Challenge 17 · intermediate · Worker pool
Thread pool saturation
An API serves 600 requests a second with 200 worker threads. Every request does a little CPU work, then waits for one of 20 database connections for a 20 ms query. Clients give up after 2 seconds. At 10 seconds a campaign lands and traffic jumps to 1,500 a second, more than the system can serve.
From 15 to 90 seconds, answer at least 800 requests a second before clients give up, keep the p99 of those answers at 1 s or less, and don't crash.
Goal · measured 00:15–01:30
0 changes · par 2- Answered before clients give up ≥ 800 answered/snot measured yet
- p99 of those answers ≤ 1.0 snot measured yet
- Time up, not restarting after a crash ≥ 100.0%not measured yet
Service map
What changed
Turn on a fault or change a defence while the run plays. Its effect shows here: the 5 seconds before against 5 seconds after it settles.
This is a simulation. Its numbers come from a model of a typical system, not from measuring real servers. The same link and seed always give the same run. How the Chaos Lab works.