Question 1 ❖ A customer has licensed 4 production cores and has 15 APIs to be deployed during phase-1 of the project. They are using CloudHub and need high availability for all their APIs. They also need to enforce rate limiting and client-id requirements on the API invocations. Assuming minimal sized vCore for deploying each API, what is the simplest deployment architecture for API policies? ➢ Policies on API proxies ➢ Policies on API implementations ➢ A combination of API proxies and API implementations Question 2 ❖ A customer has deployed API implementations and proxies with policies on a runtime fabric running in their local data center. 80% of the APIs are accessed by onprem components and 20% of the APIs are accessed by external components. Due to an internet outage, their data center gets disconnected from the internet. How many APIs will be accessible during the internet outage. ➢ 20% ➢ 80% ➢ All ➢ None Question 3 ❖ IP-based whiltelisting policies can be applied to Experience APIs ➢ Always ➢ Never ➢ In some scenarios (explain) Question 4 ❖ A Process API performs complex orchestration across multiple System and Process APIs and returns a large JSON payload. The processing does not change the underlying state of any resource. Multiple Experience APIs may call the Process API with same parameter values. What is the most performant way to implement this requirement? ➢ Caching proxy on the Process API ➢ Caching proxy on the Experience API ➢ Client-side caching Question 5 ❖ An organization does not have a CDM or Bounded Context Data Model. What is the advantage of having System API mimic the backend system? ➢ If the System API is not created, API-led connectivity cannot be achieved. ➢ System API simplifies the access to the backend system. It also isolates API clients from underlying protocols and authentication. ➢ System API provides an easy way to replace the backend system in the future without changing the System API ➢ Reusability cannot be achieved without creating the System API Question 6 ❖ An organization uses CloudHub for their API deployments. These APIs connect to onpremise backend systems and they have separate production and non-production systems. They would like their production APIs to be able to access only production backend systems and non-production APIs to be able to access only non-production backend systems. What is an appropriate solution for this? ➢ Create production and non-production APIs in different CloudHub Business groups ➢ Configure appropriate firewall rules so that only IP addresses from corresponding environments are allowed ➢ Create separate production and non-production VPCs in the CloudHub and setup connections to onpremise environments appropriately. Question 7 ❖ A native iOS mobile application directly invokes a long-running Experience API. What is the best way for the iOS mobile app to get the results from the Experience API without blocking the app? ➢ Asynch API invocation with callback ➢ Asynch API invocation with polling ➢ Event driven message processing ➢ iOS notification Question 8 ❖ An API is deployed to CloudHub and uses Object Store for storing some common org wide information. Another API needs access to the same data and knows the key. What is the best way for the second API to get this data? ➢ Directly retrieve from Object Store using the Object Store connector ➢ Use Object Store API ➢ Second API cannot retrieve the data from Object Store Question 9 ❖ An API occasionally fails due to database row locking. What are the appropriate strategies to make this API invocation fault tolerant? ➢ Timeout & Retry ➢ Retry & Circuit Breaker ➢ Retry & Fallback Invocation ➢ Opportunistic Parallel Invocations Question 10 ❖ Which API data modeling technique enables replacing of the underlying core system with minimal changes to the System APIs encapsulating the core system being replaced? ➢ Canonical data model ➢ Bounded context data model ➢ API specific data model ➢ Data model mirroring the core system Question 11 ❖ In which layer are the APIs most reusable? ➢ Experience APIs ➢ Process APIs ➢ System APIs ➢ All layers Question 12 ❖ What types of APIs are most reusable? ➢ Coarse-grained APIs ➢ Fine-grained APIs ➢ SOAP APIs ➢ REST APIs Question 13 ❖ Which of these API versioning schemes use the least resources assuming clients will be accessing multiple API versions? ➢ API version in the URL path ➢ API version in the hostname ➢ API version in the hostname and URL path Question 14 ❖ Which type of testing focuses on the API specifications? ➢ Unit testing ➢ Integration testing ➢ Resilience testing ➢ White box testing Question 15 ❖ Anypoint Visualizer is useful for all of the following except: ➢ Discover application architecture ➢ Event publishers and consumers ➢ Impact analysis of proposed changes ➢ Architecture governance - identify API-led connectivity violations Question 16 ❖ Alerts can be implemented for: ➢ API invocations ➢ API implementations ➢ Both API invocations and implementations ➢ Neither API invocations or implementations Question 17 ❖ API Policies cannot be applied to ➢ SOAP APIs running outside Mule Runtime ➢ REST APIs running outside Mule Runtime ➢ HTTP/1.x APIs running outside Mule Runtime ➢ HTTP/2 & WebSocket APIs Question 18 ❖ An organization has deployed multiple versions of an API and all versions are accessible to clients. A user is unable to request access to a version of the API for a new client from Exchange. What could be the reason for this? ➢ API version has not yet been published ➢ API version has been deleted ➢ API version has been deprecated ➢ User does not have access to the API version Question 19 ❖ An organization provides a real-time stock quotes from BSE via APIs and has directly interfaced with BSE systems for this. It has different classes of clients accessing the APIs and has a need to provide different levels of service to clients based on whether they are paying or free-use clients and for the paying clients, how much they pay for the service. What minimal policy can be applied for these APIs? ➢ SLA-based rate limiting policy ➢ Rate limiting policy ➢ Spike control policy ➢ Spike control and SLA-based rate limiting policy Question 20 ❖ Under what conditions is it beneficial to deploy to 1 x 2 vCore instead of 2 x 1 vCore workers? ➢ Improve cache hit ➢ Improve high availability ➢ Improve response time ➢ Improve TPS (transactions per second)