Salesforce-MuleSoft-Platform-Architect Practice Test
Salesforce Spring 25 Release 152 Questions
Several times a week, an API implementation shows several thousand requests per minute in an Anypoint Monitoring dashboard, Between these bursts, the
dashboard shows between two and five requests per minute. The API implementation is running on Anypoint Runtime Fabric with two non-clustered replicas, reserved vCPU 1.0
and vCPU Limit 2.0.
An API consumer has complained about slow response time, and the dashboard shows the 99 percentile is greater than 120 seconds at the time of the complaint. It also shows
greater than 90% CPU usage during these time periods.
In manual tests in the QA environment, the API consumer has consistently reproduced the slow response time and high CPU usage, and there were no other API requests at
this time. In a brainstorming session, the engineering team has created several proposals to reduce the response time for requests.
Which proposal should be pursued first?
A. Increase the vCPU resources of the API implementation
B. Modify the API client to split the problematic request into smaller, less-demanding requests
C. Increase the number of replicas of the API implementation
D. Throttle the APT client to reduce the number of requests per minute
A. Increase the vCPU resources of the API implementation
Explanation:
The issue is characterized by high CPU usage (over 90%) during bursts of traffic, leading to slow response times (>120 seconds for the 99th percentile). Since manual tests in QA reproduced the issue with no other traffic, the bottleneck is clearly CPU-bound. Increasing vCPU resources (from 1.0 to a higher limit) will directly address the CPU saturation issue. While other options (like throttling or splitting requests) may help, they are workarounds rather than solving the root cause. Scaling replicas (Option C) could help distribute load, but since the issue persists even in isolation, vertical scaling (Option A) is the most immediate and effective solution.