Matt Matejczyk

Re-implement scheduling-throughput measurement

44

Currently, we have a very crude logic for computing [scheduling-throughput](https://github.com/kubernetes/perf-tests/blob/master/clusterloader2/pkg/measurement/common/scheduling_throughput.go). We list all pods every 5s and see how many new pod very scheduled in that time. It doesn't work...

We should have a high-density test running on regular GCE

30

Currently we have only one running on kubemark - https://github.com/kubernetes/test-infra/blob/master/config/jobs/kubernetes/sig-scalability/sig-scalability-periodic-jobs.yaml#L276

Create api-availability measurement

38

### Justification See this comment - https://github.com/kubernetes/perf-tests/issues/1086#issuecomment-593918574 ### Milestones #### V0 1. Create a new "ApiAvailability" measurement 1. This measurement should periodically probe the apiserver's /healthz endpoint and record whether...

NodeKiller seems to be not working in 100 node 1.17 / master performance tests

19

Original debugging done by @jkaniuk: > In 100 nodes OSS performance tests of 1.16: > https://k8s-testgrid.appspot.com/sig-scalability-gce#gce-cos-1.16-scalability-100 > NodeKiller is consistently failing: > https://prow.k8s.io/view/gcs/kubernetes-jenkins/logs/ci-kubernetes-e2e-gci-gce-scalability-stable1/1219228169567997957 > ``` > W0120 12:25:21.234] I0120 12:25:21.234558...

Make PV tests work on kubemark clusters

11

Enabling PVs in the load test failed the kubemark presubmit because scheduler failed to schedule pods with PVs ``` Unable to schedule test-y6u59n-1/small-statefulset-0-0: no fit: 0/101 nodes are available: 1...

Scalability load test extended to exercise Deployments, DaemonSets, StatefulSets, Jobs, PersistentVolumes, Secrets, ConfigMaps, NetworkPolicies

8

**Implemented** - [x] Deployments - [x] DaemonSets - [x] StatefulSets - - [x] Jobs - [x] Persistent Volumes - [x] Secrets - [x] ConfigMaps - [ ] NetworkPolicies - WIP...