Dashboards show CPU stable and memory normal, but users report failures increasing over 20 minutes. What signal is most reliable for detecting health degradation?

A dependency doesn’t crash; it returns slowly, intermittently wrong, or times out only sometimes. Monitoring shows partial degradation, not total failure. What contributes most to gray failures?

Which activity is typically part of the CI stage?

A key risk of poorly managed test environments is:

Which of the following best indicates poor code quality in a BFSI system?

During autoscaling and restarts, services take long because they store session state locally, requiring warm-up and recovery steps. Which design improves recovery speed?

A user-profile service calls a credit-score API. The credit-score API becomes slow and starts timing out. Your upstream threads begin piling up. What’s the best immediate safeguard?

An application goes completely down when one node crashes because there is no standby instance. This is primarily a failure in:

An “Order Submit” call times out; the client retries and creates two orders. You can’t depend on the client to stop retrying. What is the fix for this situation?

Which of the following best reflects mature engineering practices in the BFSI sector?