Release the locusts - distributed performance tests
summary
Shows how a Locust-based distributed harness generates up to 700 Gbps of HTTP(S) traffic using Locust master/workers with ZeroMQ, Java workers, and orchestration via Ansible orchestration. Metrics flow into InfluxDB+Grafana with Jolokia/Telegraf per-worker telemetry. Covers concrete scaling rules (spawn rate ≥ workers, evenly divisible users), JVM and queue bottlenecks, task-ordering and bucketing artifacts, and workarounds that let large multi-node runs succeed albeit requiring careful tuning.
AI-written summary · about
chapters (5)
- 0:00Start
- 0:32performance testing is in general a testing practice performed to determine how a system performs in terms of responsiveness and stability under a particular workload....
- 0:49It can also serve to investigate, measure, validate or verify other quality attributes of the system, such as scalability, reliability and resource usage.
- 2:02complex scenarios several ways to interact with tested system configure nearly everything in HTTP message generated clients
- 16:09In this class you define a tick method that returns a tuple with the desired user count and spawn rate for None to stop the test . Locust will call the tick() method approximately once per second.
related talks
21mUnleashing the Eighth Plague: An Intro to Load Testing with Locust and Python
35mTesting Web Pages and Mobile Applications With Locust.io
49mLoad Testing with Python and Locust
21mCrunching Millions of Requests for Breakfast: A Tale of Performance Testing Prowess
23mDistributed performance testing at the push of a button
30mAbove the Clouds: Investigating Systems from 32,000 ft.
22mBuilding a Flexible, Scalable Infrastructure to Test Dolby.IO at Scale
40mPerformance Testing and Big Data