Q24 of 38 · Performance
What is spike testing and how does it differ from stress testing?
Short answer
Short answer: Spike testing applies a sudden, extreme burst of load for a short duration to see whether the system recovers gracefully. Stress testing increases load progressively until the system breaks. Spike tests the elasticity and recovery behaviour; stress tests find the breaking point.
Detail
In a stress test you ramp load gradually — say 10 VUs per minute — until error rate climbs, latency degrades, or the system crashes. The goal is finding the maximum sustainable throughput and the failure mode.
In a spike test you jump from baseline to 10× the normal load in seconds, hold briefly, then drop back. The goal is not finding the breaking point but observing: does the system auto-scale in time? Do queues fill and drain cleanly? Does latency return to baseline after the spike, or does it stay elevated (indicating a resource leak or connection pool exhaustion)?
Spike tests are particularly important for consumer applications that experience flash traffic — a social media mention, an email blast to a large list, or a TV advertisement — where load does not ramp gradually.