hardScenarioConceptMidSeniorStaff
A load test shows p99 latency climbing sharply past 500 concurrent users. How do you work out why?
6 min readUpdated 2026-08-15
Answer
Short answer
Read the throughput curve first: if throughput has flattened while latency rises, you are queueing behind a saturated resource, and the job is to find which one by walking utilisation and saturation from the load generator through the app tier to the database.
Was this answer helpful?
Likely follow-ups
- Throughput is still rising linearly at 500 users but p99 has doubled. What does that tell you?
- How would you tell a garbage-collection pause apart from lock contention in the same latency profile?
- The database CPU is at 40 percent and p99 is 4 seconds. Where do you look next?
- How do you prove a fix worked rather than moved the bottleneck?
Related questions
- Your p99 latency jumped tenfold after a deploy while p50 is unchanged. Diagnose it out loud.hardAlso on percentiles and queueing6 min
- How do you model a realistic workload for a performance test?hardAlso on percentiles4 min
- The load test was green and the launch still fell over. What did the test not model?hardAlso on performance-analysis5 min
- Traffic is a hundred times normal and some of it is real customers. What do you drop first?hardAlso on capacity5 min