Kubernetes lab
HPA autoscaling
Throw traffic at a Horizontal Pod Autoscaler and watch replicas, CPU and dropped requests over time.
Each Pod serves 100 req/s at 100% CPU, so measured CPU can't show how far over capacity the service is: the HPA needs several rounds to catch a spike. Try the flapping and node-limit scenarios too.
Learn the concept
The lesson this lab comes from, with the explanation, a practice task and a quiz.
More labs
All labs- InfrastructureLoad balancerCompare balancing algorithms, health checks and consistent hashing under simulated load.
- System designRate limiterRun one stream of requests through fixed window, sliding log, sliding counter and token bucket limiters side by side.
- System designQuorum reads and writesPick N, W and R, take replicas down, and see exactly when a read misses the latest write.