K
Q

Question

should allow starting 30 pods per node

7/14/2018

Running e2e-test on the local cluster kubernetes, with command:

go run hack/e2e.go -- --provider=local --test --check-version-skew=false --test_args="--host=https://192.168.1.5:6443 --ginkgo.focus=\[Feature:Performance\]"

Showing the errors:

[Feature:Performance] should allow starting 30 pods per node using { ReplicationController} with 0 secrets, 0 configmaps and 0 daemons [BeforeEach]
• Failure in Spec Setup (BeforeEach) [6.331 seconds]
[sig-scalability] Density
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/test/e2e/scalability/framework.go:22
  [Feature:Performance] should allow starting 30 pods per node using { ReplicationController} with 0 secrets, 0 configmaps and 0 daemons [BeforeEach]
  /go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/test/e2e/scalability/density.go:554

  Expected error:
      <*errors.errorString | 0xc421733010>: {
          s: "Namespace e2e-tests-containers-ssgmn is active",
      }
      Namespace e2e-tests-containers-ssgmn is active
  not to have occurred

  /go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/test/e2e/scalability/density.go:466
------------------------------
SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSJul 14 00:02:24.065: INFO: Running AfterSuite actions on all node
Jul 14 00:02:24.065: INFO: Running AfterSuite actions on node 1


Summarizing 2 Failures:

[Fail] [sig-scalability] Load capacity [BeforeEach] [Feature:Performance] should be able to handle 30 pods per node { ReplicationController} with 0 secrets, 0 configmaps and 0 daemons
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/test/e2e/scalability/load.go:156

[Fail] [sig-scalability] Density [BeforeEach] [Feature:Performance] should allow starting 30 pods per node using { ReplicationController} with 0 secrets, 0 configmaps and 0 daemons
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/test/e2e/scalability/density.go:466

Ran 2 of 998 Specs in 12.682 seconds
FAIL! -- 0 Passed | 2 Failed | 0 Pending | 996 Skipped --- FAIL: TestE2E (12.71s)

Seemingly, the local cluster Kubernetes has limitation of pods per node. How to fix this? The local cluster configuration is:

leeivan@master01:~/gowork/src/k8s.io/kubernetes$ kubectl get nodes
NAME       STATUS    ROLES     AGE       VERSION
master01   Ready     master    10d       v1.11.0
node01     Ready     <none>    10d       v1.11.0
node02     Ready     <none>    10d       v1.11.0
node03     Ready     <none>    10d       v1.11.0
node04     Ready     <none>    10d       v1.11.0
node05     Ready     <none>    10d       v1.11.0
leeivan@master01:~/gowork/src/k8s.io/kubernetes$ kubectl version
Client Version: version.Info{Major:"1", Minor:"11", GitVersion:"v1.11.0", GitCommit:"91e7b4fd31fcd3d5f436da26c980becec37ceefe", GitTreeState:"clean", BuildDate:"2018-06-27T20:17:28Z", GoVersion:"go1.10.2", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"11", GitVersion:"v1.11.0", GitCommit:"91e7b4fd31fcd3d5f436da26c980becec37ceefe", GitTreeState:"clean", BuildDate:"2018-06-27T20:08:34Z", GoVersion:"go1.10.2", Compiler:"gc", Platform:"linux/amd64"}

-- Ivan Lee

kubernetes

1 Answer

7/16/2018

According to kubelet documentation:

--max-pods int32
Number of Pods that can run on this Kubelet. (default 110)

So, 110 should be enough to pass the tests. But it is possible that test measures the real capacity of your nodes in terms of Allocatable.CPU and Allocatable.Memory

Also, before the test run, all the namespaces should be deleted:

// Terminating a namespace (deleting the remaining objects from it - which
// generally means events) can affect the current run. Thus we wait for all
// terminating namespace to be finally deleted before starting this test.

Looks like one of your namespaces still was active, so the test failed.

-- VAS

Source: StackOverflow

KQ

should allow starting 30 pods per node

Similar Questions

1 Answer

K
Q