Sunteți pe pagina 1din 11

Prognosis Self Monitoring V2.

1
Description ................................................................................................................................................................2
Support .....................................................................................................................................................................3
Prognosis Health Check Central ..................................................................................................................................4
Critical WVLOG Errors (Last 7 Days)..................................................................................................................................... 4
Bytes in Send Queues (NETRTR) .......................................................................................................................................... 5
Low Hard Drive Space .......................................................................................................................................................... 5
Patch Review........................................................................................................................................................................ 5
Generate Irfax ...................................................................................................................................................................... 6
Irfax Generation window ................................................................................................................................................. 6
Get WVLOG File & Get IRFAX File “Download” options .................................................................................................. 6
Internode Connection Problems.......................................................................................................................................... 7
Databases, Thresholds and Analysts Health – new in V2 .................................................................................................... 7
Prognosis Health Check - Database Surveillance display – (New in V2) .........................................................................9
Databases of concern .......................................................................................................................................................... 9
All Databases ........................................................................................................................................................................ 9
Update in Client ................................................................................................................................................................... 9
AutoSummarization Failures.............................................................................................................................................. 10
Prognosis Health Check - Alert Surveillance display – (New in V2)..............................................................................11
Thresholds .......................................................................................................................................................................... 11
Analysts .............................................................................................................................................................................. 11
Update in Client ................................................................................................................................................................. 11

Project Owners: Adam Annen (US) & Gerald Chaston (AU)


Version 2.1: 12/16/2016

Page 1 of 11
Description
The Prognosis Self Monitoring solution is a Threshold with supporting displays/dashboard and is designed to help
Prognosis administrators to simplify managing their Prognosis environment and detect issues in Prognosis from a top
level managing node. Its features cover the following areas:

• common critical errors in Prognosis error logs 'wvlog' with suggested solutions;
• old patch levels, low disk space and connectivity with Prognosis servers;
• databases wrapping too short or overdue for updates or auto-summarization failing;
• thresholds and analysts that stopped running;
• easy generate, retrieve 'irfax' and 'wvlog' diagnostics for requesting Support;
• alerts for Health Check issues found.

Prognosis Self Monitoring V2.1 includes:

Health Check Central display


Database Surveillance display\dashboard
Alerts Surveillance display\dashboard
Threshold for the key critical health check items on these displays\dashboards

Some of the key criteria behind these solutions are:

Management node only settings - No configurations or changes to monitoring nodes


- This allows scaling to very large environments and prevents the solution becoming a burden
Concentrate on areas not already monitored by other thresholds/analysts
Only show problems on the displays/dashboards/threshold
Easily maintainable
IR Support to install and train the Prognosis Administrator on it.

Platforms: UC monitored environments, Windows Infrastructure (Only 10.3 and above)

Continue to next page for ‘Support Topic’

Page 2 of 11
Support
This Solution is currently provided as a non-GA customization. It is still supported by Product Support per region.
Support will help install this Solution on one top level Node. Updates to the Solution will be emailed as they become
available.

For problems found by this Solution, the recommendation is to first review the indicated Knowledge Base (KB) Article via
the Support Portal and follow the recommendations to rectify each problem. If this does not resolve the problem, then
please log a Support Case that includes:

1. Screenshot of the Prognosis Health Check Central Display showing the problem detected.
2. IRFAX and WVLOG from the affected system/s.
3. Information and screenshots indicating the results of the recommended KB Article.

Support Portal https://support.ir.com:

• Log Support cases


• Search Knowledge Base Articles
• Download all Patches

Online Portal: https://online.prognosis.com

• Download Prognosis and latest Patches


• Online Help Link
• Training Resources
• Community Forum

Continue to next page for ‘Health Check Central’

Page 3 of 11
Prognosis Health Check Central
The Prognosis Health Check Central Display has 7 main sections. The latest solution also has 2 extra child pages called
Database Surveillance & Alert Surveillance with links to these at the top of the display. The main Health Check is still this
first display, but the supporting displays do offer tools to manage and monitor Prognosis key areas.

Critical WVLOG Errors (Last 7 Days)


This section detects some very specific wvlog errors that have been known to cause data loss or missed Alerts. Future
Alerts will be added as they are added to the product.

The Knowledge Base column on this Display has a reference number that can be used to search the Support Portal
(https://support.ir.com) Solutions tab. It will give a detailed explanation of what the error means and suggested actions
to address the error.

Note: Fixing a problem will not remove the entry from ‘Critical WVLOG errors’ window. After fixing the problem the
error stops logging new entries in these windows. All other windows on this Display update live. The errors will wrap off
after 7 days.

Page 4 of 11
Bytes in Send Queues (NETRTR)

This window monitors the PCONNECT record for the NETRTR process on each server and specifically concentrates on the
‘SEND Queues’ figure. While an occasional blip here is okay, it continually showing high values should be of concern. The
main root cause for this having a problem is a poor performing network connection between the servers. If as above you
see non-consecutive intervals below 100000, then this indicates the queue is clearing. If they were consecutive and
much higher values, then this would be an area of concern.

Low Hard Drive Space


This is key with the increase use of VMWare for Prognosis servers. An increasingly common issue is servers running out
of hard drive space. The Alert and Display show hard drives from any of the Prognosis servers having 10% or less free
space. Adjust the 10% value as needed. Prognosis running out of hard drive space can cause corruption of the Databases
and Configurations.

Patch Review
This section has two components:
1. It detects Prognosis servers where the last patch install was 90 days or more. This may indicate either a system
with less than the current Patch, or a system that may require an upgrade.
2. A Deploy Patch wizard designed to be used with the .gz patches available on the Support Portal. It provides the
ability to install the latest patch from the Managing Node to all lower nodes. It even completes the required
Prognosis Service restart.

Page 5 of 11
• Download the .gz version of the Patch to install.
• Save the file in the \Prognosis\Server\Configuration directory of the server where Prognosis Self-Monitoring
is installed.
• Click the “Deploy Patch” option and complete the fields.
Note: Most patches require a Prognosis Service restart so always best to select Y in the last box as per above
screenshot

Generate Irfax
This section is one that has no Alerts associated with it. It is included as a useful option to generate irfaxes from the
Monitoring Nodes and then download them to the Managing Node where Prognosis Self-Monitoring is installed. There is
also an option to download the wvlog.txt file from those remote servers.

It is important to provide both of these files for any Support Cases raised. Being able to do all this from the Managing
Node makes the process simpler and easier.

Irfax Generation window

1. If the credentials used to login to the Managing Node are valid for lower Nodes, just click yes, otherwise click
advanced and enter different credentials.
2. Wait for the popup command window to fully complete before downloading the irfax file with the 2nd link in that
window (The output file will be in the ‘Prognosis\Server\Configuration’ directory with the naming
‘IRFAX_<Machinename>.txt’

Get WVLOG File & Get IRFAX File “Download” options

Page 6 of 11
• This first dialog box has only 1 input field and is asking for the destination Node for the download of the
Irfax\wvlog. The field can be any Prognosis Node that is visible from the Managing Node with the Solution.
Usually you would put the nodename of the Prognosis Self Monitoring server (Be sure to put the backslash at
the start.
• The file will be saved in the \Prognosis\Server\Configuration directory with the naming format
IRFAX_<nodename>.txt or WVLOG_<nodename>.txt. If a file downloaded again from the same source Node, the
new file will overwrite the older version.
• Be sure to wait for the irfax to complete before clicking on the “Download Here” link.
• Same behavior occurs with the wvlog and the filename is WVLOG-<nodename>.txt. WVLOG download does not
rely on irfax generation.

Internode Connection Problems


This Display detects problems with Prognosis Node to Prognosis Node communications.

Internode Connection Problems caused by removing a MANAGING NODE or MANAGED NODE statement from the
NETWORK Configuration of a Prognosis server, or via true network issues between two nodes. The Managing Direction
and Arrow give a good indication of which node contains the MANAGING/MANAGED entry.

Clicking the drilldown in the ‘Affected Nodes’ field will open a more detailed Display with even more drilldown options.

Databases, Thresholds and Analysts Health – new in V2


This pane lists databases with abnormal Statuses like 'not active', '100% full', 'overdue for update'. Clicking on a Status
drills down to the Prognosis Health Check - Database Surveillance display, see below for further details about it.

This pane also lists Thresholds and Analysts that have stopped running. Click to drill down to the Prognosis Health Check
- Thresholds and Analysts Surveillance display, see below for further details of that display.

Page 7 of 11
Continue to next page for ‘Database Surveillance’

Page 8 of 11
Prognosis Health Check - Database Surveillance display – (New in V2)

This new ‘Database Surveillance’ solution lets you create an input file to control which databases to alert on. Not only
will it alert on them being stopped, it also provides growth metrics for each database. This allows you to check to see if
your databases are sized correctly. The default sizing of databases as provided with the product is not ideal for all
environments.

The driving input file for this solution is: C:\Prognosis\Server\Configuration\PSM\dbs.txt

The format for each line is: \NodeName | DatabaseName;


Note: Values after the ; character are ignored

There is a useful ‘Update in Client’ section on this display that will help you generate the text for what’s currently
running.

Databases of concern
This pane shows a database that is 100% full, some databases overdue to receive data, and a database not running.

All Databases
This pane showing all databases being monitored – it has been scrolled over to the right to show the metrics that are
being monitored – Exists, Type, Active (running ), %Full (for identifying if wrapping too short making gaps in client and
web UI reports), Last Updated date, Interval of database collection, Size, Max size, wrap and Expected Full date.

Update in Client
As mentioned above, this feature generates a command window for the selected node and gives you text you can
copy/paste into the “C:\Prognosis\Server\Configuration\PSM\dbs.txt” file. It is just an extra tool to get you the text for

Page 9 of 11
whats running. You should audit your environment and ensure all key databases are running before this step.

Note: The default databases in prognosis called WebSearch....... are special collection databases that get stopped and
started with the web search functions. You should exclude these from being monitored by this solution

I.e. WebSearchAIX, WebSearchAvailability, WebSearchSolaris, WebSearchUnix, WebSearchVmWare etc

AutoSummarization Failures
Many monitoring solutions rely on the Autosummarization mechanism to schedule regular database summaries. An
Alert for failure of the Autosummarization mechanism is not part of any other Prognosis threshold so was an important
inclusion in this self monitoring solution. This window shows the total amount of summary failures in the last 7 days for
each node. From this window, there are two useful drilldown links to the Autosummarization Central and Log areas for
the affected node.

Note: Fixing a failing summary will not clear the figure on this display immediately as it looks at the last 7 days of
failures. The threshold condition only looks at the last 1 day, so once you correct the problems, the alerts should stop
within 24 hours, and the display should reduce in the next 7 days.

Continue to next page for ‘Alert Surveillance’

Page 10 of 11
Prognosis Health Check - Alert Surveillance display – (New in V2)
Alert Surveillance works much like the Database Surveillance product in that you configure an input
file to monitor nodes for running alerts. To ensure alerts do not get stopped.

Thresholds
This window shows the currently monitored thresholds as per C:\Prognosis\Server\Configuration\PSM\Thresholds.txt. It
shows either a N or Y beside it.
Input File format: \NodeName | ThresholdName;

Analysts
This window shows the currently monitored analysts as per C:\Prognosis\Server\Configuration\PSM\Analysts.txt. It
shows either a N or Y beside it.
Input File format: <tab>\NODENAME<tab>AnalystName

Note: This input file is different to the others

Update in Client
There are links here for both thresholds and analysts (Separate input files). Note that the input file for analysts is
different <tab>\NODENAME<tab>AnalystName

Page 11 of 11

S-ar putea să vă placă și