available drive space.
● You have not installed any of the following software on the same machine:
○ SAP BusinessObjects Business Intelligence platform
○ SAP Lumira, server for teams
○ SAP Lumira, Edge edition
○ SAP Crystal Server
● Ports are free on the Windows system in the following ranges:
○ 6400-6800 for the CMS (default 6400)
○ 6100-6800 for the SIA

Predictive Analytics Installation Guide


Installing Predictive Factory PUBLIC 17
○ 2200-9800 for the SQL Anywhere database
○ 6200-6900 for the Web server (default 6405)
● If you are using Windows AD to authenticate users, Predictive Factory must be installed in the same
domain as Windows AD.

The Predictive Factory Setup program is contained within the self-extracting archive called
SAPPredictiveFactorySetup. The wizard first checks for platform requirements. If all requirements are
met, it installs the required resources on your machine.

1. Double-click SAPPredictiveFactorySetup.exe.

If any requirement is not met, a prerequisite page opens. Close the wizard and correct any missing
prerequisites before restarting the setup program.

If all the installation requirements are confirmed, the License Agreement page opens.
2. Review the license agreement, select the I accept the License Agreement check box, and click Next.
3. In the Configure Destination Folder page, enter the following information:
○ In the Destination Folder Information you can keep the suggested default, or click the browse button to
navigate to another folder for installation.
○ In the Administrator Logon Information section, enter the password for the administrator account.
The password must be at least 6 characters long and contain at least 2 of the following types of
characters: upper case, lower case, number, or punctuation mark.
○ Enter the same password in Confirm Password and click Next.
An error message displays if your password doesn't meet the security requirements. Click OK and
modify your password according to the rules provided in the message.
4. When the message appears that the installation was successful, click Finish.

Predictive Factory is automatically launched in a browser window. Log in using the administrator account and
the password you entered during the installation.

Note
A temporary license key is included in the software. Apply for a permanent license key immediately after
installing your SAP software on the SAP Support Portal at http://support.sap.com/keys-systems-
installations/keys.html .

4.3 Installing Predictive Factory on Linux

Before installing Predictive Factory, check the following requirements:

● You have installed the packagelibstdc++33-32bit to make the Predictive Factory Linux installer running
on Linux machine.
● You have installed mandatory libs 32bit glibc package containing libc.so.6 and 32bit
libstdc++ package containing libstdc++.so.5 to be able to run the Predictive Factory Linux
installer. The following table describes the commands to execute depending on the type of operating
system and the libs you want to install:

Predictive Analytics Installation Guide


18 PUBLIC Installing Predictive Factory
Operating System Commands to install the libs: Commands to install the libs:
32bit glibc package 32bit libstdc++ package
containing libc.so.6 (for ex­ containing libstdc++.so.5
ample: glibc-32bit-2.11.1-0.17.4). (for example: libstdc+
+33-3.3.3-11.9).

Suze 12.2 sudo zypper search glibc wget/libstdc+


+33-32bit-3.3.3-34.1.x86_
sudo zypper install
64.rpm
glibc-32bit
sudo rpm -Uvh libstdc+
+33-32bit-3.3.3-34.1.x86_
64.rpm

wget/libstdc+
+33-3.3.3-34.1.x86_64.rpm

sudo rpm -Uvh libstdc+


+33-3.3.3-34.1.x86_64.rpm

Suze 12.1 sudo zypper search glibc wget/libstdc+


+33-32bit-3.3.3-34.1.x86_
sudo zypper install
64.rpm
glibc-32bit
sudo rpm -Uvh libstdc+
+33-32bit-3.3.3-34.1.x86_
64.rpm

wget/libstdc+
+33-3.3.3-34.1.x86_64.rpm

sudo rpm -Uvh libstdc+


+33-3.3.3-34.1.x86_64.rpm

Suze 11.4 sudo zypper search


libstdc++

sudo zypper install


libstdc++33-32bit

Redhat 7.3 sudo yum search compat-


libstdc++

sudo yum install compat-


libstdc++-33.i686

sudo yum provides */


libstdc++.so.6

sudo yum install libstdc+


+ libstdc+
+-4.8.5-16.el7.i686

Predictive Analytics Installation Guide


Installing Predictive Factory PUBLIC 19
Operating System Commands to install the libs: Commands to install the libs:
32bit glibc package 32bit libstdc++ package
containing libc.so.6 (for ex­ containing libstdc++.so.5
ample: glibc-32bit-2.11.1-0.17.4). (for example: libstdc+
+33-3.3.3-11.9).

Redhat 7.2 sudo yum search compat-


libstdc++

sudo yum install compat-


libstdc++-33.i686

sudo yum provides */


libstdc++.so.6

sudo yum install libstdc+


+ libstdc+
+-4.8.5-16.el7.i686

sudo yum provides */


libstdc++.so.5

sudo yum install libstdc+


+ compat-libstdc+
+-33-3.2.3-72.el7.x86_64

Redhat 7.1 sudo yum search compat-


libstdc++

sudo yum install compat-


libstdc++-33.i686

sudo yum provides */


libstdc++.so.6

sudo yum install libstdc+


+ libstdc+
+-4.8.5-16.el7.i686

sudo yum provides */


libstdc++.so.5

sudo yum install libstdc+


+ compat-libstdc+
+-33-3.2.3-72.el7.x86_64

Predictive Analytics Installation Guide


20 PUBLIC Installing Predictive Factory
Operating System Commands to install the libs: Commands to install the libs:
32bit glibc package 32bit libstdc++ package
containing libc.so.6 (for ex­ containing libstdc++.so.5
ample: glibc-32bit-2.11.1-0.17.4). (for example: libstdc+
+33-3.3.3-11.9).

Redhat 7.0 sudo yum search compat-


libstdc++

sudo yum install compat-


libstdc++-33.i686

sudo yum provides */


libstdc++.so.6

sudo yum install libstdc+


+ libstdc+
+-4.8.5-16.el7.i686

sudo yum provides */


libstdc++.so.5

sudo yum install libstdc+


+ compat-libstdc+
+-33-3.2.3-72.el7.x86_64

Caution
To install the requirements, you must be logged as a root user.

Once you have fulfilled the requirements, you want to install the Predictive Factory Setup program called
<Predictive Factory archive>.

Caution
To install the Predictive Factory Setup program, you must not be logged as a root user.

1. Copy SAPPredictiveFactorySetup on your Linux machine: cp .../<Predictive Factory


archive>.tgz.
2. Unzip the SAPPredictiveFactorySetup: gunzip <Predictive Factory archive>.tgz.
3. Untar the SAPPredictiveFactorySetup: tar xvf <Predictive Factory archive>.tar.
4. Open the SAPPredictiveFactorySetup: cd <Predictive Factory archive>.
5. Launch the setup: ./setup.sh.

6. Select Installer Language English .


7. Read and validate the License Agreement.
8. Configure the Destination Folder.
9. Enter the credential and password.

Predictive Analytics Installation Guide


Installing Predictive Factory PUBLIC 21
Note
The administrator will use these credentials to log on to Predictive Factory.

10. Type Enter to launch the installation.


11. Once the installation is completed, you can start Internet Explorer and log on to Predictive Factory using
the URL: http://<LinuxMachineNameOrlpAddress>:6405/PA/login.html.

4.4 Starting and Stopping Predictive Factory Services - On


Windows

Predictive Factory Windows services are automatically started after installation. To stop and restart them if
needed:

1. In the Windows Start menu, chose SAP Business Intelligence > Stop Predictive Factory.

Note
Stop Predictive Factory may time out before it can verify that all the processes are stopped. In
this case, chose Stop Predictive Factory again until you see the confirmation that all processes
are stopped.

2. In Windows Start menu, chose SAP Business Intelligence > Start Predictive Factory.

4.5 Starting and Stopping Predictive Factory Services - On


Linux

If you need to start or stop Predictive Factory Services on Linux, type the commands:

Action Command

Start Predictive Factory Services <install dir>/sap_bobj/startservers

Stop Predictive Factory <install dir>/sap_bobj/stopservers

To have your servers start and stop automatically when the host machine is restarted, run the following script
as root user:

<install dir>/sap_bobj/init/setupinit.sh”

Note
If for some reasons you want to stop/start the databases, you will first have to stop Predictive Factory
services and then type the commands:

Predictive Analytics Installation Guide


22 PUBLIC Installing Predictive Factory
Action Command

Starting the CMS database <install dir>/sap_bobj/


sqlanywhere_startup.sh

Stopping the CMS database <install dir>/sap_bobj/


sqlanywhere_shutdown.sh

Starting the reporting database <install dir>/sap_bobj/


reportingdb_startup.sh

Stopping the reporting database <install dir>/sap_bobj/


reportingdb_shutdown.sh

4.6 Upgrading Predictive Factory

Caution
Do not uninstall the previous version of Predictive Factory: this will remove all objects created in Predictive
Factory (projects, tasks, modeling servers, etc..).

When you upgrade Predictive Factory, the setup program will recognize that a previous version is already
installed. To go on with the update, run the installation as described in the step Installing Predictive Factory on
Windows [page 17].

Note
Do not forget to clean your browser cache and to restart it after the upgrade.

Caution
In case you have declared an Automated Server as a modeling server in Predictive Factory, it is
recommended that you upgrade this Automated Server too. Indeed both the version of the Automated
Server and Predictive Factory must be the same. If not, an error will occur if the project members want to
import a model created with SAP Predictive Analytics.

4.7 Uninstalling Predictive Factory - On Windows

To remove Predictive Factory from your system, use the standard Windows uninstallation procedure:

1. In the Windows Control Panel Programs and Feature section , select SAP Predictive Analytics
Predictive Factory.
2. Click Uninstall.

Predictive Analytics Installation Guide


Installing Predictive Factory PUBLIC 23
Note
Predictive Factory uninstallation process does not delete the following two registry keys:

● HKEY_LOCAL_MACHINE\SOFTWARE\SAP BusinessObjects\Suite XI 4.0\CER


● HKEY_LOCAL_MACHINE\SOFTWARE\SAP BusinessObjects\Suite XI 4.0\CMS

There is no impact if you want to reinstall Predictive Factory or upgrade it. However, should you wish to, you
can delete these entries manually.

Note
When using an SAP HANA database, check that all objects created in the database have been deleted. For
the list of created objects, see related information.

Caution
Do not uninstall the previous version of Predictive Factory if you intend to simply do an upgrade. An
uninstallation will remove all objects created in Predictive Factory (projects, tasks, modeling servers, and so
on).

4.8 Uninstalling Predictive Factory - On Linux

It is not possible to uninstall Predictive Factory on Linux machine using an automatic way.

For more information, refer to the SAP note 2446171 .

However a workaround to solve the uninstall issue is available into the SAP note 2540065 and the SAP note
2557370 .

Predictive Analytics Installation Guide


24 PUBLIC Installing Predictive Factory
5 Installing SAP Predictive Analytics Server
for Windows

5.1 SAP Predictive Analytics Enterprise Edition

The enterprise edition of SAP Predictive Analytics is a three-tier client-server architecture. Communication
between the Automated Analytics server and the data is identical to the desktop edition, using either ODBC or
the native file system. For each client connection, a new Automated Analytics instance process is started on
the server. Depending on the server configuration, the process can be started with a specific system account,
or with the user account. Communication between the clients and server can be encrypted using SSL or TLS.

This configuration offers the following benefits:

● Users are authenticated because clients must log in before being able to use the modeling server. User
accounts can be configured to implement security policy.
● User activity monitoring and logging is possible and activated by default.
● Database connectivity needs to be configured only once on the server. Operating system rights can be
used to check access to the different resources (for example, modeling data).

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 25
● Resources are used more fully because each modeling session has a dedicated process. The process size
limit applies only to a single user.
● Network administration is simplified because all network traffic from the client is directed to the server.
This means only two TCP ports need to be opened for an Automated Analytics installation.

5.1.1 Minimum Skills Required to Perform the Installation on


Windows

The installation of SAP Predictive Analytics server requires some minimum skills that are described in the
following list.

The installation of the server must be performed by someone with a system administrator profile who:

● has administration rights on the machine,


● has the rights to set up services on the machine,
● knows how to set up ODBC drivers on Windows systems.

Before setting up the server, the person performing the installation must contact the database administrator to
obtain all necessary information about the database the application will be connected to.

5.1.2 Server Sizing

Refer to the Product Availability Matrix https://support.sap.com/content/dam/launchpad/en_us/pam/pam-


essentials/PAM_Predictive_Analytics.pdf to know the size needed for using the application server in your
environment.

5.1.2.1 Data Access

Data access for Automated Analytics is either through the ODBC layer of a relational database, or directly
against files in a file system. The application allows direct access to several types of files formats: SAS, SPSS,
Excel, Minitab, delimited text, and fixed length text. For efficiency, the application does not create a separate
analytics data store but reads data from the existing data sources, and saves output and models back into the
data source.

When working in a client-server environment, all data are accessed by the server. The data source (ODBC, ...)
must be installed (drivers, ...) and configured on the server.

Note
To know how to install and configure the data source, refer to the Predictive Analytics Data Access guide,
section Connecting to your Database Management System on Linux or Windows.

Predictive Analytics Installation Guide


26 PUBLIC Installing SAP Predictive Analytics Server for Windows
5.1.2.2 CPU Usage

The training of a single model consumes the resources of a single CPU for the duration of the training. Each
model being trained will spawn a separate thread, each being assigned to an available CPU. Training 2 models
simultaneously on the same CPU will approximately double the training time of each model. Therefore, 1 CPU
(500MHz or above) is recommended per concurrent user.

To know more about the recommended CPU, refer to the SAP Product Availability Matrix http://
service.sap.com/sap/support/pam .

5.1.2.3 Memory Consumption

A minimum of 512MB of RAM per concurrent user for the server is recommended. Additionally, each client
should have 128MB-256MB RAM to run the User Interface. In general, a model consumes about 100KB of RAM
per variable during training, or about 10MB per 100 variables. The application graphical user interface
consumes about 30MB of RAM.

By default the Data Cache is parameterized to use 500 MB RAM on a 64-bit architecture. The settings can be
changed in <INSTALLATION_DIR>/DataCacheManager.cfg.

5.1.2.4 Hard Disk Space

The server software requires about 700 MB for the server and about 200 MB for each client installed on
separate machines. No additional storage is required for data because the application does not create a
separate data store.

However, approximately 10MB to 100MB of disk space will be required per saved model, depending on the
model type and number of variables in the model. Note that Social models may require much more space. The
models are stored in database or on the file system either on the server or somewhere accessible from the
server. Be careful that users must be granted Write access to those spaces in order to be able to save their
work. As a best practice, modeling results are preferably saved as text files on a file system in the Development
phase, the database format is usually relevant only for Production models.

By default the Data Cache is parameterized to use 2 GB on the disk on a 64-bit architecture. The settings can
be changed in <INSTALLATION_DIR>/DataCacheManager.cfg.

To know more about the recommended Hard Disk space, refer to the SAP Product Availability Matrix http://
service.sap.com/sap/support/pam

5.1.3 Network Requirements

Each of the three components involved in the application, which are the name server, the SAP Predictive
Analytics server and the client application, can be located on a different machine. Note that in most installation
settings, both the name server and the SAP Predictive Analytics server are located on the same server.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 27
The communication protocol used under the CORBA framework is the TCP protocol. The first requirement is
that the:

● The SAP Predictive Analytics server must be able to access the name server.
● The client application must be able to access both the name server and the SAP Predictive Analytics
server.

From a network administration point of view it means that:

● The name server and the server will be assigned a specific TCP port. By default, the startup scripts we use
fix them, and they should probably set properly depending on the network strategy.
● Both of these ports must be accessible from the client machine. For example, network firewalls should
allow communication on these ports between client machine and the servers.

5.1.4 Java Runtime Environment for Client

You need to install Java Runtime Environment (JRE) with support for Java Web Start (JNLP Protocol), which is
available on the Java website. To know which version to install, refer to the SAP Product Availability Matrix
http://service.sap.com/sap/support/pam .

5.1.5 Java Web Start Deployment

In order to provide Java Web Start deployment of SAP Predictive Analytics Client applications, a Web Server
needs to be installed, configured and started on the machine where the Server is installed. Java Web Start has
been tested on the following server configurations:

● Windows IIS (Internet Information Service), on Microsoft Windows systems


● Apache Server (http://www.apache.org ). For more information on how to install a Apache server, refer
to the Apache documentation (http://httpd.apache.org/docs/current/en/install.html ).

5.2 Server Installation

Caution
For upgrade information when a previous version of the server or an SAP InfiniteInsight server is already
installed, please refer to the Predictive Analytics Administrator Guide, section Operations Guide provided on
the SAP Help Portal (SAP Help Portal).

Caution
As the Windows installer for server creates a Service and a User group, it must be run with an Administrator
user account .

Predictive Analytics Installation Guide


28 PUBLIC Installing SAP Predictive Analytics Server for Windows
1. Double-click the file PASERVER<version number>_<patch number>.EXE, for example
PASERVER2000_0.EXE for version 2.0, patch level 0.

The InstallShield Wizard is displayed.


2. Follow the instructions on the screen.
3. On the screen Select Destination Location, you can change the folder where SAP Predictive Analytics will be
set up.

Note
If a previous version of the server is already installed, the installer prevents you from using the same
directory. You need to select another folder or to uninstall the previous version first.

4. Click Next.
5. On the screen Select Components, select Authenticated Server in the drop-down list. The necessary
components are automatically checked. If needed, you can select additional components.

Option Description

Authenticated Server SAP Predictive Analytics server, which provides user authentication and connection

Java Web Start Necessary files to deploy a client through Java Web Start technology

Register Authenticated Server as Automatically register the Authenticated server as a Windows service
Services

Remote Assistant the client sample application that communicates with a modeling server.

6. When you have selected the components you want to install, click Next.
7. On the screen Server IP Address, indicate the IP address the clients should use to connect to this server. By
default, the displayed IP address is that of the machine, however you may need to modify it.
8. Click Next.

The screen Authentication Type appears.


9. Following the authentication type you have selected, the follow-on steps are different:

Authentication Type Next steps

System Authentication 1. Click Next. The screen Windows Domain is displayed.


2. Enter the name of the domain to be used for authen­
tication. The current domain name is displayed by de­
fault, you may need to change it. Contact your admin­
istrator to get the relevant information.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 29
Authentication Type Next steps

Encrypted File or File 1. Click Next.


2. The screen Administrator Account Selection is dis­
played. You have to choose one of the two possible
options:
○ Automatic: The screen Authenticated Server
Administrator Account Credentials will appear on
next step and you will have to provide with the
necessary credentials.
○ Manual: You will create the Authenticated Server
Administrator Account at the end of the installa­
tion process.

10. Click Next.

The screen Select Start Menu Folder is displayed.


11. Select the Start Menu folder in which you want to create the program shortcuts.
12. Click Next. The screen Select Additional Tasks allows you to create shortcuts on the desktop for the
components you have selected to install
13. Click Next.
The screen Ready to Install sums up the settings you have selected for the installation.
14. Check whether the program settings are correct. If so, click Install.

Note
By default the Data Cache repository is C:\Documents and Settings\<user>\Local Settings
\Temp. Check that you have enough memory on the disk (2 GB). If it is not the case, change the
repository or reduce the size in <INSTALLATION_DIRECTORY>/DataCacheManager.cfg.

Once the installation has completed, you must retrieve, register, and update your server license. You then
restart the server services.

● When the install is complete, register your license keycode. A temporary license key is included in the
software. Apply for a permanent license key immediately after installing your SAP software on the SAP
Support Portal at http://support.sap.com/keys-systems-installations/keys.html .
● To register the keycode, edit the License.cfg file in the folder C:\Program Files\SAP Predictive
Analytics\. Enter new keycodes in the format described in the License.cfg file:
KeyCode<TAB><TheReceivedKeycode>.
● Once you have registered your license keycode, in order for it to be effective, you need to restart the service
called SAP BUSINESSOBJECTS PREDICTIVE ANALYTICS – AUTOMATED ANALYTICS DIRECTORY, which
will automatically restart the other linked service: SAP BUSINESSOBJECTS PREDICTIVE ANALYTICS –
AUTOMATED ANALYTICS SERVER. This is described in the section Restarting the Services [page 36]

Caution
If you are using server authentication based on a fixed password list, encrypted or not, note that default
users are no longer declared at installation time. You need to create users as described in the Authentication
section.

Predictive Analytics Installation Guide


30 PUBLIC Installing SAP Predictive Analytics Server for Windows
5.3 Performing a Silent Installation

You can use the SAP Predictive Analytics server self-extractor to create a response file required for a silent
installation. Using a silent installation, system administrators can run a script from the command line to
automatically install SAP Predictive Analytics server on any machine in their system without the setup program
prompting them for information or displaying the progress bar. The silent installation is primarily geared
towards users with network administration roles. A silent installation is particularly useful when you need to
push multiple installations in your corporate network. Once you have created a silent installation response file,
you can add the silent installation command to your installation scripts.

1. Choose Start Run and type cmd to open a Command Prompt window.

% <filename.exe> /SAVEINF=C:\Tmp\SilentInstall.inf

Note
<filename.exe> represents the name of the .exe file.

The SAP Predictive Analytics server Setup program opens.


2. Follow the installation wizard to select your setup options.
3. On the Start installation page, choose Next.
The setup program writes your installation options to the SilentInstall.inf file, and closes.
4. To run the silent installation, open a Command Prompt window and enter the following command:

% <filename.exe> /SILENT /LOADINF= C:\Tmp\SilentInstall.inf

To prevent any pop-up window to be displayed during installation, replace /SILENT by /VERYSILENT in
the command above.

5.4 Checking Installation

Once the installation is done, you can run the script KxCheckConfig.bat located in the installation directory.
This script is used to communicate with the Support team in case of issues.

Double-click the file KxCheckConfig.bat. It automatically generates a KxCheckConfig.log text file that
contains the full description of the machine and a snapshot of configuration and log files from the installation
directory.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 31
5.5 Configuration

If you need to modify the configuration of the Authenticated Server, you will need to modify the following
options. The following table summarizes the location of parameters in a distribution:

Description File Location

TCP ports (and eventually IP address) to be used by the ${INSTALLATION DIR}\EXE\Servers


Server, and SSL Activation switch \AuthServer\KxServiceManager.ini

Name of Log configuration file, Authentication mode and ${INSTALLATION DIR}\EXE\Servers


other Authentication parameters \AuthServer\KxAuth.cfg

List of users if File Authentication is selected ${INSTALLATION DIR}\EXE\Servers


\AuthServer\users

Mapping between authenticated users and physical users if Not available


specified in the KxAuth.cfg

Default configuration file used for Automated Analytics in­ ${INSTALLATION DIR}\EXE\Servers\CORBA
stances.
\KxCORBA.cfg

See the related topic for a description of the available op­


tions for each instance.

Note
After modifying the configuration, you will need to restart the server.

5.5.1 Basic Configuration

The basic server configuration option can be found in the file KxAuth.cfg located in the folder $
{INSTALLATION DIR}\EXE\Servers\AuthServer\.

############# AuthServer configuration


AuthServer.TimeOut=4000
AuthServer.MaxConnections=50
AuthServer.ExeName=KxCORBA3
AuthServer.ExeStartDirectory=../CORBA
AuthServer.DomainName=KXEN01
AuthServer.Admin=root
AuthServer.Admin=admin

The available options are detailed in the following table.

Predictive Analytics Installation Guide


32 PUBLIC Installing SAP Predictive Analytics Server for Windows
Option Default Value Description

AuthServer.TimeOut 4000 Internal timeout parameter used when


starting internal sever instance. If the
started instance (after user authentica­
tion) can not be reached after this
specified number of milliseconds, then
a time-out occurs and the connection
fails. In most cases it is due to a bad
configuration or installation of the in­
stance.

AuthServer.MaxConnections 50 Maximum number of concurrent con­


nections on the server.

This can be used to hard-limit the num­


ber of Instance processes launched at
any time.

AuthServer.ExeName KxCORBA3 Name of the server instance executable


file. It should not be changed.

AuthServer.ExeStartDirecto ../CORBA Name and path of the instance execu­


ry tion directory. It should not be changed.

AuthServer.InstanceConfigF KxCORBA.cfg Can be used to specify a different con­


ile figuration file for server instances
started by SAP Predictive Analytics
server. If the name contains the string
%u it will be replaced by the authenti­
cated user name starting the instance,
so that each user can have a separate
configuration. It can be used to specify
different Data Sources (FileStoreRoot)
or different License for each user.

AuthServer.InstanceConfigP . Specify the path to be searched for


ath configuration file. By default the config­
uration file are located in the same di­
rectory that the KxCORBA3 executable,
but this allows you to give a specific di­
rectory for all configuration files.

AuthServer.DomainName KXEN01 Name of the Windows domain to be


used for Windows Authentication.

AuthServer.Admin root Logging name(s) of the server adminis­


trators. Server administrators can see
AuthServer.Admin admin and disconnect all the user connections
(in the Administration Console).

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 33
5.5.2 Configuration of Server Instances
The configuration of an instance, which consists of a CORBA process, requires to set:

● the proper TCP port,


● and the proper configuration for the server itself.

The server instance can be configured using the configuration file called KxCORBA.cfg, located in the folder
<installation_folder>\EXE\Servers\CORBA\. This file is loaded at start-up, and includes several Key/
Values properties. The first line should not be edited, blank lines are accepted, and each valid line should
include: one of the accepted keyword, the equal sign and the associated value.

The main usual configuration step is the definition of the FileStoreRoot to be used on the local system. It is
generally a good idea to set up entries for common data storage (for example, c:/data) and also a specific
entry to store built models (for example, c:/Models).

The following table describes the available properties.

Option Description

KxAdmin The root name of the file or table used to store the models in
a directory or ODBC store. This file is automatically gener­
ated when "commit" or "save" operations are performed on
Predictive Analytics models. It is used to quickly retrieve the
list of available models in a store. By default it is KxAdmin
so that it will creates KxAdmin.txt file or KxAdmin ta­
bles.

MessageFile Such a message file allows for a quick and easy localization
of applications. New message files can be added for new lan­
guage. The convention used for such message file name is:
<Name>_<language>.umsg, where language is an ISO
code of a language (for example, fr for French, us for English
American, uk for English).

MessageDirectory This option allows you to load all message files from a single
directory. All .umsg file found in that directory will be loaded
as message.

FileStoreRoot Such entries are used by the server to retrieve the main file
store to be displayed when client applications ask for top-
level file directory.

Config Load another configuration file. This is used to load the Li­
cense file, or to load some Data Access library, but could be
used also to base different configuration files (one for each
user) on some common settings.

LogConf This option gives the name of the log system configuration
file to be used. For a description of the logging configuration
file itself, refer to the SAP Predictive Analytics Operation
Guide, section Predictive Analytics Logging System.

Example
To display two directories in the file selection drop down, add the following lines:

FileStoreRoot=c:/data

Predictive Analytics Installation Guide


34 PUBLIC Installing SAP Predictive Analytics Server for Windows
FileStoreRoot=c:/Models

5.5.3 Port Configuration

The SAP Predictive Analytics default installation uses two TCP ports for Client/Server communication.

By default the Name Server is set on port 12345 and the SAP Predictive Analytics server is set on port 12347.

If these TCP ports cannot be used (for network policies reasons or because they are already used), then they
must be changed. To change the TCP ports when installing the application, you will need to:

1. stop both services,


2. update the configuration files,
3. clean the log files up,
4. start again both services.

Caution
If the deployment of the client is managed by Web Start, refer to the Web Start documentation to update it.

5.5.3.1 Stopping the Services

To Stop the SAP Predictive Analytics Server Service

1. Go to the Windows Services panel: Start Settings Control Panel Administrative Tools Services .
2. Right-click SAP Predictive Analytics - Automated Analytics Server.
3. In the contextual menu, select Stop.

To Stop the Name Service

1. Go to the Windows Services panel: Services Start Settings Control Panel Administrative Tools .
2. Right-click SAP Predictive Analytics - Automated Analytics Directory.
3. In the contextual menu, select Stop.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 35
5.5.3.2 Updating the Configuration Files

To Update the Configuration File for theSAP Predictive


Analytics Server

1. Go to <INSTALLATION DIR>\EXE\Servers\AuthServer.
2. Open the file KxServiceManager.ini in a text editor.
The port is the last element of the following parameters:

-ORBendPoint giop:tcp::12347.

-ORBendPoint giop:tcp::192.168.1.1:12347.

The default value is 12347.


3. Save the new port number and close the editor.

To Update the Configuration Files for the Name Server

1. Go to <INSTALLATION DIR>\EXE\Servers\AuthServer.
2. Open the file KxServiceManager.ini in a text editor.
The port is the last element of the parameter NameService=corbaname::localhost:12345 . The
default value is 12345.
3. Save the new port number and close the editor.
4. Go to <INSTALLATION DIR>\EXE\Namesrv.
5. Open the file KxServiceManager.ini in a text editor.
The port is the last element of the following parameters:

-start 12345.

-ORBendPointNoPublish giop:tcp::12345.

The default value is 12345.


6. Save the new port number and close the editor.

5.5.3.3 Restarting the Services

To Restart the Name Service

1. Go to the Windows Services panel: Start Settings Control Panel Administrative Tools Services .
2. Right-click SAP Predictive Analytics - Automated Analytics Directory.
3. In the contextual menu, select Restart.

Predictive Analytics Installation Guide


36 PUBLIC Installing SAP Predictive Analytics Server for Windows
To Restart the SAP Predictive Analytics Server Service

1. Go to the Windows Services panel: Start Settings Control Panel Administrative Tools Services
2. Right-click SAP Predictive Analytics - Automated Analytics Server.
3. In the contextual menu, select Restart.

5.5.4 Communication Channel Encryption

Caution
Connection through Transport Layer Security (TLS) or Secure Sockets Layer (SSL) is not activated by
default when you install SAP Predictive Analytics Server. You have to enable it explicitly.

To enable the encryption through TLS or SSL, you’ll need the following elements:

● A Certification Authority certificate (CA certificate)


● The server private key
● The server certificate
● The server pass phrase (optional)

Note
If your company can’t provide you with these elements, you can generate self-signed certificates with
OpenSSL . Note that for security reasons, it is recommended to use certificates issued by a certificate
authority.

5.5.4.1 Enabling the Encryption for Automated Analytics


Server

To deploy the certificates on the Automated Analytics server, you’ll need to combine the private key and the
server certificate into a single file with the .cer extension.

1. Edit the KxServiceManager.ini file located in the installation folder of the Automated Analytics server.
For example, <InstallationFolder>\Server\Servers\AuthServer.
2. Locate the line: CommandLine = E:\apps\PredictiveAnalytics\Server\EXE\Servers
\AuthServer\KxAuthServer.exe [...].
3. Add the following options at the end of the command line: -cacert %CACertificate% -servercert
%ServerCertificate% -password %PassPhrase%.
4. Replace the placeholders by the following values:

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 37
Placeholder Name Value

%CACertificate% Name of the file containing the CA certificate

%ServerCertificate% Name of the file combining the private key and the server certificate
%PassPhrase% The server private key pass phrase. Only use this option if you have set up a passphrase
for the server.

5. Replace tcp by ssl in the following options:


○ -ORBendPointNoPublish giop:tcp:<ADDRESS>:<PORT>
○ -ORBendPointPublish giop:tcp:<ADDRESS>:<PORT>

You should then have:


○ -ORBendPointNoPublish giop:ssl:<ADDRESS>:<PORT>
○ -ORBendPointPublish giop:ssl:<ADDRESS>:<PORT>
6. Save the file.
7. Restart the Windows service named SAP Predictive Analytics - Automated Analytics Server

5.5.5 Authenticated Server Based on the System

This is the recommended setting. When you have selected this type of authentication during the installation,
the configuration files are automatically updated according to the information provided.

On Windows systems, the System Authentication is based on the Windows Domain Login services. To configure
the Authenticated Server to use Windows authentication, the Authenticator Name has to be System and the
Domain key must be properly configured. For example here is a configuration file that authenticates users on
Windows domain MYDOM01:

AuthServer.DomainName=MYDOM01
############## Authenticator configuration
Authenticator.Name=System
#Authenticator.System.PamServiceName=login

5.5.5.1 Adding Windows Users to the User Group

In order to allow users to use the SAP Predictive Analytics Server, they must be added to the users group
KxenUsers created during the installation process.

1. Right-click My Computer.
A contextual menu is displayed.
2. Select the Manage option.
The window Computer Management opens.

Note
On Windows Server 2016, when you select the Manage option, the window Server Manager opens.
Select Tools Computer Management to open the window Computer Management.

Predictive Analytics Installation Guide


38 PUBLIC Installing SAP Predictive Analytics Server for Windows
3. Open the item System Tools > Local Users and Groups > Groups.

4. Double-click the KxenUsers group.


The group properties window opens.
5. Click the Add button.
The window Select Users, Computers or Groups opens.
6. Enter the list of users that should be granted the right to use the server in the field Enter the object names
to select.

7. Click OK.

5.5.6 Authenticated Server Based on a Fixed Encrypted


Password List

An authenticated server based on a fixed encrypted password list works on same principle than a standard file­
based one except that it supports two types of passwords, which are encrypted and non-encrypted. In the
password store, the encrypted passwords are preceded by a star (*).

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 39
Sample Code
The following example shows a fixed encrypted password list containing both encrypted and non-encrypted
passwords:

Key=Value
admin=admin
user1=user1
user2=user2
port=*901555fb06e346cb065ceb9808dcfc25

To implement this type of authentication, a password store such as a 'users' password file with passwords
stored in clear must be used exactly as in the standard file­based authentication. At this point, it works exactly
as an authenticated server based on a fixed file password list. Differences come when a user changes their
password, which is then automatically stored as an encrypted one.

It can be used with an existing 'users' password file used on a server based on the 'File' authentication.
Passwords are progressively encrypted when users change them using the Remote Control Panel.

1. Open the file KxAuth.cfg located in the folder <INSTALLATION DIR>/KxAuthServer.


2. In the section Authenticator configuration, check that the line Authenticator.Name=Encrypted
is not commented , that is, does not begin with a hash sign (# ):
3. You can customize the options listed in the following table.

Options for the Authentication Based on an Encrypted Password List


Option Description Default Value

Authenticator.Encrypted.F Sets the name of the store containing users


ileName the user logins and passwords

Authenticator.Encrypted.S Indicates how are stored the user log­ KXEN.FileStore


toreType ins and passwords, for example in a
flat file, in a database, in a SAS file This value means that the data is
stored in a flat file.

Authenticator.Encrypted.S Sets the path where the user data is . (dot)


toreName stored
This value means that the data is
stored in the same folder as the
KxAuth.cfg file.

The following lines show the settings to use in the KxAuth.cfg file to set up an encrypted authentication.

#Authenticator.Name=File
#Authenticator.File.FileName=users
#Authenticator.File.StoreType=KXEN.FileStore
#Authenticator.File.StoreName=.
#Authenticator.File.User=
#Authenticator.File.Password=
Authenticator.Name=Encrypted
Authenticator.Encrypted.FileName=users
Authenticator.Encrypted.StoreType=KXEN.FileStore
Authenticator.Encrypted.StoreName=.
#Authenticator.Encrypted.User=
#Authenticator.Encrypted.Password=

Predictive Analytics Installation Guide


40 PUBLIC Installing SAP Predictive Analytics Server for Windows
#Authenticator.Name=System
#Authenticator.System.PamServiceName=login

5.5.7 Authenticated Server Based on a Fixed Password List

When you have selected this type of authentication during installation, the authentication is based on a fixed
file named users, which contains the login and password of each user and is located in the folder
<INSTALLATION DIR>\EXE\Servers\AuthServer\users.

To activate and customize this option, follow the following instructions:

1. Open the file KxAuth.cfg located in the same folder as the users file.
2. Uncomment the line Authenticator.Name=File by removing the hash sign (#) at the beginning of the
line.
3. You can then customize the following options:

Option Description

Authenticator.File.FileName This option allows you to set the name of the store containing the users'
logins and passwords. By default, the store name is users .

Authenticator.File.StoreType This option allows you to indicate how are stored the user logins and pass­
words, for example in a flat file, in a database, in a SAS file. The default
value is KXEN.FileStore, which means that the data is stored in a flat
file.

Authenticator.File.StoreName This option allows you to set the path where the user data is stored. The
default value is a dot (.), which means that the data is stored in the same
folder as the KxAuth.cfg file.

The following lines show the settings to use in the KxAuth.cfg file to set up a fixed password list
authentication.

############## Authenticator configuration


#Authenticator.Name=Void
Authenticator.Name=File
Authenticator.File.FileName=users
Authenticator.File.StoreType=KXEN.FileStore
Authenticator.File.StoreName=.
Authenticator.File.User=
Authenticator.File.Password=
#Authenticator.Name=System
#Authenticator.System.PamServiceName=login

5.5.8 No Authentication

Another authentication scheme called "Void" is available, but it should only be used with care, generally for
technical integration with other system. This "Void" authenticator accept any combination of user and
password as a valid one, it does not perform any actual identity check.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 41
This authentication scheme can be used when the connections of the authenticated server are only created by
another software components, such as web server, or another application where identity is already managed.
The other software components might want to create connection as some user, but do not want or need to do
an actual authentication.

The goal of specifying a user in a void authentication connection is to be able to trace the work and process on
a per-user basis since most logs and trace information reference the connected user.

Note
Although it is possible to enter an empty user/password in such a case, it is not a recommended practice for
traceability and efficiency purpose: all logs will be mixed in the same file, without any distinction.

If this kind of authentication is used, the software integrator or application developer should ensure that the
server cannot be physically accessed another way by using proper firewalls, network configuration, …, as
allowing a connection from another route will prevent data security.

Also in such a configuration, it is strongly recommended to not start the server as administrator or super user,
as no impersonification will be done after authentication. The identity of the instance started will be the one
used to start the server process itself.

5.6 Managing the Server on Windows

All users that will be allowed to connect to SAP Predictive Analytics should have the right to “Logon as Service”.
To simplify this procedure, the default installation procedure creates a Windows Group of users “KxenUsers”,
with the appropriate rights. Each user granted to connect to SAP Predictive Analytics should be added to this
group (using Windows administration tools).

Also, the users should be allowed to write log files to the temporary directory. The server instances started will
typically write their log files in a temporary directory, which depends on some user environment variables (its
uses the first defined among: TMP, TEMP, or USERPROFILE environment variable, or by default the Windows
temporary directory).

There are two ways to manage the server on a Windows platform:

● By using Windows Services, if the server has been registered as a Windows service,
● By launching a bat file (for debugging purpose).

Related Information

Launching a .bat File [page 43]


Using Windows Services [page 43]

Predictive Analytics Installation Guide


42 PUBLIC Installing SAP Predictive Analytics Server for Windows
5.6.1 Launching a .bat File

For debugging or testing purpose, the runserver.bat file is also available in the folder <INSTALLATION
DIR>\EXE\Servers\AuthServer and allows you to manually start the server without using SSL.

5.6.2 Using Windows Services

If the server has been registered as a Windows service during the installation, use Windows Services to start.

1. Open the Windows Services administration panel.


2. In the Services panel, the two services created by SAP Predictive Analytics are:
○ Automated analytics server
○ Automated analytics directory
3. Right-click the server name.
A contextual menu is displayed offering several actions.
4. Use the contextual menu to start, pause, stop or restart the server, depending on its current status.

Note
When the server is registered as a Windows Service, it will be run under the “Local System” user name.
By default, this user has the correct security policy to start the server instance as another
user(Windows security policy “Replace a process level token”). However, if you need to change the
name of the user who has started this service, the correct security policy “Replace a process level
token” must be set for this user. For more information about process level token refer to the Automated
Analytics Server Installation Guide for Windows.

5.7 Deploying the Client

SAP Predictive Analytics client applications are Java graphical user interface that connect to SAP Predictive
Analytics servers. The client application can be installed:

● Through Java Web Start: in this mode, a Web server needs to be installed and configured on the machine
hosting the SAP Predictive Analytics server. It will deliver, through web pages, the necessary content in
order for the client computer to run the application.
● With standalone installers on each targeted client system.

We recommend using the Web Start facility, as it will allow automatic client version migration in case of a server
upgrade.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 43
Using Java Web Start

Java Web Start is an application launcher for the client application. It allows you to start the client from a web
site.

The updates are automatically managed through the Java Web Start system. This system avoids having to
manually install the client on every computer of the local network. As an extra, if the client is updated on the
server, all the distributed applications will automatically be updated too.

The installation and configuration of Java Web Start is described in the section Installing Java Web Start .

Manually Install Clients for SAP Predictive Analytics Server

For information on the installation and configuration of an SAP Predictive Analytics client, read the section
Installing SAP Predictive Analytics Client.

5.8 Starting the Administration Console

The Remote Control Panel is a component that allows the server administrator to manage the users activity
on the server by:

● Listing the active connections, with information such as user, last activity time stamp, number of current
models
● Listing the state of models being within each connection
● Ending an existing connection. This should be used with care, for example in order to free the associated
process when some pending connection remains active while a client crashed.

To start the Remote Control Panel, follow the instructions below:

In the Start menu, select All Programs SAP Business Intelligence SAP Predictive Analytics Server
Remote Control Panel .

5.9 Logging

Automated Analytics offers a logging system which is activated by default and can be deactivated if needed.
The goal of this logging system is for an administrator or a support team to be able to trace and analyze the
application activity on within the kernel. Note that the activity of the client user interfaces is only seen through
events fired within the kernel.

It can be deactivated within a server, or in the Shell utility.

Predictive Analytics Installation Guide


44 PUBLIC Installing SAP Predictive Analytics Server for Windows
The default Log Files for the server are:

● kxen_auth_log.txt: the default activity trace for the server, which can be used mainly for debugging
purpose
● kxenusagelog.txt: the goal of this log file is to have a simpler monitoring of overall modeling server
activity. This log will include one line for the following events: user connection, user disconnection, training
a model, applying a model, saving a model.

These files are located in the server user temporary directory (usually "C:\Documents and Settings\user
\Local Settings\Temp" or “C:\WINDOWS\Temp” when the SAP Predictive Analytics server is started as a
Service)

The logging system is described in detail in the Operations Guide.

5.10 Troubleshooting

This topic lists all troubleshooting in case a problem occurs during the configuration of Automated Analytics
Server.

Admin Console

If Java Runtime Environment is not available for you platform, you have to use a Windows client to start the
Admin Console.To know which Java version to install, refer to the SAP Product Availability Matrix http://
service.sap.com/sap/support/pam .

Java Web Start

If parameters have been modified, the users (or admin) will need to clean the cache to insure a correct running.

5.11 Setting User Security for Windows

In order to be able to create processes as other user, the user account used to run Automated Analytics Server
process must have the privilege “Replace a process Level Token”.

This procedure should be used in case of a specific configuration or startup of Automated Analytics Server,
since the default installer start it as a Windows Service using the account Local System, which already owns
this right.

To set this privilege, follow the instructions below:

1. Go to the Control Panel.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Windows PUBLIC 45
2. In the panel Administrative Tools, open the Local Security Policy.
3. In the panel Local Security Settings, select the option Local Policies > User Right Assignment.
4. In the right panel, check that the account used to run the application is listed for the policy Replace a
process level token.

5. If the account is not listed, it can be added by double clicking the policy and adding the proper user.
6. To commit this modification then logoff/logon.

Predictive Analytics Installation Guide


46 PUBLIC Installing SAP Predictive Analytics Server for Windows
6 Installing SAP Predictive Analytics Server
for Linux

6.1 SAP Predictive Analytics Server

6.1.1 SAP Predictive Analytics Enterprise Edition

The enterprise edition of SAP Predictive Analytics is a three-tier client-server architecture. Communication
between the Automated Analytics server and the data is identical to the desktop edition, using either ODBC or
the native file system. For each client connection, a new Automated Analytics instance process is started on
the server. Depending on the server configuration, the process can be started with a specific system account,
or with the user account. Communication between the clients and server can be encrypted using SSL or TLS.

This configuration offers the following benefits:

● Users are authenticated because clients must log in before being able to use the modeling server. User
accounts can be configured to implement security policy.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 47
● User activity monitoring and logging is possible and activated by default.
● Database connectivity needs to be configured only once on the server. Operating system rights can be
used to check access to the different resources (for example, modeling data).
● Resources are used more fully because each modeling session has a dedicated process. The process size
limit applies only to a single user.
● Network administration is simplified because all network traffic from the client is directed to the server.
This means only two TCP ports need to be opened for an Automated Analytics installation.

6.1.2 Requirements

6.1.2.1 Minimum Skills Required to Perform the


Installation

The installation of Automated Analytics server requires some minimum skills that are described in the following
list.

The installation of Automated Analytics server must be performed by someone with a system administrator
profile who:

● knows the main Linux administration commands (managing users, groups, and so on),
● has administration rights on the machine,
● has the rights to set up services on the machine,
● knows how to set up ODBC drivers on Linux systems.

Before setting up the server, the person performing the installation must contact the database administrator to
obtain all necessary information about the database the application will be connected to.

6.1.2.2 User Accounts Requirements

Before installing the server, you must create:

● a kxenusers group
● a kxenadmin user, belonging to the kxenusers group.

Note
Any users that will access to the server need to be added to this group.

6.1.2.3 Server Sizing

Refer to the Product Availability Matrix https://support.sap.com/content/dam/launchpad/en_us/pam/pam-


essentials/PAM_Predictive_Analytics.pdf to know the size needed for using the application server in your
environment.

Predictive Analytics Installation Guide


48 PUBLIC Installing SAP Predictive Analytics Server for Linux
6.1.2.3.1 Data Access

Data access for Automated Analytics is either through the ODBC layer of a relational database, or directly
against files in a file system. The application allows direct access to several types of files formats: SAS, SPSS,
Excel, Minitab, delimited text, and fixed length text. For efficiency, the application does not create a separate
analytics data store but reads data from the existing data sources, and saves output and models back into the
data source.

When working in a client-server environment, all data are accessed by the server. The data source (ODBC, ...)
must be installed (drivers, ...) and configured on the server.

Note
To know how to install and configure the data source, refer to the Predictive Analytics Data Access guide,
section Connecting to your Database Management System on Linux or Windows.

6.1.2.3.2 CPU Usage

The training of a single model consumes the resources of a single CPU for the duration of the training. Each
model being trained will spawn a separate thread, each being assigned to an available CPU. Training 2 models
simultaneously on the same CPU will approximately double the training time of each model. Therefore, 1 CPU
(500MHz or above) is recommended per concurrent user.

To know more about the recommended CPU, refer to the SAP Product Availability Matrix http://
service.sap.com/sap/support/pam .

6.1.2.3.3 Memory Consumption

A minimum of 512MB of RAM per concurrent user for the server is recommended. Additionally, each client
should have 128MB-256MB RAM to run the User Interface. In general, a model consumes about 100KB of RAM
per variable during training, or about 10MB per 100 variables. The application graphical user interface
consumes about 30MB of RAM.

By default the Data Cache is parameterized to use 500 MB RAM on a 64-bit architecture. The settings can be
changed in <INSTALLATION_DIR>/DataCacheManager.cfg.

6.1.2.3.4 Hard Disk Space

The server software requires about 700 MB for the server and about 200 MB for each client installed on
separate machines. No additional storage is required for data because the application does not create a
separate data store.

However, approximately 10MB to 100MB of disk space will be required per saved model, depending on the
model type and number of variables in the model. Note that Social models may require much more space. The

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 49
models are stored in database or on the file system either on the server or somewhere accessible from the
server. Be careful that users must be granted Write access to those spaces in order to be able to save their
work. As a best practice, modeling results are preferably saved as text files on a file system in the Development
phase, the database format is usually relevant only for Production models.

By default the Data Cache is parameterized to use 2 GB on the disk on a 64-bit architecture. The settings can
be changed in <INSTALLATION_DIR>/DataCacheManager.cfg.

To know more about the recommended Hard Disk space, refer to the SAP Product Availability Matrix http://
service.sap.com/sap/support/pam

6.1.2.4 Network Requirements

Each of the three components involved in the application, which are the name server, the SAP Predictive
Analytics server and the client application, can be located on a different machine. Note that in most installation
settings, both the name server and the SAP Predictive Analytics server are located on the same server.

The communication protocol used under the CORBA framework is the TCP protocol. The first requirement is
that the:

● The SAP Predictive Analytics server must be able to access the name server.
● The client application must be able to access both the name server and the SAP Predictive Analytics
server.

From a network administration point of view it means that:

● The name server and the server will be assigned a specific TCP port. By default, the startup scripts we use
fix them, and they should probably set properly depending on the network strategy.
● Both of these ports must be accessible from the client machine. For example, network firewalls should
allow communication on these ports between client machine and the servers.

6.1.2.5 Java Runtime Environment for Client

You need to install Java Runtime Environment (JRE) with support for Java Web Start (JNLP Protocol), which is
available on the Java website. To know which version to install, refer to the SAP Product Availability Matrix
http://service.sap.com/sap/support/pam .

6.1.2.6 Java Web Start Deployment

In order to provide Java Web Start deployment of Automated Analytics client applications, a Web server needs
to be installed, configured and started on the machine where Automated Analytics server is installed.

Automated Analytics Java Web Start has been tested on the following server configurations:

● Windows IIS (Internet Information Service), on Microsoft Windows systems


● Apache Server (http://www.apache.org ). For more information on how to install an Apache server, refer
to the Apache documentation (http://httpd.apache.org/docs/current/en/install.html ).

Predictive Analytics Installation Guide


50 PUBLIC Installing SAP Predictive Analytics Server for Linux
6.1.2.7 RedHat

To know which version is supported by SAP Predictive Analytics, refer the Product Availability Matrix athttp://
service.sap.com/sap/support/pam .

Note
as a prerequisite to Linux products, you need to install libstdc++.so.6.0.19.

6.1.3 Installation

6.1.3.1 Installing the Server

Caution
For upgrade information when a previous version of the server or a SAP InfiniteInsight server is already
installed, please refer to the Predictive Analytics Administrator Guide, section Operations Guide provided on
the SAP Help Portal (http://help.sap.com/pa/).

6.1.3.1.1 Standard Installation

There is no specific installation procedure for the executable files themselves; they should simply be
uncompressed and extracted in the correct location.

Note
For additional information on how to install SAP Predictive Analytics server on AIX or Solaris platforms,
please refer to the SAP Note 2174673 .

1. Create a group kxenusers and an account kxenadmin.


2. Log as kxenadmin.
3. Run the command umask 007.
4. From the home directory of the kxenadmin user run the command gunzip -c </directory/
distribution_file.extension> | tar -xvf -.
This command creates a directory named AutomatedAnalytics_<OS_Name> in the home directory of
kxenadmin user.
5. If you need to create a global shortcut (outside of the kxenadmin home directory)
a. Log in as root user.
b. Go to the directory where you want to create the shortcut.
c. Create a symbolic link to the installation directory of Automated Analytics.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 51
Example
In the instructions, we will consider that Automated Analytics features are set up in the /softs/
directory:cd /softs

Example of a symbolic link to the installation directory of Automated Analytics.


ln -s /home/kxenadmin/AutomatedAnalytics_<OS_Name> AutomatedAnalytics

When the install is complete, register your license keycode. A temporary license key is included in the software.
Apply for a permanent license key immediately after installing your SAP software on the SAP Support Portal at
http://support.sap.com/keys-systems-installations/keys.html .

To register the keycode, edit the License.cfg file located under the root folder where the SAP Predictive
Analytics tar archive was extracted. Enter new keycodes in the format described in the License.cfg file. You
have to stop then start the server for this to be effective.

Caution
If you are using server authentication based on a fixed password list, encrypted or not, note that default
users are no longer declared at installation time. You need to create users as described in the Authentication
section. Refer to the related information below.

6.1.3.1.2 Installation Directory Content

Once the installation has been completed the following folders will be available:

Folder Content

Doc documentation

KJWizardCORBA the remote assistant for client/server modeling

KJWizardJNI the modeling assistant, which is a Java GUI based on Java Native Inter­
face, is available on some Linux platforms only - for the list of platforms,
see the PAM.

KxShell a command line interpreter to test the C++ API

KxenServer the server instance

Messages messages files used by the kernel

NameServer a name server

Samples some sample data files

libs libraries

include include and IDL files

KxNodeCodeGenerator node code generator

KxCORBAShell remote Shell utility

Predictive Analytics Installation Guide


52 PUBLIC Installing SAP Predictive Analytics Server for Linux
Folder Content

KxAuthServer Automated Analytics server: authenticate users

WebStart simplify the deployment of the Automated Analytics client through a


Web page

ODBCShell a simple command line utility to test ODBC connections

odbcconfig folder containing ODBC configuration template files, and where the
ODBC configuration should be set

EXE runtimes for the code generator

Resources resources required by the kernel (messages, additional configuration


files, ...)

KCG stand-alone C source code generator

6.1.3.1.3 Start-up Script Installation

Once SAP Predictive Analytics has been installed, the folder /softs/AutomatedAnalytics_<OS_Name>
exists:

1. Edit /softs/AutomatedAnalytics_<OS_Name>/KxenServer_initd to put the proper path into the


shell variable <KXEN_INSTALLATION_PATH>.
2. Copy the updated version of /softs/AutomatedAnalytics_<OS_Name>/KxenServer_initd
into /etc/init.d or /etc/rc.d

# cp /softs/AutomatedAnalytics_<OS_Name>/KxenServer_initd /etc/init.d

3. Create the link from /etc/init.d/KxenServer_initd to the folders /etc/rc?.d

# cd /etc/rc3.d; ln ../init.d/KxenServer_initd S93KxenServer_initd


# cd /etc/rc0.d; ln ../init.d/KxenServer_initd K93KxenServer_initd
# cd /etc/rc1.d; ln ../init.d/KxenServer_initd K93KxenServer_initd
# cd /etc/rc2.d; ln ../init.d/KxenServer_initd K93KxenServer_initd
# cd /etc/rc5.d; ln ../init.d/KxenServer_initd K93KxenServer_initd

6.1.3.2 Checking the Installation

Once the installation is done, you can use the script KxCheckConfig.sh (located in the SAP Predictive
Analytics root directory) to check the installation and verify if the required OS patches are installed on the
machine.

This script can also be used to communicate with the support by generating a report of the installation in a
report.txt text file. This file contains the full description of the machine and a snapshot of the configuration
and log files from the installation directory.

./KxCheckConfig.sh -v > report.txt

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 53
6.1.4 Configuration

The configuration of the server is included in several files. The following table summarizes the location of the
parameters in a distribution:

Description File Location

TCP ports and possibly IP addresses to be used by the ${SAP_PA_HOME}/KxenCorbaConfig.sh


server, and SSL activation switch

Name of the log configuration file, authentication mode and ${SAP_PA_HOME}/KxAuthServer/KxAuth.cfg


other authentication parameters

List of users if the file authentication is selected ${SAP_PA_HOME}/KxAuthServer/users

Mapping between authenticated users and physical users if ${SAP_PA_HOME}/KxAuthServer/mapping


specified in the KxAuth.cfg

Default configuration file used for server instances ${SAP_PA_HOME}/KxenServer/KxCORBA.cfg

For a description of the available option for each instance,


see the related information below.

Note
If you modify the configuration or update the license, you will need to restart the server.

6.1.4.1 Assisted Configuration

A script is provided to help you configure the server after installation.

It goes through all the steps listed in the related links section below and saves the parameters in a .ini file that
will allow you to automatically reuse the parameters when installing a new version. The script modifies all the
necessary configuration files and creates back-up copies of the default files as well as of the modified files.

Files and folders related to the configuration script


Description Location File

Configuration script ${SAP_PA_HOME}/ PostInstallScript.sh


PostInstall/

File containing the parameters provided ${SAP_PA_HOME}/ PostInstall.ini


by the user when running the PostInstall/
PostInstall script

Folder containing the back-up copies of ${SAP_PA_HOME}/ N/A


the default configuration files PostInstall/Org/

Folder containing the back-up copies of ${SAP_PA_HOME}/ N/A


the modified configuration files PostInstall/Bkp/

Predictive Analytics Installation Guide


54 PUBLIC Installing SAP Predictive Analytics Server for Linux
6.1.4.1.1 Using the Configuration Script

1. Launch the script PostInstall.sh . The script asks you whether you want to reload an existing
configuration from a previous installation.
2. If you want to reload a configuration, enter Y , else enter N .
3. When prompted to select the type of component you want to configure, enter S to select the server.
4. When prompted, enter the values requested by the script. Once the script has been successfully
completed, the following line is displayed.

INFO: The installation has been checked, the following script can be executed:
INFO: - kxen.server

6.1.4.2 Basic Configuration

The basic server configuration option can be found in the file KxAuth.cfg located at the root of the installation
folder. This file contains keys and values separated by an equal sign, as shown in the following example.

Example

############## AuthServer configuration


AuthServer.TimeOut=4000
AuthServer.MaxConnections=50
AuthServer.ExeName=KxCORBA3
AuthServer.ExeStartDirectory=../CORBA
AuthServer.DomainName=KXEN01
AuthServer.Admin=rootAuthServer.Admin=admin

Available options for basic configuration


Option Default Value Description

AuthServer.TimeOut 4000 Internal time-out parameter used when starting internal instance.
If the instance started after user authentication cannot be reached
after this specified number of milliseconds, then a time-out occurs
and the connection fails. In most cases it is due to a bad configura­
tion or installation of the instance.

AuthServer.MaxConnections 50 Maximum number of concurrent connections on the server.

This can be used to hard-limit the number of Instance processes


launched at any time.

AuthServer.ExeName KxCORBA3 Name of the Instance executable file. It should not be changed.

AuthServer.ExeStartDirectory ../CORBA Name and path of the Instance execution directory. It should not
be changed.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 55
Option Default Value Description

AuthServer.InstanceConfigFile KxCORBA.cfg Can be used to specify a different configuration file for instances
started by Authenticated Server. It can be used for example in a
context where the server is used both as Authenticated and as a
regular server. More over if the name contains the string '%u' it will
be replaced by the authenticated user name starting the instance,
so that each user can have a separate configuration. It can be used
to specify different Data Sources (FileStoreRoot) or different Li­
cense for each user.

AuthServer.InstanceConfigPath . Specify the path to be searched for configuration file. By default


the configuration file are located in the same directory that the
KxCORBA3 executable, but this allow to give a specific directory
for all configuration files.

AuthServer.UserMappingFile­ (none) Name of a text file which describe the user mapping. See User
Name Mapping in the related information below.

AuthServer.DomainName KXEN01 Name of the Windows domain to be used for Windows authentica­
tion. See Authenticated Server based on the System in the related
information below.

AuthServer.Admin root Logging name(s) of the server administrators. Server administra­


tors can see and disconnect all the user connections in the Admin­
AuthServer.Admin admin
istration Console.

6.1.4.3 Configuration of SAP Predictive Analytics Server


Instances

The configuration of a server instance, which consists of a CORBA process, requires to set:

● the proper TCP port,


● and the proper configuration for the server itself.

The server instance can be configured using the configuration file called KxCORBA.cfg, located in the folder
<installation_folder>/KxenServer/. This file is loaded at start-up, and includes several Key/Values
properties. The first line should not be edited, blank lines are accepted, and each valid line should include: one
of the accepted keyword, the equal sign and the associated value.

The main usual configuration step is the definition of the FileStoreRoot to be used on the local system. It is
generally a good idea to set up entries for common data storage such as /tmp/data/ and also a specific entry
to store built models for example, /tmp/models/).

Available properties for server instances configuration


Option Description

KxAdmin The root name of the file or table used to store the models in a directory or ODBC store. This file is
automatically generated when "commit" or "save" operations are performed on Automated
Analytics models. It is used to quickly retrieve the list of available models in a store. By default it is
KxAdmin so that it will creates KxAdmin.txt file or KxAdmin tables.

Predictive Analytics Installation Guide


56 PUBLIC Installing SAP Predictive Analytics Server for Linux
Option Description

MessageFile Such a message file allows for a quick and easy localization of applications. New message files can
be added for new language. The convention used for such message file name is:
<Name>_<language>.umsg, where language is an ISO code of a language. For example, fr for
French, en for English, de for German.

MessageDirectory This option allows you to load all message files from a single directory. All .umsg file found in that
directory will be loaded as message.

FileStoreRoot Such entries are used by the server to retrieve the main file store to be displayed when client appli­
cations ask for top-level file directory.

Config Load another configuration file. This is used to load the license file, or to load some data access
library, but could be used also to base different configuration files on some common settings, for
example one for each user.

LogConf This option gives the name of the log system configuration file to be used. See the related informa­
tion section below for a description of the logging configuration file itself.

Example
To display two directories in the file selection drop down, add the following lines:

FileStoreRoot=/tmp/data/
FileStoreRoot=/tmp/models/

6.1.4.4 TCP Port and Address Configuration

1. Edit the file KxenCorbaConfig.sh located in the folder /softs/AutomatedAnalytics_<OsName>/.


2. Modify the value of the option NS_PORT. The default value is 12345.

8 # Port used by the CORBA Name Server


9 # This one must also be set on the client's GUI startup scripts.
10 #
11 NS_PORT=12345
12 export NS_PORT

3. Modify the value of the option KXEN_PORT. The default value is 12346.

16 # Port used by the server itself


17 #
18 KXEN_PORT=12346
19 export KXEN_PORT

4. Modify the value of the option KXEN_AUTH_PORT. The default value is 12347.

23 # Port used by the authenticated server


24 #
25 KXEN_AUTH_PORT=12347
26 export KXEN_AUTH_PORT

5. Optionally you can set the public address. This is especially useful when the network configuration includes
NAT (Network Address Translation), for example when the server is on a private network exposed on

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 57
another public address. Modify and uncomment the variables <OMNIORB_USEHOSTNAME> and
<KX_PUBLIC_ADDRESS> as shown below.

Sample Code
In this example, the public address of the server is 192.234.18.9.

OMNIORB_USEHOSTNAME=192.234.18.9
export OMNIORB_USEHOSTNAME
KX_PUBLIC_ADDRESS=192.234.18.9
export KX_PUBLIC_ADDRESS

6.1.4.5 Communication Channel Encryption

Caution
Connection through Transport Layer Security (TLS) or Secure Sockets Layer (SSL) is not activated by
default when you install SAP Predictive Analytics Server. You have to enable it explicitly.

To enable the encryption through TLS or SSL, you’ll need the following elements:

● A Certification Authority certificate (CA certificate)


● The server private key
● The server certificate
● The server pass phrase (optional)

Note
If your company can’t provide you with these elements, you can generate self-signed certificates with
OpenSSL . Note that for security reasons, it is recommended to use certificates issued by a certificate
authority.

6.1.4.5.1 Enabling the Encryption for Automated Analytics


Server

To deploy the certificates on the Automated Analytics server, you’ll need to combine the private key and the
server certificate into a single file with the .cer extension.

1. Edit the KxAuthServer.sh file located in <installation folder>/KxAuthServer.


2. Locate the line: exec KxAuthServer [...].
3. Add the following options at the end of the command line: -cacert %CACertificate% -servercert
%ServerCertificate% -password %PassPhrase%.
4. Replace the placeholders by the following values:

Predictive Analytics Installation Guide


58 PUBLIC Installing SAP Predictive Analytics Server for Linux
Placeholder Name Value

%CACertificate% Name of the file containing the CA certificate

%ServerCertificate% Name of the file combining the private key and the server certificate
%PassPhrase% The server private key pass phrase. Only use this option if you have set up a passphrase
for the server.

5. Replace tcp by ssl in the following options:


○ -ORBendPointNoPublish giop:tcp:<ADDRESS>:<PORT>
○ -ORBendPointPublish giop:tcp:<ADDRESS>:<PORT>

You should then have:


○ -ORBendPointNoPublish giop:ssl:<ADDRESS>:<PORT>
○ -ORBendPointPublish giop:ssl:<ADDRESS>:<PORT>
6. Save the file.

6.1.4.6 Authentication Types

6.1.4.6.1 Authenticated Server Based on the System

If you choose to use the PAM authentication then you should update /softs/
AutomatedAnalytics_<OsName>/KxAuthServer/KxAuth.cfg as follow and start the Automated
Analytics server under the root user account so that you can use the system password.

1. Comment the lines starting with Authenticator.File by adding # at the beginning of each line.
2. Uncomment the lines starting with Authenticator.Name by removing the # at the beginning of each line.

############## Authenticator configuration


#Authenticator.Name=Void
#Authenticator.Name=File
#Authenticator.File.FileName=users
#Authenticator.File.StoreType=KXEN.FileStore
#Authenticator.File.StoreName=.
#Authenticator.File.User=
#Authenticator.File.Password=
Authenticator.Name=System
#Authenticator.System.PamServiceName=login

6.1.4.6.1.1 About PamServiceName

The PamServiceName option refers to the name of the PAM Service used for authentication. This option, for
which the value is set to login is commented by default in the configuration file.

The PAM Service name has to be a valid entry defined in the system's PAM configuration file pam.conf located
in /etc/. This configuration file defines services by listing the modules to be applied for each service. Refer to
the PAM documentation for more details.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 59
It is possible to define a new PAM service for SAP Predictive Analytics. For example, if you want to implement a
specific authentication strategy to be able to add or remove an authentication module.

6.1.4.6.1.1.1 Defining a New PAM Service

1. Define the proper lines in the /etc/pam.conf file, starting by kxenlogin.


2. Edit the KxAuthServer/KxAuth.cfg file.
3. Uncomment the line starting with Authenticator.System.PamServiceName.
4. Set the value to kxenlogin.

6.1.4.6.2 Authenticated Server Based on a Fixed Password


List

By default, the authentication is based on a fixed file named users, which contains the login and password of
each user and is located in the folder /softs/AutomatedAnalytics_<OsName>/KxAuthServer/users.

6.1.4.6.2.1 Activating and/or Customizing this Option

1. Open the file KxAuth.cfg located in the same folder as the users file.
2. In the section Authenticator configuration , check that the line Authenticator.Name=File is not
commented, that is, does not begin with a hash sign (# ).
3. You can customize the options listed in the table below.

Options for the Authentication Based on a Fixed Password List


Option Description

Authenticator.File.FileName Sets the name of the store containing the users' logins and passwords. By
default, the store name is users .

Authenticator.File.StoreType Indicates how the users' logins and passwords are stored, for example in a
flat file, in a database, in a SAS file. The default value is
KXEN.FileStore, which means that the data are stored in a flat file.

Authenticator.File.StoreName Sets the path where the users' data are stored. The default value is a dot
(. ), which means that the data are stored in the same folder as the
KxAuth.cfg file.

Sample Code

############## Authenticator configuration


#Authenticator.Name=Void

Predictive Analytics Installation Guide


60 PUBLIC Installing SAP Predictive Analytics Server for Linux
Authenticator.Name=File
Authenticator.File.FileName=users
Authenticator.File.StoreType=KXEN.FileStore
Authenticator.File.StoreName=.
Authenticator.File.User=
Authenticator.File.Password=
#Authenticator.Name=System
#Authenticator.System.PamServiceName=login

6.1.4.6.3 Authenticated Server Based on a Fixed Encrypted


Password List

By default, the authentication is based on a fixed file named users, which contains the login and password of
each user and is located in the folder /softs/AutomatedAnalytics_<OsName>/KxAuthServer/users.

An authenticated server based on a fixed encrypted password list works on same principle than a standard file­
based one except that it supports two types of passwords, which are encrypted and non-encrypted. In the
password store, the encrypted passwords are preceded by a star (*).

Sample Code
The following example shows a fixed encrypted password list containing both encrypted and non-encrypted
passwords:

Key=Value
admin=admin
user1=user1
user2=user2
port=*901555fb06e346cb065ceb9808dcfc25

To implement this type of authentication, a password store such as a 'users' password file with passwords
stored in clear must be used exactly as in the standard file­based authentication. At this point, it works exactly
as an authenticated server based on a fixed file password list. Differences come when a user changes their
password, which is then automatically stored as an encrypted one.

It can be used with an existing 'users' password file used on a server based on the 'File' authentication.
Passwords are progressively encrypted when users change them using the Remote Control Panel.

1. Open the file KxAuth.cfg located in the folder <INSTALLATION DIR>/KxAuthServer.


2. In the section Authenticator configuration, check that the line Authenticator.Name=Encrypted
is not commented , that is, does not begin with a hash sign (# ):
3. You can customize the options listed in the following table.

Options for the Authentication Based on an Encrypted Password List


Option Description Default Value

Authenticator.Encrypted.F Sets the name of the store containing users


ileName the user logins and passwords

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 61
Option Description Default Value

Authenticator.Encrypted.S Indicates how are stored the user log­ KXEN.FileStore


toreType ins and passwords, for example in a
flat file, in a database, in a SAS file This value means that the data is
stored in a flat file.

Authenticator.Encrypted.S Sets the path where the user data is . (dot)


toreName stored
This value means that the data is
stored in the same folder as the
KxAuth.cfg file.

The following lines show the settings to use in the KxAuth.cfg file to set up an encrypted authentication.

#Authenticator.Name=File
#Authenticator.File.FileName=users
#Authenticator.File.StoreType=KXEN.FileStore
#Authenticator.File.StoreName=.
#Authenticator.File.User=
#Authenticator.File.Password=
Authenticator.Name=Encrypted
Authenticator.Encrypted.FileName=users
Authenticator.Encrypted.StoreType=KXEN.FileStore
Authenticator.Encrypted.StoreName=.
#Authenticator.Encrypted.User=
#Authenticator.Encrypted.Password=
#Authenticator.Name=System
#Authenticator.System.PamServiceName=login

6.1.4.6.4 No Authentication

Another authentication scheme called "Void" is available, but it should only be used with care, generally for
technical integration with other system. This "Void" authenticator accept any combination of user and
password as a valid one, it does not perform any actual identity check.

This authentication scheme can be used when the connections of the authenticated server are only created by
another software components, such as web server, or another application where identity is already managed.
The other software components might want to create connection as some user, but do not want or need to do
an actual authentication.

The goal of specifying a user in a void authentication connection is to be able to trace the work and process on
a per-user basis since most logs and trace information reference the connected user.

Note
Although it is possible to enter an empty user/password in such a case, it is not a recommended practice for
traceability and efficiency purpose: all logs will be mixed in the same file, without any distinction.

If this kind of authentication is used, the software integrator or application developer should ensure that the
server cannot be physically accessed another way by using proper firewalls, network configuration, …, as
allowing a connection from another route will prevent data security.

Predictive Analytics Installation Guide


62 PUBLIC Installing SAP Predictive Analytics Server for Linux
Also in such a configuration, it is strongly recommended to not start the server as administrator or super user,
as no impersonification will be done after authentication. The identity of the instance started will be the one
used to start the server process itself.

6.1.4.7 User Mapping

Once a user is authenticated through the authentication module, the server will start a new instance for that
user connection. The system account used for this new instance can be:

● The same system account as the one used to start the server. This occurs with “void” authenticator for
example.
● The system account of the authenticated user, this is the default with the system authentication.
● A different system account, as specified in a “User-Mapping” file. This file will specify, for each
authenticated user name the name of a system account to be used. This feature is available only on Linux.

The mapping file contains the name of the logical authenticated user and the name of the actual user account
to be used separated by an equal sign.

Example
For example, here is a sample file which maps 5 users on 2 different accounts:

Key=Value
user1=kxuser1
user2=kxuser1
user3=kxuser1
user4=kxuser2
user5=kxuser2

If a user is not present in this mapping file, then by default it will be denied, unless a specific entry is present
in this mapping file:

*=kxusern

This indicates that “all other users” will be mapped to physical user “kxusern ”.

To be used by authenticated server, the mapping file need to be declared in the configuration file
(KxAuth.cfg). Uncomment the UserMappingFile entry, and specify the correct filename.

Example


AuthServer.UserMappingFileName=mymapping.txt

User mapping can be used to simplify user account management, or to implement specific authorization
policies.

Other special rules for user mapping allow creating specific mapping policies.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 63
To map specific users as themselves:

Key=Value
user1=*

The authenticated server will start a CORBA process as user1 when user1 authenticates himself. To map all the
other users as themselves, use the following syntax.

Key=Value
user1=kxuser1
user5=kxuser2
*=*

Any other users than user1 and user5 will be mapped as themselves.

6.1.4.8 User Environment Setup

The SAP Predictive Analytics server allows multiple users to run server processes via impersonation.

These users can have different needs in the global configuration of their work environment under Linux
Operating Systems.

One example is the need to access different databases such as Teradata and Oracle. The access to specific
databases is set up in the user environment in variables like ODBCINI and the shared library path.

The idea is to allow the set-up of the execution environment of SAP Predictive Analytics.

Note
This setup is not automatically done by the assisted configuration script.

6.1.4.8.1 Options Allowing Set-up Configuration

The following options can be set up in the configuration file KxAuth.cfg

Option Description Default Value

AuthServer.ShellStarter the full path to the shell to use to launch the /bin/sh
KxLauncher_start.sh for all the users.

AuthServer.StartAsLoginShel starts the shell as a login shell. A login shell will load the false
l user's environment as if the user logs in with rlogin.

AuthServer.ShellOption the option to pass to the shell if needed. For example, on empty
some operating systems bash needs to be launched as an
interactive shell with the option -i in order to source its
configuration scripts.

Predictive Analytics Installation Guide


64 PUBLIC Installing SAP Predictive Analytics Server for Linux
So for example if the AuthServer.ShellStarter option above is /bin/bash and the
AuthServer.ShellOption is -i, the process run will be equivalent to launch, for each user:

/bin/bash -i KxLauncher_start.sh

6.1.4.8.2 Guidelines

The guidelines listed below must be followed when defining variables:

● Any redefinition of the shared library path and the <$PATH> variable must take into account the previous
values of this variables:

LD_LIBRARY_PATH=/usr/local/unixODBC_v2.4/libs:$LD_LIBRARY_PATH
export LD_LIBRARY_PATH

● When the .kxen_profile is sourced, the current directory for the shell is the server directory.
● The environment variable definition is using the POSIX standard shell sh semantic, which first defines the
variable, then exports it. To guarantee a proper functioning, you should use the following syntax to define
your variables:

Code Syntax

MyVariable=ok
export MyVariable

● Use the <$HOME> variable rather than the tilde character ( ~ ) to refer to the Home directory:

ODBCINI=$HOME/MyOdbcini.ini

● Export ODBCINI

6.1.4.8.3 Troubleshooting

If the server does not start for a user, follow the procedure described below:

1. Log on the machine as the user.


2. Make sure the user is in the KXEN group by executing the command: groups
3. Go to $KXINSTDIR/KxenServer.
4. Execute the command: /bin/sh KxLauncher.sh KxCORBA3 -NoRegister
The error should be displayed in the standard output.

6.1.5 Reloading the Configuration

You can reload part of the configuration of the server without restarting it.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 65
Execute the following command: /softs/AutomatedAnalytics_<OsName>/kxen.server
reload_configuration
This will reload only the options of the server such as identification, maximum connection, timeout, but will not
reload the other options like the log files.

Note
Sending the signal SIGUSR1 to the KxAuthServer process will have the same result.

6.1.6 Managing the Server on Linux

The system account used to start the autenticated server should have sufficient rights to perform
authentication and impersonation, that is, start the server instance as another user. In particular, it means that
most of the time it should be started with the root user account if:

● System authentication is to be used through Pluggable Authentication Module (PAM). Access to Linux
System account password required root privileges.
● Server instances are going to be started with a different system account than the authenticated server.

So the only case where a different user (who is not identified as root) can be used is when void or File/
Encrypted authentications are used, and all users are mapped to the same used as the one used to start the
processes.

Before starting Automated Analytics server, make sure that:

● The user account used to run the server can write in the tmp directory of the application installation. It is
the case if
○ the user who runs SAP Predictive Analytics is also the one used to install the software
○ if the user running SAP Predictive Analytics is root (which is ensure by the post installation
configuration script).
● Users that will be connecting to the server should have rights also to write their logs on the file system. By
default it is the tmp directory of the installation. Typically this could be done by following the steps listed
below:
1. Create a Linux group dedicated to Automated Analyticss users (for example kxenusrs).
2. Add the concerned Linux accounts to this group.
3. Set this group as the group owner of the tmp directory. This should be done if the application has been
set up with a user account that has this group set as its primary group.
4. Allow the group to write there. This is automatically done when using the post installation
configuration script.

chgrp kxenusrs tmp


chmod g+w tmp

Caution
The three procedures described below should be executed as root.

Predictive Analytics Installation Guide


66 PUBLIC Installing SAP Predictive Analytics Server for Linux
6.1.6.1 Starting the Server and its Name Server

Enter the command: /softs/AutomatedAnalytics_<OsName>/kxen.server start

6.1.6.2 Checking Whether the Server is Running

Enter the command: /softs/AutomatedAnalytics_<OsName>/kxen.server check

6.1.6.3 Stopping the Server and its Name Server

Enter the command: /softs/AutomatedAnalytics_<OsName>/kxen.server stop

6.1.7 Deploying the Client

SAP Predictive Analytics client applications are Java graphical user interface that connect to SAP Predictive
Analytics servers. The client application can be installed:

● Through Java Web Start: in this mode, a Web server needs to be installed and configured on the machine
hosting the SAP Predictive Analytics server. It will deliver, through web pages, the necessary content in
order for the client computer to run the application.
● With standalone installers on each targeted client system.

We recommend using the Web Start facility, as it will allow automatic client version migration in case of a server
upgrade.

Using Java Web Start

Java Web Start is an application launcher for the client application. It allows you to start the client from a web
site.

The updates are automatically managed through the Java Web Start system. This system avoids having to
manually install the client on every computer of the local network. As an extra, if the client is updated on the
server, all the distributed applications will automatically be updated too.

The installation and configuration of Java Web Start is described in the section Installing Java Web Start .

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 67
Manually Install Clients for SAP Predictive Analytics Server

For information on the installation and configuration of an SAP Predictive Analytics client, read the section
Installing SAP Predictive Analytics Client.

6.1.7.1 Using Java Web Start

Java Web Start is an application launcher for the client application. It allows you to start the client from a web
site.

The updates are automatically managed through the Java Web Start system. This system avoids having to
manually install the client on every computer of the local network. As an extra, if the client is updated on the
server, all the distributed applications will automatically be updated too.

The installation and configuration of Java Web Start is described in the section Installing Java Web Start.

6.1.7.2 Manually Install Clients for SAP Predictive


Analytics Server

SAP Predictive Analytics clients can also be manually installed.

1. Use the SAP Predictive Analytics installer to install a client.

Note
For more detailed information on the installation and configuration of a SAP Predictive Analytics client,
refer to the section Installing SAP Predictive Analytics Client.

2. To connect the client to the server, add the argument -authenticated on the command line.

6.2 Monitoring

6.2.1 Logging

Automated Analytics offers a logging system which is activated by default and can be deactivated if needed.
The goal of this logging system is for an administrator or a support team to be able to trace and analyze the
application activity on within the kernel. Note that the activity of the client user interfaces is only seen through
events fired within the kernel.

It can be deactivated within a server, or in the Shell utility.

Predictive Analytics Installation Guide


68 PUBLIC Installing SAP Predictive Analytics Server for Linux
The default log files for the server are:

● kxen_auth_log.txt: the default activity trace for the server, which can be used mainly for debugging
purpose
● kxenusagelog.txt: the goal of this log file is to have a simpler monitoring of the overall Automated
Analytics activity. This log will include one line for each of the following events: user connection, user
disconnection, training of a model, application of a model, saving of a model.

These files are located in the tmp directory of the SAP Predictive Analytics distribution.

For a detailed description of the logging system, see Operations Guide.

6.2.2 Starting the Administration Console

The administration console is a tool that allows the server administrator to follow the users' activity on the
server by:

● listing active connections, with information such as user, last activity time stamp, number of current
models,
● listing the state of models being within each connection,
● disconnecting an existing connection. This should be used with care, in case some pending connection
remains active while a client crashed for example, in order to free the associated process.

/softs/AutomatedAnalytics_<OsName>/KJWizardCORBA/KxControlPanel.sh serverhost where


serverhost is the name of the host running the name server.

6.3 Troubleshooting

6.3.1 Displaying the Server Log File

Enter the command: cat nohup.out


The log displays the list of problems found when starting the servers.

6.3.2 Error: ./tmp/KxAuthServer_.pid is not running

When starting the server with the command ./kxen.server start, you may encounter this error. In that
case, follow the procedure described below to identify the problem.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Server for Linux PUBLIC 69
6.3.2.1 Missing openssl Library

If the log contains the following line:

KxAuthServer: error while loading shared libraries: libssl.so.0.9.8: cannot open


shared object file: …

This means that the required SSL library is not available on the system.

6.3.2.1.1 Correcting the Issue

1. Check that there is a SSL library installed on the system.


On a RedHat Enterprise 5, use the command:

rmp –qa | grep ssl

2. Display the list of the files installed by the openssl package and filter it to display only the required libraries.
On a RedHat Enterprise 5, enter the following commands:

rpm –ql openssl<installed_version> | grep libssl


rpm –ql openssl<installed_version> | grep libcryp

where <installed_version> is the version of the openssl package listed in the previous step.
3. Go to the installation directory.
4. Go to the libs folder.
5. Create the symbolic links using the name of the library indicated as missing.
If the library lib_ssl.so.1.0.0 is missing, enter the following commands:

ln –s <lib64 path>/libssl.so.<version number> libssl.so.1.0.0


ln –s <lib64 path>/libcrytpo.so.<version number> libcrypto.so.1.0.0

where <lib64 path> is the location of the lib64 folder provided by the commands at step 2.
6. Restart the servers using the commands:

./kxen.server stop
rm nohup.out
./kxen.server start

The servers should start. If they do not, check the log file again and contact your administrator or the
support.

Predictive Analytics Installation Guide


70 PUBLIC Installing SAP Predictive Analytics Server for Linux
7 Installing SAP Predictive Analytics
Desktop

7.1 SAP Predictive Analytics Desktop Edition

SAP Predictive Analytics desktop edition is a stand-alone process with a two-tier architecture.

Automated Analytics can access data in flat files on the native file system, SAS, and SPSS files, or be
configured to access Database Management Systems using ODBC.

Expert Analytics enables you to connect to various data sources such as flat files, relational databases, in-
memory databases, and SAP BusinessObjects universes. In addition, you can operate on different volumes of
data from a small matrix of data in a CSV file to a very large dataset in SAP HANA.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Desktop PUBLIC 71
7.2 Installing Predictive Analytics

7.2.1 Installation Prerequisites

Before installing SAP Predictive Analytics Desktop Edition, make sure the following requirements are met:

● Check with the SAP Product Availability Matrix http://service.sap.com/sap/support/pam that your
operating system installed on your machine is supported by SAP Predictive Analytics.
● If you have already installed SAP Predictive Analytics on your machine, uninstall it before installing SAP
Predictive Analytics.
● If you want to simply perform an upgrade of your SAP Predictive Analytics Desktop Edition, you don’t need
to uninstall the previous version. The new version will erase the old one during the installation process.

Note
If you have already installed SAP Lumira on your machine, you do not need to uninstall it. SAP Predictive
Analytics is installed in addition in a separate folder.

● You must have Administrator rights to install SAP Predictive Analytics on the computer.
● Sufficient disk space must be available on various resources. For a detailed list of supported environments
and hardware requirements, see the Product Availability Matrix at: http://service.sap.com/pam .

Note
It's possible to install on your computer both an SAP Predictive Analytics Server version and a Desktop
version. Both versions can be installed on separate directories.

7.2.2 Installing SAP Predictive Analytics Using the Setup


Program

The SAP Predictive Analytics Setup program is an installation wizard that guides you through the installation of
the required SAP Predictive Analytics resources on your computer.

The SAP Predictive Analytics Setup program is contained within the self-extracting archive -
SAPPredictiveAnalysisSetup.exe. The program automatically recognizes your computer's operating
system and checks for platform requirements. It updates files as required.

1. Navigate to the SAP Predictive Analytics self-extracting archive - PADESKTOP<version


number>_<patch level>.EXE, for example PADESKTOP2002_0.EXE for version 2.2 patch level 0 - and
double-click it.
The User Account Control dialog box appears with a warning message.
2. Choose Yes in the confirmation prompt.

The SAP Predictive Analytics Setup program is extracted from the archive. The Installation Manager
performs a verification check for all of the installation prerequisites. A Prerequisites page opens only if the
verification fails for any requirement. Close the wizard and correct any missing prerequisite before
relaunching PADESKTOP<version number>_<patch level>.EXE.

Predictive Analytics Installation Guide


72 PUBLIC Installing SAP Predictive Analytics Desktop
If all of the installation prerequisites are confirmed, the Define Properties page opens.
3. Select the setup language from the drop-down list.
4. Specify the destination folder for installing SAP Predictive Analytics.

○ To accept the default installation directory, choose Next .


○ To install SAP Predictive Analytics in a different location, choose Browse. Select the required folder and
choose Next.

The License Agreement page appears.


5. Review the license agreement and select I accept the License Agreement.
6. Choose Next.
The Ready to Install page appears. You can go back to modify your installation information if required.
7. To begin the installation, choose Next.
The installation is complete when the Finish Installation page opens.
8. To automatically launch the program, select Launch SAP Predictive Analytics after installation completes.
9. To exit this installation, choose Finish.

After completing the installation steps, you can register your installation.

Note
A temporary license key is included in the software. Apply for a permanent license key immediately after
installing your SAP software on the SAP Support Portal at http://support.sap.com/keys-systems-
installations/keys.html .

The first time you start SAP Predictive Analytics, a keycode dialog appears where you can enter your keycode.
You can enter a new keycode at any time. On the SAP Predictive Analytics Welcome screen, select Help
Enter keycode and choose Save.

7.2.3 Performing a Silent Installation

You can use the SAP Predictive Analytics self-extractor to create a response file required for a silent
installation.

Using a silent installation, system administrators can run a script from the command line to automatically
install SAP Predictive Analytics on any machine in their system without the setup program prompting them for
information or displaying the progress bar. The silent installation is primarily geared towards users with
network administration roles. A silent installation is particularly useful when you need to push multiple
installations in your corporate network. Once you have created a silent installation response file, you can add
the silent installation command to your installation scripts.

1. Choose Start Run and type cmd to open a Command Prompt window.
2. Retrieve the SAP Predictive Analytics self-extracting archive.
3. Run the following command:

<filename.exe> -w <response_filepath>\response.ini

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Desktop PUBLIC 73
Note
<filename.exe> represents the name of the .exe file. <response_filepath> represents the file path
where you want to save the response file.

The SAP Predictive Analytics Setup program opens.


4. Follow the installation wizard to select your SAP Predictive Analytics setup options.
5. On the Start installation page, choose Next.
The setup program writes your installation options to the response.ini file, and closes.

Tip
You can now open response.ini in a text editor to review your setup selections.

6. To run the silent installation, open a Command Prompt window and enter the following command:

<filename.exe> -r <response_filepath>\response.ini

The parameter -r requires the name and location of the response file as specified in Step 3.

You can also apply other optional parameters listed in the table below.

Parameter Name Description Example

-s hides the self-extraction progress bar -s


during the silent installation

userkeycode adds the “User key code” to the in­ userkeycode=XXXXXYYYYYYY-


stalled License.cfg file ZZZZZZZ-….

licensekeyfile adds the “License Key File” to the in­ licensekeyfile=c:\temp


stalled License.cfg file \License.cfg

Note
Adding the user information and/or the license Key automatically thanks to the parameters
Userkeycode and licensekeyfile speeds up the installation process. You can use one or both
parameters in the same command.

Example

<filename.exe> -s –r <response_filepath> \response.ini licensekeyfile=c:\temp


\License.cfg userkeycode=XXXXXYYYYYYY-ZZZZZZZ-…

Predictive Analytics Installation Guide


74 PUBLIC Installing SAP Predictive Analytics Desktop
7.3 Configuring Trace Logs

7.3.1 Configuring Trace Logs for Expert Analytics

You use this procedure to enable the SAP Predictive Analytics application to record information about the
execution of the application. This log information helps you identify issues when the application fails or
encounters a problem.

By default the error messages and trace messages are written to the folder %TEMP%\sappa\logs in your
machine. However, you can change the default location of the folder, where the installation information is
written by performing the following steps:

1. Create a folder in any location for generating logs.

Note
Ensure that you have "write" permission to the folder.

For example, C:\logs.


2. Create the BO_Trace.ini file and add the following trace details to it.

active=false;
severity='E';
importance=xs;
size=1000000;
keep_num=437;
alert=true;

The table below lists the general parameters used for configuring server tracing.

Parameter Possible Values Description

active false, true If set to true, trace messages that


meet the threshold set in the
importance parameter will be traced.
If set to false, trace messages will not
be traced based on their
"importance" level. Default value is
false.

importance '<<', '<=', '==', '>=', '>>', xs, s, m, l, xl Specifies the threshold for tracing
messages. All messages beyond the
Note threshold will be traced. Default value
is m (medium).
importance = xs or importance =
<< are the most verbose options
available while importance = xl or
importance = >> are the least.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Desktop PUBLIC 75
Parameter Possible Values Description

alert false, true If set to true, trace messages that


meet the threshold set in the severity
parameter will be traced. If set to
false, the trace messages will not be
traced based on their "severity" level.
Default value is true.

severity ' ', 'W', 'E', 'A', success, warning, error, Specifies the threshold severity over
assert which massages can be traced.
Default value is 'E'.

size Possible values are integers >=1000 Specifies the number of messages in
a trace log file before a new one is
created. Default value is 100000.

keep_num Possible values are integers >=1000 Specifies the number of logs to keep.

administrator Strings or integers Specifies an annotation to use in the


output log file. For example, if

administrator = "hello"

this string is inserted into the log file.

log_dir For example, C:\logs. Specifies the output log file directory.
By default log files are stored in the
Logging folder.

always_close on, off Specifies if the log file should be


closed after a trace is written to the
log file. Default value is off.

3. Save and close the BO_trace.ini file.


4. Place the BO_Trace.ini file under C:\logs.
5. Set up the following environment variables:
○ BO_TRACE_LOGDIR = C:/logs
○ BO_TRACE_CONFIGDIR = C:/logs
○ BO_TRACE_CONFIGFILE = C:/logs/BO_Trace.ini
6. Restart the application.

The application logs are generated in the specified location. For example, C:\logs.

7.3.2 Configuring Trace Logs for Automated Analytics

Automated Analytics writes error and trace messages to the file kxenwizard_log_<user name>.txt in the
user temporary directory (for example, C:\Users\<user name>\AppData\Local\Temp).

Predictive Analytics Installation Guide


76 PUBLIC Installing SAP Predictive Analytics Desktop
By default, the log format corresponds to the standard used by SAP systems (ITSAM-17). You can revert to the
log format used by previous versions of SAP InfiniteInsight using the following procedure.

1. Exit the application.


2. Edit the log configuration file logconf.txt found in the folder <installation folder>\Automated
\EXE\Clients\KJWizardJNI\ (the default installation folder is C:\Program Files\SAP
BusinesObjects Predictive Analytics\Desktop <version>\).
3. Comment the first Root.format and Root.DateFormat lines, and uncomment the second occurance of
these lines by removing the hash sign (#).

Root.format="#2.0#%D#%t#%l#%C#%K##%I#%T#%P##KxShell#%U#######Main#PLAIN##%M#"
Root.DateFormat="%Y %m %d %H:%M:%S:%V"
#Root.format="[%D] %C [ID %I] [Lvl %L] [%E] '%M'"
#Root.DateFormat="%b-%d %H:%M:%S"

4. Save the logconf.txt file and restart the application.

For more information about log file configuration, see the Predictive Analytics Administrator Guide,
sectionAutomated Analytics Operations Guide on the SAP Help Portal at http://help.sap.com/pa.

7.4 Important Considerations for Using SAP HANA with


Expert Analytics

Important considerations and requirements for using Expert Analytics with the SAP HANA database.

Security requirements for publishing to SAP HANA

Before users can publish content to SAP HANA, they must be assigned specific privileges and roles. These
roles and privileges are also required for retrieving data from SAP HANA. Use the SAP HANA Studio application
to assign user roles and privileges. For information on administrating the SAP HANA database and using SAP
HANA Studio see SAP HANA Database – Administration Guide. For information on user security see the SAP
HANA Security Guide (Including SAP HANA Database Security).

The user account used to log into the SAP HANA system from SAP Predictive Analytics must be assigned the
MODELING role (in SAP HANA).

Note
This action can only be performed by a user with ROLE_ADMIN privileges on the SAP HANA database.

When an SAP Predictive Analytics user logs into the SAP HANA system, the internal _SYS_REPO account must:

● Be granted the SELECT SQL Privileges.


● Have the Grantable to others option selected in the (SAP Predictive Analytics) user's schema.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Desktop PUBLIC 77
Related Information

Configuring _SYS_REPO for the SAP Predictive Analytics User [page 78]
Granting Object Privileges to Access HANA Online Data Source [page 78]
Granting Privileges to Run Demand Forecasting Functionality with AFL [page 79]
Granting Privileges to Run PAL Algorithms with AFL [page 79]
Granting Privileges to Run OFL Functionality with AFL [page 80]
Supported OLAP Measures [page 80]
Using SAP APL Functions with Expert Analytics [page 80]

7.4.1 Configuring _SYS_REPO for the SAP Predictive


Analytics User

How to configure _SYS_REPO in SAP HANA Studio.

An account for the SAP Predictive Analytics user is already defined in the SAP HANA system.

1. From the system connection in the SAP HANA Studio Systems view, choose Security Users .
2. Double-click the _SYS_REPO account.
3. On the Object Privileges tab, click the + icon, and enter the name of the user's schema, choose OK.
4. Choose the SELECT privilege.
5. If enabled, you have the option to select the Grantable to Others radio button.
6. Choose Deploy.

Note
Users can also open an SQL editor in SAP HANA Studio and run the following SQL statement:

GRANT SELECT ON SCHEMA <user_account_name> TO _SYS_REPO WITH GRANT OPTION

7.4.2 Granting Object Privileges to Access HANA Online Data


Source

Object (_SYS_REPO , _SYS_BI , _SYS_BIC ) privileges are granted by the SAP HANA administrator for the SAP
Predictive Analytics user.

An account for the SAP Predictive Analytics user is already defined in the SAP HANA system.

1. From the system connection in the SAP HANA Studio Systems view, choose Security Users .
2. Double-click the <HANA Online user account>.
3. On the Object Privileges tab, click the + icon, select _SYS_REPO, and choose OK.
4. Under Privileges for '_SYS_REPO', choose the SELECT privilege.

Predictive Analytics Installation Guide


78 PUBLIC Installing SAP Predictive Analytics Desktop
Perform the same steps for the _SYS_BI and _SYS_BIC objects.

7.4.3 Granting Privileges to Run Demand Forecasting


Functionality with AFL

The SAP HANA administrator must grant privileges for the SAP Predictive Analytics user to run UDF (Unified
Demand Forecast library) with AFL (Application Function Library) in Expert Analytics.

Ensure that the user has an account in the SAP HANA system and a UDF license.

Note
For further information on implementing UDF, see the SAP Help page, Unified Demand Forecast (UDF).

Take the following steps to grant the necessary privilege for Expert Analytics users to work with the Demand
Forecasting component:

1. Launch SAP HANA Studio.


2. Select the system that the Expert Analytics product will use.
3. Open SQL console in HANA Studio.
4. Execute the following SQL statement: GRANT AFL__SYS_AFL_UDFCORE_AREA_EXECUTE;

7.4.4 Granting Privileges to Run PAL Algorithms with AFL

The SAP HANA administrator must grant privileges for the SAP Predictive Analytics user to run PAL (Predictive
Analysis Library) Algorithms with AFL (Application Function Library).

An account is already defined in the SAP HANA system for the user. Take the following steps to grant the
necessary privileges to the user:

1. From the system connection in the SAP HANA Studio Navigator window, choose Security > Users.
2. Double-click the <HANA Online user account>.
3. On the Object Privileges tab, click the + icon, select AFL_WRAPPER_GENERATOR(SYSTEM), and choose
OK.
4. Under Privileges for 'AFL_WRAPPER_GENERATOR(SYSTEM)', select EXECUTE.
5. On the Granted Roles tab, click the + icon, select AFL__SYS_AFL_AFLPAL_EXECUTE, and choose OK.

For more information on how to install AFL and create the AFL_WRAPPER_GENERATOR(SYSTEM) procedure,
see the SAP HANA Predictive Analysis Library (PAL) Reference Guide.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Desktop PUBLIC 79
7.4.5 Granting Privileges to Run OFL Functionality with AFL

The SAP HANA administrator must grant privileges for the SAP Predictive Analytics user to run OFL
(Optimization Function Library) with AFL (Application Function Library) in Expert Analytics.

Ensure that an account is already defined for the user in the SAP HANA system and take the following steps to
grant the necessary privileges for Expert Analytics users to use the HANA Optimization Function:

1. Launch SAP HANA Studio.


2. Select the system that the Expert Analytics product will use.
3. Open SQL console in HANA Studio.
4. Execute the following SQL statement: GRANT AFL__SYS_AFL_OFL_AREA_EXECUTE TO <USER_NAME>

7.4.6 Supported OLAP Measures

OLAP measures that are supported by SAP HANA.

SAP HANA supports only the following measures of aggregation in OLAP data sources:

● SUM
● MIN
● MAX
● COUNT

If your dataset contains an aggregation on a measure that is not listed above, the aggregation will be ignored by
SAP HANA during publication and it will not be part of the final published artifact.

7.4.7 Using SAP APL Functions with Expert Analytics

You can use SAP APL (SAP Automated Predictive Library) functions in Expert Analytics when connected to
SAP HANA.

● When creating a new user schema, you need to make sure the required privileges are granted.

Recommendation
If you are creating or updating a new user schema, it is recommended to drop the existing table types
before creating new table types.

You can find a list of the available table types and their column details in the SAP Automated Predictive
Library Reference Guide on SAP Help Portal at http://help.sap.com/pa.

To understand more about APL privileges, go to the following section in the SAP APL user guide,.

Predictive Analytics Installation Guide


80 PUBLIC Installing SAP Predictive Analytics Desktop
7.5 Important Considerations for Using SAP
BusinessObjects Universes with Expert Analytics

To acquire data from universes that exist on BI platforms, ensure that the Web Intelligence Server is running.
For the complete list of supported BI platforms, see the SAP Product Availability Matrix

7.6 Connecting Automated Analytics to your Database


Management System

Automated Analytics can be configured to access data in a Database Management System using ODBC. To
configure your installation, refer to the guide Predictive Analytics Data Access guide, section Connecting to your
Database Management System on Windows available on the SAP Help Portal at https://help.sap.com/pa.

7.7 Installing and Configuring R

7.7.1 Installing R and the Required Packages

R is an open-source programming language and software environment for statistical computing.

Expert Analytics supports the following R algorithms and dependent packages for both SAP HANA and
agnostic platforms. For more information about supported platforms and technologies, click to access the
Product Availability Matrix page and search for “Predictive Analytics”, http://service.sap.com/sap/
support/pam .

SAP HANA (Online):

R Algorithm Dependent Packages

HANA R-Apriori arules

HANA R-CNR Tree rpart

HANA R-Multiple Linear Regression stats

HANA R-Triple Exponential Smoothing stats

HANA R-Bagging Classification adabag, rpart

HANA R-Boosting Classification adabag, rpart

HANA R-Random Forest Classification randomForest

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Desktop PUBLIC 81
R Algorithm Dependent Packages

HANA R-Random Forest Regression randomForest

Agnostic (Offline):

R Algorithm Dependent Packages

R-CNR Tree rpart

R-Apriori arules

R-K-Means stats

R-Linear Regression stats

R-Multiple Linear Regression stats

R-Exponential Regression stats

R-Geometric Regression stats

R-MONMLP Neural Network monmlp

R-NNet Neural Network nnet

R-Single Exponential Smoothing stats

R-Double Exponential Smoothing stats

R-Triple Exponential Smoothing stats

R-Bagging Classification adabag, rpart

R-Boosting Classification adabag, rpart

R-Random Forest Classification randomForest

R-Random Forest Regression randomForest

Prerequisites:

To use open-source R algorithms in your analysis, you need to install the R environment and configure it with
the application.

SAP Predictive Analytics provides an option to install and configure the current version of R and the required
packages from within the application. Ensure that you are connected to the internet while installing R.

Before installing R (and corresponding required packages), ensure that the following requirements are met:

● The existing R is uninstalled and the registry entries and the R installation folder are removed from the
machine.
● The R environment variables (R_LIBS, R_HOME) and R path variables are removed.

Predictive Analytics Installation Guide


82 PUBLIC Installing SAP Predictive Analytics Desktop
Installation:

To install the R environment and the required packages, perform the following steps:

1. Launch SAP Predictive Analytics.


2. Open Expert Analytics.
3. From the File menu, choose Install and Configure R.
4. Select Install R.
5. Read the open-source R license agreement, important instructions, and select I agree to install R using the
script.
6. Select Ok.

Note
If you have already installed R, you can use this procedure to install the required R packages, or manually
install the following required packages through your R console.

7.7.2 Configuring R

After you have installed R, configure the R environment to enable R algorithms in the application.

To configure R, perform the following steps:

1. Launch SAP Predictive Analytics.


2. Open Expert Analytics.
3. From the File menu, choose Install and Configure R.
4. On the Configuration tab, select Enable Open-Source R Algorithms.
5. Choose Browse to select the R installation folder.
For example, C:\Users\Public\R-3.1.2.
6. Choose Ok.
The "User Account Control" dialog box appears with a warning message.
7. Choose Yes in the confirmation prompt.

Recommendation
● When installing R packages, check that the folder containing the installed R packages exists under the
file path in File Install and Configure R Configuration .
You can specify the exact location where you want an R package installed using the following command
in R studio: install.packages("PackName",lib="PATH")
For example: install.packages("recomandable",lib="C:/Users/Public/R-3.1.2/
library"). After the installation, check the subfolder in the R file path. You should see a folder called
recomandable.
● Alternatively, if you have R packages installed in multiple locations, make sure that the Microsoft
Windows® environment variable, R_LIBS, is pointing to each location. Add the folder locations to the
R_LIBS variable, separating each one by a semicolon so they can be found by Expert Analytics.
For example: R_LIBS=%R_HOME%\library;%YOUR_HOME_DIRECTORY%\Documents\R\win-
library\3.1

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Desktop PUBLIC 83
If the R_LIBS variable is not already available, you need to create this.

Note
You can use the .libPaths() function to display all locations where R packages are installed.

7.7.3 Important Considerations for Using Expert Analytics


with R Algorithms in SAP HANA Online Mode

SAP HANA supports in-database data mining through R integration and the Predictive Analysis Library (PAL),
as well as the Automated Predictive Library (APL) and Unified Demand Forecast (UDF). When using Expert
Analytics with R algorithms in the SAP HANA online mode, the following considerations are important:

● To use R algorithms in the SAP HANA database, you must install and configure R on SAP HANA. For
information on how to install and configure R on SAP HANA, see the SAP HANA R integration guide
available at http://help.sap.com/hana_platform.
● Ensure that the user privilege Create R script is granted.

7.8 Uninstalling SAP Predictive Analytics

To remove an installation of SAP Predictive Analytics from your machine, you can use the standard procedure
or the silent one.

Note
On Windows 10 only, once the uninstallation is completed, the path C:\Program Files\SAP Predictive
Analytics\ still exists, and contains:

● The Desktop folder


● The file Setup.exe, which is stored under the Desktop folder.

You can remove these elements manually or reinstall SAP Predictive Analytics on the same location.

7.8.1 Performing a Standard Uninstallation

1. Choose Start Control Panel Programs .


2. Choose Uninstall a program.
3. Right-click SAP Predictive Analytics and choose Uninstall.
The SAP Predictive Analytics Setup wizard appears.

Predictive Analytics Installation Guide


84 PUBLIC Installing SAP Predictive Analytics Desktop
4. On the Confirm Uninstall page, choose Next.
5. To complete the uninstallation, choose Finish.

7.8.2 Performing a Silent Uninstallation

Using a silent uninstallation, system administrators can use the command line to automatically uninstall SAP
Predictive Analytics on any machine in their system without the need to use the Microsoft Windows
uninstallation wizard. The silent uninstallation is primarily geared towards users with network administration
roles. A silent uninstallation is particularly useful when you need to uninstall the software automatically in your
corporate network, for example to update the software to a newer release.

1. Choose Start Run and type cmd to open a Command Prompt window.
2. Run the following command.

cd "<installation folder>" ; "setup.exe" -u product.predictiveanalytics-4.0-


core-32 -q

where <installation folder> is the location where Desktop has been installed.

If Desktop is set up in C:\Program Files\SAP Predictive Analytics\Desktop, type the


command :cd “C:\Program Files\SAP Predictive Analytics\Desktop” ; "setup.exe" -u
product.predictiveanalytics-4.0-core-32 -q.

After a while, you can check that the installation directory is empty and that SAP Predictive Analytics
Desktop 3.3 is no longer displayed in Control Panel All Control Panel Items Programs and Features .

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Desktop PUBLIC 85
8 Installing SAP Predictive Analytics Client

8.1 Installing the SAP Predictive Analytics Client on


Windows

You need to log on as an Administrator to run the installation.

The SAP Predictive Analytics client is also called the Remote Assistant.

1. Execute the file named PACLIENT<version number>_<patch number>.EXE, for example


PACLIENT2000_0.EXE for version 2.0, patch level 0.
The Welcome dialog appears.
2. Click Next.
3. In the Corba Name Server Information dialog, enter the host and port.
This panel shows the settings to be used in the CORBA client application to connect to the server. By
default, these parameters are set for a local server. Typically you need to edit at least the field Name Server
host address to indicate on which machine the server is running. The values set here will be used in the
Windows shortcut created to start the client application. These parameters can be modified later in this
shortcut.
○ Name Server host address: this parameter indicates the address to be used to connect to the machine
on which the Name Server is running. It should be a valid IP address.
○ Name Server TCP port
Contact the administrator who has installed the server to get the correct parameters.
4. Click Next.
5. In the Select Destination Location dialog, enter the location where you want to install the remote assistant
and click Next.
6. In the Select Components dialog, select the components you want to install.

Option Description

Remote Assistant The client for SAP Predictive Analytics Enterprise edition.

Remote Shell Interpreter A simple command line interpreter that uses the C++ Library located on the server

Documentation Installs a link to SAP Predictive Analytics documentation on the SAP Help Portal.

Java Runtime Environment This is needed to execute the Java graphical interface. You need to accept the terms of
1.7 [...] SUN License agreement (http://java.sun.com/j2se/1.4.2/j2re-1_4_2_16-license.txt ).

7. Click Next.
The Select Start Menu Folder dialog is displayed.
8. Click Next.
The Select Additional Tasks dialog is displayed.
9. Click Next.
The Ready to Install dialog displays a summary of the settings you have selected for the installation.

Predictive Analytics Installation Guide


86 PUBLIC Installing SAP Predictive Analytics Client
10. Click Install.
The Installing panel is displayed and a progress bar allows you to follow the installation process.
Once the installation process is over, the Finish dialog is displayed.
11. Click Finish.

8.1.1 Performing a Silent Installation

You can use the SAP Predictive Analytics client self-extractor to create a response file required for a silent
installation. Using a silent installation, system administrators can run a script from the command line to
automatically install SAP Predictive Analytics client on any machine in their system without the setup program
prompting them for information or displaying the progress bar. The silent installation is primarily geared
towards users with network administration roles. A silent installation is particularly useful when you need to
push multiple installations in your corporate network. Once you have created a silent installation response file,
you can add the silent installation command to your installation scripts.

1. Choose Start Run and type cmd to open a Command Prompt window.

% <filename.exe> /SAVEINF=C:\Tmp\SilentInstall.inf

Note
<filename.exe> represents the name of the .exe file.

The SAP Predictive Analytics client Setup program opens.


2. Follow the installation wizard to select your setup options.
3. On the Start installation page, choose Next.
The setup program writes your installation options to the SilentInstall.inf file, and closes.
4. To run the silent installation, open a Command Prompt window and enter the following command:

% <filename.exe> /SILENT /LOADINF= C:\Tmp\SilentInstall.inf

To prevent any pop-up window to be displayed during installation, replace /SILENT by /VERYSILENT in
the command above.

8.2 Installing the SAP Predictive Analytics Client on Linux

Before installing SAP Predictive Analytics client, make sure that the correct Java Runtime Environment (JRE) is
set up on your computer.

To know which version is supported by SAP Predictive Analytics, refer the Product Availability Matrix athttp://
service.sap.com/sap/support/pam

You can use one of the following versions of the JRE:

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Client PUBLIC 87
● For all platforms, you can download the SAP JVM version that is needed from the SAP Software Download
Center :
1. On the homepage, click Download Software.
2. Enter the SAP JVM version that is relevant in the Enter search term field.
● For all platforms except IBM AIX, you can download the Oracle JRE version that is needed from the Oracle
website

Unzip and detar the client, for example in soft/:

$ gunzip -c </directory/containing/the *.TAR | tar -xvf -

Note
For additional information on how to install SAP Predictive Analytics client on AIX or Solaris platforms,
please refer to the note 2174673 .

8.3 Upgrading the Client

● When upgrading to a new version of the SAP Predictive Analytics client, you do not need to uninstall the
previous version: a new version can be set up in parallel with a previous version. If you want to uninstall the
previous version, first install the new version and check that it works before you uninstall the previous one
● You do not need a new license.

8.4 Checking the Connection between the Client and


Server

1. Go to Start SAP Business Intelligence SAP Predictive Analytics Client Remote Assistant .
2. Enter your user name and password.

3. On the home screen of the application, click the button.


4. Click the Help menu option About. The About SAP Predictive Analytics dialog opens.
If the Authenticated Server is running correctly, the field Modeling Server Status is set to Modeling Server
running. This panel also allows you to see the Product Version of the connected server.

You can also check that the correct licenses are in place on the server by verifying the tabs Generator
Licenses, Component Licenses, and Data Access Licenses.

Predictive Analytics Installation Guide


88 PUBLIC Installing SAP Predictive Analytics Client
8.5 Starting the Client for Linux Hosts

1. To connect to the SAP Predictive Analytics server / standard UI to perform modeling:

$ AutomatedAnalyticsClient_<SYSARCH>_v<version number>/KJWizardCORBA/
KJWizardCORBA-Authenticated.sh <Name of the server>

2. To connect to the SAP Predictive Analytics server / Control Panel UI

$ AutomatedAnalyticsClient_<SYSARCH>_v<version number>/KJWizardCORBA/
KxControlPanel.sh <Name of the server>

8.6 Troubleshooting

8.6.1 Installation Error Message

If the following error message is displayed during the installation of the remote application, it indicates that a
previous version of the remote assistant is already installed and currently running:

An error occurred while trying to replace the existing file:


DeleteFilefailed; code 32.
The process cannot access the file because it is being used by another process.
Click Retry to try again, Ignore to skip this file (not recommended), or Abort
to cancel installation.

To solve the problem, follow the instructions below:

1. Click Abort.
2. Stop the remote assistant currently running:

○ Start the Task Manager.


○ Select the application named KJWizardCorba.exe.
○ Click End Task.
3. In the existing installation directory, by default <INSTALLATION DIR>\Client <version number>
\EXE\Clients\KJWizardCORBA\, execute the file named unins000.exe to uninstall the previous
version of the remote assistant.
4. Repeat the procedure described in Installing the SAP Predictive Analytics Client on Windows [page 86].

8.6.2 Invalid or Missing License

If your license file is missing or out of date, the following warning message appears:

Invalid license to run <application name>.


Please contact your SAP representative.

Predictive Analytics Installation Guide


Installing SAP Predictive Analytics Client PUBLIC 89
To solve the problem, follow the instruction below:

1. Check that a valid license file named License.cfg exists in the installation directory of the server to be
connected to.
2. Check that the client was properly restarted to take into account the latest license file.

8.6.3 Full Client System Application Installation (Server and


Client)

If the Server application and the Remote Assistant are set up on the same system, check the following points:

○ The Name Server application is correctly running.


○ The Server application is correctly running.

8.6.4 Server and Remote Assistant Installed on Different


Systems

If the server and the Remote Assistant applications are set up on two different systems, check the following
points:

○ The client network system is correctly connected to the remote server. From a command tool application,
execute a ping application to your server IP, it should return a valid reply message.
○ Check that no firewall application, or any protection applications, blocks the server port number and/or
Remote Assistant.

8.6.5 Common Check Point

1. Go to Start SAP Business Intelligence SAP Predictive Analytics Client Remote Assistant <version
number> and right-click to display the Properties menu option.
2. In the Properties dialog, check if the input settings defined during the installation process correspond to
your server network address.

Example
With a local server, the Target shortcut value should be:

C:\Program Files\SAP Predictive Analytics\Client <version number>\EXE\Clients


\KJWizardCORBA\KJWizardCorba.exe"
-ORBInitialHost pvs0389 -ORBInitialPort 12345 -ServiceName FactoryEntries3 -
authenticated

Predictive Analytics Installation Guide


90 PUBLIC Installing SAP Predictive Analytics Client
9 Installing Java Web Start for Windows

9.1 About Java Web Start

Java Web Start allows you to use SAP Predictive Analytics in a client/server mode. The server application can
be launched from a local computer via the web interface. When the user accesses SAP Predictive Analytics
Java Web Start for the first time, the client install script runs and installs the client version of SAP Predictive
Analytics Java Web Start.

9.1.1 Requirements

The following components are required to install SAP Predictive Analytics to run on Java Web Start.

On the server:

● A web server such as Apache Web Server or Windows Internet Information Services (IIS).
● A valid full installation of SAP Predictive Analytics Server. For more information, please refer to the server
installation documentation provided on the SAP Help Portal at http://help.sap.com/pa/.
● A valid SAP Predictive Analytics License.

On the client machine(s):

● A Java Runtime Environment installed on the client machine to connect to the server.

To run the client application with Java Web Start, configure the network connection between the server and the
client machine(s). To configure the connection, open the following TCP ports:

● The HTTP port (usually, 80).


● The Name Server (usually port 12345).
● The server port (by default: 12347 for the authenticated server).

Java Web Start must be installed by a web server administrator, who can administrate and install a web server.

9.2 Installing the Web Server

You can install an Apache Server, or Windows Internet Information Services (IIS). How you install Windows IIS
depends on the Windows Operating System you have. You might also require your original Windows installation
media if IIS was not included at install time.

Predictive Analytics Installation Guide


Installing Java Web Start for Windows PUBLIC 91
9.2.1 Installing Microsoft IIS

Internet Information Services (IIS) provides a platform for hosting websites, services and applications. IIS is
not turned on by default when Windows is installed. Access the IIS Manager through the Microsoft
Management Console or Administrative Tools in the Control Panel, depending on your Windows Operating
System.

1. Go to Windows Control Panel and access the panel for configuring Features (Windows 2008 Enterprise
Server) or Internet Information Services.
2. Check the box Web Server (IIS)/Internet Information Services (the name for this option depends on your
OS).
3. Click Next.
○ In Windows 2008 Enterprise Server OS, an installation wizard appears, follow the instructions from the
wizard to complete the installation of IIS.
○ On a non-Windows 2008 Enterprise Server OS, IIS will launch automatically

9.2.2 Installing Apache Server

1. Download the package from the Apache HTTP Server Project (http://httpd.apache.org/)
2. Install the downloaded Apache Server package according to the Apache documentation (http://
httpd.apache.org/docs/current/en/platform/windows.html).

9.3 Setting Up the Web Server

9.3.1 Publishing with Microsoft IIS

By default, SAP Predictive Analytics is installed in c:\Program Files\SAP Predictive Analytics


\Server <version number>\. This folder should contain the following sub-folders:

● EXE\WebStart\
● EXE\Clients\KJWizardCORBA\
● Help\

Since the application will be accessible at http://<myserver.com>/PredictiveAnalytics/, these local


folders must be published by the web server as follows:

Local Directory Published as... Gives access to...

EXE\WebStart\ http://<myserver.com>/ The home page of SAP Predictive


PredictiveAnalytics/ Analytics Web Start.

Predictive Analytics Installation Guide


92 PUBLIC Installing Java Web Start for Windows
Local Directory Published as... Gives access to...

EXE\Clients\KJWizardCORBA\ http://<myserver.com>/ The SAP Predictive Analytics Remote


PredictiveAnalytics/ Assistant.
KJWizard

Help\ http://<myserver.com>/ The documentation.


PredictiveAnalytics/
Documentation

Note
If you cannot see the above directories, you probably did not perform a full install of the SAP Predictive
Analytics Server software.

When typing the URL in a web browser, the index page of SAP Predictive Analytics Web Start displays (with links
to the documentation, the SAP Predictive Analytics Remote Assistant and the Administration Console).

Note
● <myserver.com> refers to the domain name of your website. For instance if the application is deployed
on a local computer, by default the domain name is <localhost>. So the URL to type is: http://
localhost/PredictiveAnalytics/.
● The final slash ("/") is mandatory when typing the URL.

9.3.1.1 To Publish the Web Start Folders on Windows

Publish the SAP Predictive Analytics Web Start index page and the KJWizardCORBA and Help folders.

1. Open IIS Manager: Go to the Windows Administrative Tools panel and click Internet Information Services
(IIS) Manager/Internet Information Services (depending on your OS).
The Internet Information Services (IIS) Manager displays.
2. Publish the SAP Predictive Analytics Web Start index page: Expand the Hostname and Sites folder, right-
click the folder in which you want to deploy SAP Predictive Analytics Java Web Start (for instance, Default
Web Site), and in the contextual menu, choose Add Virtual Directory and do the following:

Adding the Web Start Virtual Directory


For Windows 2008 Enterprise Server OS Other Windows OS

○ Expand the Hostname and Sites folder, ○ Open Local computer Web Sites
○ Right-click the folder in which you want to deploy SAP ○ Right-click the folder in which you want to deploy SAP
Predictive Analytics Java Web Start (for instance, De­ Predictive Analytics Web Start (for instance, Default
fault Web Site), and in the contextual menu, choose
Web Site) and in the contextual menu, select New
Add Virtual Directory and enter the information
shown in the table below: Virtual Directory .
○ In the Virtual Directory Creation Wizard, click Next
and enter the information shown in the table below:

Predictive Analytics Installation Guide


Installing Java Web Start for Windows PUBLIC 93
Add Virtual Directory Settings
Field Setting

Alias Predictive Analytics

Physical path c:\Program Files\SAP Predictive


Analytics\Server <version number>\EXE
\WebStart

3. Click OK.
The folder Predictive Analytics displays.
4. Set the permissions for the directory as follows:

Setting the Virtual Directory Permissions


For Windows 2008 Enterprise Server OS Other Windows OS

Right-click the folder Predictive Analytics and choose Edit ○ Click Next and in the Access permissions panel, verify
permissions and make sure that the Read-Only box is that the boxes Read Scripts and Run Scripts are
checked in the Attributes area. checked.
○ Click Next
○ Click Finish

5. Publish the KJWizardCORBA and Help folders as follows:

Publishing the KJWizardCORBA and Help folders


For Windows 2008 Enterprise Server OS Other Windows OS

In IIS, right-click the Predictive Analytics folder and enter In IIS, right-click the Predictive Analytics folder, select
the following information: New Virtual Directory and enter the following infor­
mation:

Setting the Alias and Path


For the folder... Enter the alias Browse to...

KJWizardCORBA KJWizard c:\Program Files\SAP


Predictive Analytics
\Server <version number>
\EXE\Clients
\KJWizardCORBA

Help Documentation c:\Program Files\SAP


Predictive Analytics
\Server <version number>
\Help

Note
By default, IIS searches for the index.htm file, not for index.html. Change this default setting.

Predictive Analytics Installation Guide


94 PUBLIC Installing Java Web Start for Windows
9.3.1.2 To Configure IIS to Search For the index.html File

By default, IIS searches for an index.htm file first. In the Default Documents pane of IIS, you can define the file
index.html and move it in the search order to ensure that IIS searches for the index.html file first.

1. In the IIS Administrative Tools pane, click Internet Information Services (IIS) Manager.
2. In the Connections pane, expand the server name, expand Sites, and then navigate to the Home pane of the
Web site you want to configure default documents, and double-click Default Document. In the Documents
tab, click Add.
3. In the Actions pane, you can either select Add Default Document and add the file name index.html, or, if
the file already exists in the default documents list, select the index.html file and then click Move Up to
force IIS to search for index.html first.

9.3.1.3 To Configure the MIME Types

A MIME type indicates to the server what kind of file is associated with a particular extension. The
application/x-java-jnlp-file file must be associated with the jnlp extension (so the application can
be launched), and the application/json-file file must be associated with the json extension.

In IIS Manager, Add the following MIME types:

Configuring MIME types


File Name Extension MIME Type

jnlp application/x-java-jnlp-file

json application/json-file

The MIME types are configured. To remove a MIME type: select the MIME type in the MIME types list and click
Remove in the Actions pane.

9.3.2 Publishing with Apache HTTP Server

We assume you have installed SAP Predictive Analytics in [path_to_Predictive_Analytics]. This folder
should contain the following sub-folders:

● EXE/WebStart
● EXE/Clients
● Help

Since the application will be accessible at http://<myserver.com>/PredictiveAnalytics/, these local


folders must be published by the web server as follows:

Predictive Analytics Installation Guide


Installing Java Web Start for Windows PUBLIC 95
Local Directory Published as... Gives access to...

EXE/WebStart http://<myserver.com>/ The home page of SAP Predictive


PredictiveAnalytics/ Analytics Web Start.

EXE/Clients http://<myserver.com>/ The SAP Predictive Analytics Remote


PredictiveAnalytics/ Assistant.
KJWizard

Help http://<myserver.com>/ The documentation.


PredictiveAnalytics/
Documentation

When typing the URL in a web browser, the index page of SAP Predictive Analytics Web Start displays (with links
to the documentation, the SAP Predictive Analytics Remote Assistant and the Administration Console).

Note
● <myserver.com> refers to the domain name of your website. For instance if the application is deployed
on a local computer, by default the domain name is <localhost> . So the URL to type is: http://
localhost/PredictiveAnalytics/ .
● The final slash ("/") is mandatory when typing the URL.

9.3.2.1 Publishing the Web Start Folders

In the following commands, replace [path_to_Predictive_Analytics] with the installation path.

If you have installed the Apache server in APACHE_HOME, edit the APACHE_HOME/conf/httpd.conf

1. Add the following lines:

Alias /PredictiveAnalytics/KJWizard/ "[path_to_Predictive_Analytics]/EXE/


Clients/"
Alias /PredictiveAnalytics/Documentation/"[path_to_Predictive_Analytics]/
Help/"
Alias /PredictiveAnalytics/Documentation/"[path_to_Predictive_Analytics]/Exe/
WebStart/"

2. To define the access rights for those folders for the entire web directory (and all its sub-folders) use the
following syntax:

<Directory "[path_to_Predictive_Analytics]/">
Satisfy any
Order allow,deny
Allow from all
</Directory>

3. To define specific access rights for each sub-folder of your web directory, use the following syntax:

<Directory "[path_to_Predictive_Analytics]/KJWizardCORBA/">
Satisfy any
Order allow,deny
Allow from all
</Directory>
<Directory "[path_to_Predictive_Analytics]/WebStart/">
Satisfy any

Predictive Analytics Installation Guide


96 PUBLIC Installing Java Web Start for Windows
Order allow,deny
Allow from all
</Directory>
<Directory "[path_to_Predictive_Analytics]/Doc/">
Satisfy any
Order allow,deny
Allow from all
</Directory>

● For specific access rights, indicate if you want to allow or deny access for each of the folders.
● The final slash ("/") is mandatory when typing the folders location.

9.3.2.2 Configuring the MIME Types

A MIME type indicates to the server what kind of file is associated with a particular extension. The JNLP MIME
type allows launching the application, whereas the JSON MIME type allows displaying the online help.

Edit the APACHE_HOME/conf/mime.types file.

○ To associate the application/x-java-jnlp-file with the jnlp extension, add the following line:

application/x-java-jnlp-file jnlp

○ To associate the application/json-file with the json extension, add the following line:

application/json-file json

Note
On Apache, this JNLP MIME type will be applied on all the web sites hosted on the server.

9.4 Configuring the Remote Assistant

Optionally, you may want to change some parameters of the launcher such as the default connection port (see
the ORBInitialPort value).

Refer to the section Installing SAP Predictive Analytics Client, which describes how to modify default settings in
the installation according to your needs.

9.5 Setting the SSL Certificates

Certificates allow clients to authenticate the server they are connected to. If there is a certificate deployed on
one of the authenticated servers, you must activate it.

Predictive Analytics Installation Guide


Installing Java Web Start for Windows PUBLIC 97
For more information on using SSL with SAP Predictive Analytics please refer to the SSL Support
Documentation.

1. Edit the file KJWizard_auth.jnlp.


2. Remove the comment tags <!-- and --> in the following line:

<!-- <argument>-checkcertificates</argument> -->

Should look like this:

<argument>-checkcertificates</argument>

9.6 Firewall Settings

In order to ensure a good communication between SAP Predictive Analytics Client application and the
Authenticated Server, the firewall must be set to allow communication between the different software (HTTP
Server, Name Server and Authenticated Server).

To configure the firewall, proceed as follows:

1. Open the Firewall Settings panel. By default, the firewall should be ON and should allow exceptions (if it is
not the case, uncheck Don't allow exceptions).
2. In the Exceptions tab, add the following ports:

Adding Ports to the Firewall Settings


Port Name to Add Value

HTTP 80

Name Server 12345

Predictive Analytics Authenticated 12347


Server

Note that the port may have been changed in the Registry. In this case the port number should be set
accordingly.

9.7 Starting the Remote Assistant and the Administration


Console

Make sure your web server and your SAP Predictive Analytics server are running.

1. Point your web browser to the defined URL, in our example:


http://myserver.com/PredictiveAnalytics/
2. Click Start Predictive Analytics Client.
3. Click Start Predictive Analytics Administration Console.

Predictive Analytics Installation Guide


98 PUBLIC Installing Java Web Start for Windows
4. After the installation and configuration are complete, please read the chapter How to Launch Predictive
Analytics Java Web Start for the First Time provided in the documentation.

9.8 Allowing Client Access to SAP Predictive Analytics Java


Web Start Interface

Here is the information you should provide to your users that want to access the SAP Predictive Analytics
Server from a client machine.

Provide your users with the following information: To access the SAP Predictive Analytics Java Web Start
Interface, go to the following URL:

http://<myserver.com>/PredictiveAnalytics.

Note
myserver.com refers to the domain name of your website. For instance, if SAP IniniteInsight is deployed on
a local computer, by default the domaine name is localhost. So the URL to type is http://localhost/
InfiniteInsight/.

The final slash character ("/") is mandatory when typing the URL.

The client installation program will run and install the client software. The user can start SAP Predictive
Analytics from the client machine by clicking Start SAP Predictive Analytics Client.

Predictive Analytics Installation Guide


Installing Java Web Start for Windows PUBLIC 99
10 Installing Java Web Start for Linux

10.1 About Java Web Start


Java Web Start allows you to use SAP Predictive Analytics in a client/server mode. The server application can
be launched from a local computer via the web interface. When the user accesses SAP Predictive Analytics
Java Web Start for the first time, the client install script runs and installs the client version of SAP Predictive
Analytics Java Web Start.

10.1.1 Requirements
The Java Web Start technology is based on distributing application through standard web servers. The
following components are required to install SAP Predictive Analytics Web Start:

On the server:

● A web server (Apache Web Server for example). See the sections below to install Apache Web Server (see
page 4 ).
● A valid installation of SAP Predictive Analytics. For more information, please refer to the server installation
documentation provided on the SAP Help Portal at http://help.sap.com/pa/.
● A valid SAP Predictive Analytics License.

On the client machine(s):

● A Java Runtime Environment properly installed on the client machine that will require to connect to the
authenticated server. To know the supported version, refer to the SAP Product Availability Matrix http://
service.sap.com/sap/support/pam

To run SAP Predictive Analytics Client application with Java Web Start, you need to configure the network
connection between the server and the client machine(s). To configure the connection, open the following TCP
ports:

● The HTTP port (usually, 80).


● The Name Server (usually port 12345).
● The server port (by default: 12347 for the authenticated server).

SAP Predictive Analytics Web Start must be installed by a web server administrator, who can administrate and
install a web server.

10.2 Installing the Web Server


To Install an Apache Server, download and install the package from the Apache HTTP Server Project (http://
httpd.apache.org/).

Predictive Analytics Installation Guide


100 PUBLIC Installing Java Web Start for Linux
Note
For more information on the installation, please refer to the Apache documentation (http://
httpd.apache.org/docs/current/en/install.html).

10.3 Publishing with Apache HTTP Server

We assume you have installed SAP Predictive Analytics in [path_to_Predictive_Analytics]. This folder
should contain the following sub-folders:

● EXE/WebStart
● EXE/Clients
● Help

Since the application will be accessible at http://<myserver.com>/PredictiveAnalytics/, these local


folders must be published by the web server as follows:

Local Directory Published as... Gives access to...

EXE/WebStart http://<myserver.com>/ The home page of SAP Predictive


PredictiveAnalytics/ Analytics Web Start.

EXE/Clients http://<myserver.com>/ The SAP Predictive Analytics Remote


PredictiveAnalytics/ Assistant.
KJWizard

Help http://<myserver.com>/ The documentation.


PredictiveAnalytics/
Documentation

When typing the URL in a web browser, the index page of SAP Predictive Analytics Web Start displays (with links
to the documentation, the SAP Predictive Analytics Remote Assistant and the Administration Console).

Note
● <myserver.com> refers to the domain name of your website. For instance if the application is deployed
on a local computer, by default the domain name is <localhost> . So the URL to type is: http://
localhost/PredictiveAnalytics/ .
● The final slash ("/") is mandatory when typing the URL.

10.3.1 Publishing the Web Start Folders

In the following commands, replace [path_to_Predictive_Analytics] with the installation path.

If you have installed the Apache server in APACHE_HOME, edit the APACHE_HOME/conf/httpd.conf

Predictive Analytics Installation Guide


Installing Java Web Start for Linux PUBLIC 101
1. Add the following lines:

Alias /PredictiveAnalytics/KJWizard/ "[path_to_Predictive_Analytics]/EXE/


Clients/"
Alias /PredictiveAnalytics/Documentation/"[path_to_Predictive_Analytics]/
Help/"
Alias /PredictiveAnalytics/Documentation/"[path_to_Predictive_Analytics]/Exe/
WebStart/"

2. To define the access rights for those folders for the entire web directory (and all its sub-folders) use the
following syntax:

<Directory "[path_to_Predictive_Analytics]/">
Satisfy any
Order allow,deny
Allow from all
</Directory>

3. To define specific access rights for each sub-folder of your web directory, use the following syntax:

<Directory "[path_to_Predictive_Analytics]/KJWizardCORBA/">
Satisfy any
Order allow,deny
Allow from all
</Directory>
<Directory "[path_to_Predictive_Analytics]/WebStart/">
Satisfy any
Order allow,deny
Allow from all
</Directory>
<Directory "[path_to_Predictive_Analytics]/Doc/">
Satisfy any
Order allow,deny
Allow from all
</Directory>

● For specific access rights, indicate if you want to allow or deny access for each of the folders.
● The final slash ("/") is mandatory when typing the folders location.

10.3.2 Configuring the MIME Types

A MIME type indicates to the server what kind of file is associated with a particular extension. The Jnlp mime
type allows launching the application, whereas the Json mime type allows displaying the online help.

Edit the APACHE_HOME/conf/mime.types file.

○ To associate the application/jnlp-file with the jnlp extension, add the following line:

application/x-java-jnlp-file jnlp

○ To associate the application/json-file with the json extension, add the following line:

application/json-file json

Note
On Apache, this JNLP MIME type will be applied on all the web sites hosted on the server.

Predictive Analytics Installation Guide


102 PUBLIC Installing Java Web Start for Linux
10.4 Configuring the Remote Assistant
To run SAP Predictive Analytics properly, you need to set up the Remote Assistant.

To configure the Remote Assistant, proceed as follows:

1. Go to the folder [path_to_PredictiveAnalytics]/WebStart/


2. Copy the file index_auth_xx.html
3. Paste it as index.html in the same folder
4. Open the KJWizard_auth.jnlp file
5. Change the value of the codebase attribute from localhost to the value corresponding to your web server
address:

<?xml version="1.0" encoding="UTF-8"?>


<jnlp spec="1.0+" codebase="http://localhost/PredictiveAnalytics/"
href="KJWizard_auth.jnlp">

</application-desc>
</jnlp>

6. Change the value of the argument ORBInitialHost from localhost to the value corresponding to your
SAP Predictive Analytics server address.

<?xml version="1.0" encoding="UTF-8"?>



<application-desc main-class="com.kxen.KJWizard.KJWizard">
<argument>-ORBInitialPort</argument>
<argument>12345</argument>
<argument>-ORBInitialHost</argument>
<argument>localhost</argument>
<argument>-authenticated</argument>
</application-desc>
</jnlp>

7. Optionally, you may also want to change other parameters of the launcher like the default connection port
(see the ORBInitialPort value). To do so, refer to the section Installing SAP Predictive Analytics Client,
which describes how to modify the default settings and configure the installation according to your needs.

10.5 Configuring the Administration Console


1. Go to the folder [path_to_PredictiveAnalytics]/WebStart/
2. Edit the file KxAdminConsole.jnlp
3. Change the value of the attribute codebase from localhost to the value corresponding to your server.<?xml
version="1.0" encoding="UTF-8"?><jnlp spec="1.0+" codebase="http://localhost/
PredictiveAnalytics/" href="KxAdminConsole.jnlp">...</jnlp>

<?xml version="1.0" encoding="UTF-8"?>


<jnlp spec="1.0+" codebase="http://localhost/PredictiveAnalytics/"
href="KxAdminConsole.jnlp">
...
</jnlp>

Predictive Analytics Installation Guide


Installing Java Web Start for Linux PUBLIC 103
4. Change the value of the argument ORBInitialHost from localhost to the value corresponding to your
server.

<application-desc main-class="com.kxen.KxConfiguration.KxConfiguration">
<argument>-ORBInitialPort</argument>
<argument>12345</argument>
<argument>-ORBInitialHost</argument>
<argument>localhost</argument>
<argument>-authenticated</argument>
</application-desc>

10.6 Setting the SSL Certificates

Certificates allow clients to authenticate the server they are connected to. If there is a certificate deployed on
one of the authenticated servers, you must activate it.

For more information on using SSL with SAP Predictive Analytics please refer to the SSL Support
Documentation.

1. Edit the file KJWizard_auth.jnlp.


2. Remove the comment tags <!-- and --> in the following line:

<!-- <argument>-checkcertificates</argument> -->

Should look like this:

<argument>-checkcertificates</argument>

10.7 Firewall Settings

In order to ensure a good communication between the Client application and the Authenticated Server, the
firewall must be set to allow communication between the different software (HTTP Server, Name Server and
Authenticated Server).

Available firewalls in Linux environments are numerous. Please refer to the user guide of your installed firewall
to know how opening applicative ports.

Open these applicative ports:

● port 80 for HTTP


● port 12345 for NameServer
● port 12347 for Authenticated Server

Predictive Analytics Installation Guide


104 PUBLIC Installing Java Web Start for Linux
10.8 Starting the Remote Assistant and the Administration
Console

Make sure your web server and your SAP Predictive Analytics server are running.

1. Point your web browser to the defined URL, in our example:


http://myserver.com/PredictiveAnalytics/
2. Click Start Predictive Analytics Client.
3. Click Start Predictive Analytics Administration Console.
4. After the installation and configuration are complete, please read the chapter How to Launch Predictive
Analytics Java Web Start for the First Time provided in the documentation.

Predictive Analytics Installation Guide


Installing Java Web Start for Linux PUBLIC 105
11 Running Multiple Instances of Java Web
Start

11.1 Overview

A standard SAP Predictive Analytics installation contains the following components:

● The client application


● The server application itself composed of:
○ The CORBA Name service (the default port is 12345)
○ The application authenticated server (the default port is 12347)
● The Java Web Start launcher, that requires a Web server to be operational (the default port is 80).

You can have several versions of the application server running simultaneously, for example:

● To test the latest version of the application without impacting the current production environment.
● To use a feature of version Y when it is not possible to change the current version X (thus to have both X
and Y running at the same time).

This guide provides the steps to configure the system in order to run several SAP Predictive Analytics servers
simultaneously.

For simplicity reasons, this documentation is based on a clear-cut example to clarify the different concepts and
steps.

It is, however, relatively simple to use as a basis for installing more than two versions of the application server.

11.1.1 General Multiple Installation Architecture

This demonstration scenario shows SAP Predictive Analytics version 2.0 and SAP InfiniteInsight version 6.1.0
running at the same time on a Linux server.

The Linux Server is running Apache HTTP Server.

The application Authenticated servers will be accessed through the application Java Web Start launcher.

Predictive Analytics Installation Guide


106 PUBLIC Running Multiple Instances of Java Web Start
The server architecture is the following:

Note
Please ensure the different services receive connections on different port numbers, in order to avoid
dysfunctions.

It is also necessary to have two web URLs setup for the application Web Start launchers, one for each
application version.

11.2 Installing SAP Predictive Analytics

This section provides guidance for the specific case of installing and configuring multiple versions of the
application.

Predictive Analytics Installation Guide


Running Multiple Instances of Java Web Start PUBLIC 107
For further details on how to install the application server, please refer to the section Installing Java Web Start.
It provides information on further aspects of the installation process that will not be repeated in this document.

11.2.1 Server Side

The installation process does not differ from a normal installation. Please refer to the section Installing Java
Web Start to guide you through the different steps.

On Microsoft Windows, you must choose between installing the application as a Windows service or as a
standard application:

● On a test environment, the software may be installed as a standard application.


● It is recommended to install the application as a Windows service on a production environment.

11.2.2 Client Side

There are two possibilities for the client machines:

● Install the client application in order to connect to the server. If you choose to install the client application,
please refer to the section Installing SAP Predictive Analytics Client.
● Use a Web browser to connect to the Java Web Start launcher deployed on the server.

Note
Using a web browser is the recommended mode if several users need to connect to the application server. It
allows simplifying the configuration and update processes (only the server needs to be configured).

11.3 Configuring the Application Server

This section describes the configuration steps to have two versions of the application server running
simultaneously. For each server, the configuration will cover the following elements:

● License files
● Connection ports
● Log files
● Java Web Start files

11.3.1 Managing Your Software License(s)

Ensure that the configuration files point to the right licenses. In case of problems using one of the servers,
check if the license configuration is correct.

Predictive Analytics Installation Guide


108 PUBLIC Running Multiple Instances of Java Web Start
11.3.2 Log Files Configuration

After configuring the Windows services and the application server port numbers, it is necessary to make sure
the different application servers will not write their outputs in the same log files. This would make them
particularly difficult to read.

The default file describing the log policy of the application servers is named logconf.txt.

11.3.2.1 Configuring the Log Files

The file logconf.txt is located in AuthServer. This file has the two following log file names declared.

Root.Directory $KXTEMPDIR
Root.Filename kxen_auth_log.txt
...
Usage.Directory $KXTEMPDIR
Usage.Filename kxenusagelog.txt

Authenticated servers use two different log entries:

● The Root entries are used for tracing the activity of the server which can later be used for debugging
purposes.
● The Usage entries are sparser and report a more global view of the activity on the server. They consist of
one line per event on the following events: user connection, user disconnection, training a model, applying
a model, saving a model.

For the demonstration scenario used in this documentation we chose to change the name of the files so that
the log files remain located in the same folder under different names.

So we get two logconfig.txt files, one for each software version/install, in the following directories:

● C:\Program Files\SAP InfiniteInsight\InfiniteInsightV <version>\EXE\Servers


\AuthServer
● C:\Program Files\SAP Predictive Analytics\Server <version>\EXE\Servers
\AuthServer

For clarity, rename the log files in the respective logconfig.txt files. In this example, the logconfig.txt
file for SAP InfiniteInsight v6.1.0 was edited to reflect the following changes:

Root.Filename kxen_610_auth_log.txt
...
Usage.Filename kxen_610_usagelog.txt

The logconfig.txt file for SAP Predictive Analytics was edited to reflect the following changes:

Root.Filename pa_20_auth_log.txt
...
Usage.Filename pa_20_usagelog.txt

Predictive Analytics Installation Guide


Running Multiple Instances of Java Web Start PUBLIC 109
11.4 Microsoft Windows Application Mode

The schema below provides a partial view of the installation folders on a Microsoft Windows platform. It shows
the files that need to be configured in order to run multiple application versions simultaneously.

Please use it as a reference to retrieve the files mentioned in the procedure steps below

Note
Each server must run on a different port number. If necessary, edit the configuration in order to allocate the
right port numbers to the different services before launching the applications.

11.4.1 Configuring the Service Ports and Starting the Server

Do the following:

● Configure the CORBA Named Service Port


● Configure the Authenticated Server Port
● Start the server

11.4.1.1 To Configure the CORBA Name Service Port

When you run more than one installation, you must make sure that each installation uses a unique CORBA
Name Service port.

Predictive Analytics Installation Guide


110 PUBLIC Running Multiple Instances of Java Web Start
1. For each installation, open the file nameserv.bat (in the <application_server_folder>\EXE
\Namesrv folder).
2. For each installation, edit the line set NS_PORT=12345 to set a different (unique) number. For example:
set NS_PORT=12445. Note that it is possible to enter any free port number here.

Your different applications should use different CORBA Named Service port numbers. Our example uses the
following:

CORBA Name Service Port Settings


Installation CORBA Name Service Port

SAP Predictive Analytics 2.0 set NS_PORT=12345

InfiniteInsight v6.1.0 set NS_PORT=12445

11.4.1.2 To Configure the Authenticated Server

When you run more than one version of the application, you must make sure that each installation uses a
unique authenticated server port.

1. For each installation, open the runserver.bat file (in the <application_server_folder>\Servers
\AuthServer folder).
2. For each installation, edit the line set KXEN_PORT=12347 to set a different (unique) number. For example:
set KXEN_PORT=12447. Note that it is possible to enter any free port number here.
3. Edit the line set NS_PORT=12345 so that it matches the line in the nameserv.bat file described in step 2
of the procedure To Configure the CORBA Name Service Port.

11.4.1.3 Starting the Servers

For each version of the software you are running, do the following:

1. Double-click the file namesrv.bat to start the CORBA name server.


2. Run the script corresponding to your configuration: double-click the file runserver.bat to start the
authenticated server.

11.5 Microsoft Windows Service Mode

The SAP Predictive Analytics Windows Services are registered under the following names:

SAP Predictive Analytics Directory (v<version number>) for the CORBA Name Service.

SAP Predictive Analytics - Automated Analytics Server (v<version number>) for the Server.

Predictive Analytics Installation Guide


Running Multiple Instances of Java Web Start PUBLIC 111
11.5.1 Configuring the Service Ports and Registering the
Services

For each version of the application you are running, do the following:

● Configure the CORBA Named Service Port


● Configure the Authenticated Server Port
● Register the Services

11.5.1.1 To Configure the CORBA Name Service

For each version of the application you are running, do the following:

1. Open the file KxRegisterNameService.bat.


2. Edit the line set NS_PORT=12345 so that the port number used is unique in your installation. For example,
set NS_PORT=12445 (you may enter any free port number here).

11.5.1.2 To Configure the Authenticated Server

For each version of the application you are running, do the following:

1. Open the file KxRegisterAuthService.bat.


2. Edit the line set KXEN_PORT=12347 so that the port number used is unique in your installation. For
example, set KXEN_PORT=12447(you may enter any free port number here).
3. For each application, edit the line set NS_PORT=12345 so that it matches the value used in the
corresponding KxRegisterNameService.bat file (use the port number of step 2 of the procedure To
Configure the CORBA Name Service).

11.5.1.3 To Register the Services

For each version of the application you are running, do the following:

1. In the NameSrv folder, double-click the file KxRegisterNameService.bat to run it.


2. In the AuthServer folder, double-click the file KxRegisterAuthService.bat to run it.

Predictive Analytics Installation Guide


112 PUBLIC Running Multiple Instances of Java Web Start
11.6 Linux Services

This document is aimed at system administrators, therefore it does not provide detailed information on how to
install daemons on the system. It will focus on the configuration of the port numbers used by SAP Predictive
Analytics services.

The required files are located at the root of the application installation folder:

● KxenCorbaConfig.sh: that controls the port numbers used by SAP InfiniteInsight ® .


● KxenServer_initd: that may be helpful to install a daemon. The script is commented and only requires
setting the KXEN_INSTALLATION_PATH variable, do not hesitate to refer to it.

The schema below provides a partial view of the installation folder of the application on a Linux platform. It
shows the files that need to be configured in order to run multiple application versions simultaneously.

Please use it as a reference to retrieve the files mentioned in the procedure steps below:

11.6.1 To Configure the Services

For each version of the application you are running, do the following:

1. Open the file KxenCorbaConfig.sh.


2. Replace the line NS_PORT=12345 by NS_PORT=12445(It is possible to enter any free port number here).
3. To configure Authenticated Server, replace the line KXEN_AUTH_PORT=12347 by
KXEN_AUTH_PORT=12447 (It is possible to enter any free port number here).

11.7 Connecting to the Server

If you have installed the SAP Predictive Analytics Remote Assistant on a computer, to connect to the server you
need to edit its configuration so it can connect to the server. These changes must be done on all the computers
on which the remote assistant is installed

Predictive Analytics Installation Guide


Running Multiple Instances of Java Web Start PUBLIC 113
Note
We recommend you to connect to the SAP Predictive Analytics Server via the SAP Predictive Analytics Java
Web Start launcher. With this option, only the host server needs to be configured. All the client machines
retrieve the right configuration thanks to the Java Web Start launcher.

11.7.1 On MS-Windows Platforms

The diagram below provides a partial view of the installation folder of the application on a Microsoft Windows
client-machine. It shows the files that need to be configured so the Remote Assistant can connect to a custom
server.

Please use it as a reference to retrieve the files mentioned in the procedure steps.

11.7.1.1 To Configure the Remote Assistant

1. Open the file KJWizardCorba.ini.


2. Edit the line arg.x=12345 to match the port number of step 2 of the procedure To Configure the CORBA
Name Service described in section Ms-Windows Services.
3. Click Programs > SAP Business Intelligence > SAP Predictive Analytics
Server>Remote Assistant 2.0 to connect to the application server.

Predictive Analytics Installation Guide


114 PUBLIC Running Multiple Instances of Java Web Start
11.7.1.2 To Configure the Control Panel

1. Open the file KxControlPanel.ini.


2. Edit the line arg.x=12345 to match the port number of step 2 of the procedure To Configure the CORBA
Name Service described in section Ms-Windows Services).
3. Click Programs > SAP Business Intelligence > SAP Predictive Analytics
Server>Control Pane 2.0 to connect to the server.

11.7.2 On Linux Platforms

The diagram below provides a partial view of the installation folder of the application on a Linux client-machine.
It shows the files that need to be configured so the Remote Assistant can connect to a custom server.

By default the file KxenCorbaConfig.sh contains the configuration for the port numbers. If you have already
configured this file, the Remote Assistant should access the server correctly.

If you have not configured it yet, please follow the procedure.

11.7.2.1 To Configure the Remote Assistant and the Control


Panel

1. Edit the file KxenCorbaConfig.sh.

Predictive Analytics Installation Guide


Running Multiple Instances of Java Web Start PUBLIC 115
2. Edit the line NS_PORT=12345 so that it matches the port number of step 2 of the procedure To Configure
the KXEN Services described in section Linux Services. We recommend making a copy of the file before
editing it.
3. To start the Remote Assistant and connect to the server, run the following command: $ KJWizardCORBA/
KJWizardCORBA.sh
4. To start the Control Panel and connect to the server, run the following command:$ KJWizardCORBA/
KxControlPanel.sh

11.8 Setting Multiple Java Web Start Launchers

Once have successfully configured and started the services, you need to configure the Java Web Start launcher
so that the remote computers can access the servers by simply connecting to a given URL and clicking the
Start SAP Predictive Analytics button.

The easiest way to achieve this is by deploying each Java Web Start launcher on a different web context, for
instance: make SAP Predictive Analytics 2.0 accessible from http://host/PA_20/ and SAP InfiniteInsight ®
version 6.1.0 from http://host/KXEN_610/.

11.8.1 JNLP Files

For each SAP Predictive Analytics version you need to deploy, you must configure the Java Web Start launcher
related to it. Depending on whether you are running an authenticated server or not, you will have to edit
respectively two files or one file.

This chapter only details the settings needed to run multiple SAP Predictive Analytics versions. For information
on the installation of the Web Start launcher and on the configuration of the Web server refer to the Java Web
Start Installation Guide for Windows or Linux.

The following JNLP files need to be edited:

● KJWizard_auth.jnlp: used to connect to an SAP Predictive Analytics Server.


● KxAdminConsole.jnlp: used to control the connections to the server.

Predictive Analytics Installation Guide


116 PUBLIC Running Multiple Instances of Java Web Start
For Windows platforms, they are located in the following directory:

For Linux platforms, they are located in the following directory:

There are three parameters of the JNLP files that need to be edited to correspond to the settings of your
server:

● The codebase: indicates where the Java Web Start Launcher is deployed.
● The ORBInitialPort: indicates which port the server is listening to.
● The ORBInitialHost: indicates the host on which the server is deployed.

Predictive Analytics Installation Guide


Running Multiple Instances of Java Web Start PUBLIC 117
Typical JNLP file structure:

<?xml version="1.0"
encoding="UTF-8"?><jnlp spec="1.0+"
codebase="http://localhost/KXEN/" href="xxx.jnlp">

<argument>-ORBInitialPort</argument>
<argument>12345</argument>
<argument>-ORBInitialHost</argument>
<argument>localhost</argument>

For the demonstration scenario used in this documentation, we need two Java Web Start Launchers, one for
each software version.

11.8.1.1 Configuring the SAP Java Web Start Launchers

The example below shows the changes made on the files KJWizard_auth.jnlp and KxAdminConsole.jnlp for SAP
Predictive Analytics:

<?xml version="1.0"
encoding="UTF-8"?><jnlp spec="1.0+"
codebase="http://host/PA_20/" href="xxx.jnlp">

<argument>-ORBInitialPort</argument>
<argument>12345</argument>
<argument>-ORBInitialHost</argument>
<argument>host</argument>

Note
Please ensure to make these changes on both files.

The example below shows the changes made on the files KJWizard_auth.jnlp and KxAdminConsole.jnlp for SAP
InfiniteInsight ® version 6.1.0:

<?xml version="1.0"
encoding="UTF-8"?><jnlp spec="1.0+"
codebase="http://host/KXEN_610/" href="xxx.jnlp">

<argument>-ORBInitialPort</argument>
<argument>12445</argument>
<argument>-ORBInitialHost</argument>
<argument>host</argument>

Note
Please ensure to make these changes on both files.

Predictive Analytics Installation Guide


118 PUBLIC Running Multiple Instances of Java Web Start
11.8.2 Web Server Mapping

The last step is to ensure that the web server has been set to reflect the right mapping between the SAP
Predictive Analytics installation folders and the web URLs.

For Apache HTTP, the configuration used for this demonstration scenario is the following:

#
# THIS SECTION DEFINES THE HTTP CONFIGURATION TO PUBLISH KXEN VIA WEBSTART
#
ALIAS /PA_20/KJWIZARD/ "/SOFTS/PA_V20/KJWIZARDCORBA/"
ALIAS /PA_20/DOCUMENTATION/ "/SOFTS/PA_V20/DOC/"
ALIAS /PA_20/ "/SOFTS/PA_20/WEBSTART/"
<DIRECTORY "/SOFTS/PA_V20/">
ORDER ALLOW,DENY
ALLOW FROM ALL
</DIRECTORY>
ALIAS /KXEN_610/KJWIZARD/ "/SOFTS/KXEN_V6_1_0/KJWIZARDCORBA/"
ALIAS /KXEN_610/DOCUMENTATION/ "/SOFTS/KXEN_V6_1_0/DOC/"
ALIAS /KXEN_610/ "/SOFTS/KXEN_V6_1_0/WEBSTART/"
<DIRECTORY "/SOFTS/KXEN_V6_1_0/">
ORDER ALLOW,DENY
ALLOW FROM ALL
</DIRECTORY>

Once these steps are correctly followed, you can send an e-mail to the users providing them the URLs from
which they can run the software.

Note
Make sure to specify which URL matches which software version.

Predictive Analytics Installation Guide


Running Multiple Instances of Java Web Start PUBLIC 119
Important Disclaimers and Legal Information

Hyperlinks
Some links are classified by an icon and/or a mouseover text. These links provide additional information.
About the icons:

● Links with the icon : You are entering a Web site that is not hosted by SAP. By using such links, you agree (unless expressly stated otherwise in your
agreements with SAP) to this:

● The content of the linked-to site is not SAP documentation. You may not infer any product claims against SAP based on this information.
● SAP does not agree or disagree with the content on the linked-to site, nor does SAP warrant the availability and correctness. SAP shall not be liable for any
damages caused by the use of such content unless damages have been caused by SAP's gross negligence or willful misconduct.

● Links with the icon : You are leaving the documentation for that particular SAP product or service and are entering a SAP-hosted Web site. By using such
links, you agree that (unless expressly stated otherwise in your agreements with SAP) you may not infer any product claims against SAP based on this
information.

Beta and Other Experimental Features


Experimental features are not part of the officially delivered scope that SAP guarantees for future releases. This means that experimental features may be changed by
SAP at any time for any reason without notice. Experimental features are not for productive use. You may not demonstrate, test, examine, evaluate or otherwise use
the experimental features in a live operating environment or with data that has not been sufficiently backed up.
The purpose of experimental features is to get feedback early on, allowing customers and partners to influence the future product accordingly. By providing your
feedback (e.g. in the SAP Community), you accept that intellectual property rights of the contributions or derivative works shall remain the exclusive property of SAP.

Example Code
Any software coding and/or code snippets are examples. They are not for productive use. The example code is only intended to better explain and visualize the syntax
and phrasing rules. SAP does not warrant the correctness and completeness of the example code. SAP shall not be liable for errors or damages caused by the use of
example code unless damages have been caused by SAP's gross negligence or willful misconduct.

Gender-Related Language
We try not to use gender­specific word forms and formulations. As appropriate for context and readability, SAP may use masculine word forms to refer to all genders.

Predictive Analytics Installation Guide


120 PUBLIC Important Disclaimers and Legal Information
Predictive Analytics Installation Guide
Important Disclaimers and Legal Information PUBLIC 121
go.sap.com/registration/
contact.html

© 2018 SAP SE or an SAP affiliate company. All rights reserved.


No part of this publication may be reproduced or transmitted in any
form or for any purpose without the express permission of SAP SE
or an SAP affiliate company. The information contained herein may
be changed without prior notice.
Some software products marketed by SAP SE and its distributors
contain proprietary software components of other software vendors.
National product specifications may vary.
These materials are provided by SAP SE or an SAP affiliate company
for informational purposes only, without representation or warranty
of any kind, and SAP or its affiliated companies shall not be liable for
errors or omissions with respect to the materials. The only
warranties for SAP or SAP affiliate company products and services
are those that are set forth in the express warranty statements
accompanying such products and services, if any. Nothing herein
should be construed as constituting an additional warranty.
SAP and other SAP products and services mentioned herein as well
as their respective logos are trademarks or registered trademarks of
SAP SE (or an SAP affiliate company) in Germany and other
countries. All other product and service names mentioned are the
trademarks of their respective companies.
Please see https://www.sap.com/about/legal/trademark.html for
additional trademark information and notices.