Check: observability | ComputeSphere Learn

Goal: Confirm you can pick the right signal for a question and find it on a service.

Four questions: three on Module 1, one on Path 5, Module 7. Every answer explains itself, right or wrong.

Users say checkout was slow at about 10:00. Your app is one service. What's the best order to investigate?Metrics to confirm when, then logs from that momentSet up tracing and wait for it to happen againThe deploy log

Which of these is saturation?p95 response time of 800 msMemory at 480 MB on a 512 MB Flex spherelet3% of requests answered with 500

You run a background worker. Where do you watch its error rate?The Traffic tab's Status breakdownIts runtime logsThe Metrics tab

From Path 5: your app needs about three minutes to start, and every deploy restarts it before it's ready. What fixes it?Raise Failure thresholdRaise Initial delay above the start-up timeTurn the health check off

Retry as often as you like.