Bug 1913275 - [4.5] Unable to schedule a pod due to Insufficient ephemeral-storage
Summary: [4.5] Unable to schedule a pod due to Insufficient ephemeral-storage
Keywords:
Status: CLOSED EOL
Alias: None
Product: OpenShift Container Platform
Classification: Red Hat
Component: kube-scheduler
Version: 4.5
Hardware: Unspecified
OS: Unspecified
medium
medium
Target Milestone: ---
: 4.5.z
Assignee: Jan Chaloupka
QA Contact: RamaKasturi
URL:
Whiteboard:
Depends On: 1913263
Blocks:
TreeView+ depends on / blocked
 
Reported: 2021-01-06 12:13 UTC by Jan Chaloupka
Modified: 2021-01-18 14:13 UTC (History)
8 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of: 1913263
Environment:
Last Closed: 2021-01-18 14:13:47 UTC
Target Upstream Version:
Embargoed:


Attachments (Terms of Use)

Comment 1 Jan Chaloupka 2021-01-06 12:19:13 UTC
Origin PR https://github.com/openshift/origin/pull/25781

Comment 2 RamaKasturi 2021-01-12 12:58:41 UTC
Verified bug in the payload below and i see that it works.

[knarra@knarra openshift-client-linux-4.7.0-fc.2]$ oc get clusterversion
NAME      VERSION                                           AVAILABLE   PROGRESSING   SINCE   STATUS
version   4.5.0-0.ci.test-2021-01-12-111301-ci-ln-5zmmb2t   True        False         74m     Cluster version is 4.5.0-0.ci.test-2021-01-12-111301-ci-ln-5zmmb2t

Below are the steps followed to verify the bug:
=================================================
1) Launch cluster bot with the PR
2) Edit file cat /etc/kubernetes/manifests/kube-controller-manager-pod.yaml, add LocalStorageCapacityIsolation=false to --feature-gates line on all master nodes one by one, wait for the kube-scheduler pods to restart.
3) Now create a pod using the yaml file below.

[knarra@knarra openshift-client-linux-4.7.0-fc.2]$ cat /tmp/ephermal.yaml 
apiVersion: v1
kind: Pod
metadata:
  name: nginx
  labels:
    name: nginx
spec:
  containers:
  - name: nginx
    image: quay.io/openshifttest/nginx@sha256:3936fb3946790d711a68c58be93628e43cbca72439079e16d154b5db216b58da
    schedulerName: default-scheduler
    ports:
    - containerPort: 80
    resources:
      requests:
        ephemeral-storage: 4096M
      limits:
        ephemeral-storage: 4096M
  initContainers:
  - name: init-myservice
    image: quay.io/openshifttest/busybox@sha256:afe605d272837ce1732f390966166c2afff5391208ddd57de10942748694049d
    command: ['sh', '-c', "echo waiting for myservice; sleep 7;"]
    resources:
      requests:
        cpu: 500m
        ephemeral-storage: 2M
        memory: 1024M

4) Verify that pod is in running state and not in pending.

[knarra@knarra openshift-client-linux-4.7.0-fc.2]$ ./oc get pods -o wide
NAME    READY   STATUS    RESTARTS   AGE   IP            NODE                                       NOMINATED NODE   READINESS GATES
nginx   1/1     Running   0          34s   10.128.2.14   ci-ln-5zmmb2t-f76d1-9lspz-worker-b-p59dn   <none>           <none>

Adding /lgtm to the PR as the bug works well.

Comment 3 Jan Chaloupka 2021-01-15 11:07:25 UTC
PR ready to be merged, just waiting for the right labels.


Note You need to log in before you can comment on or make changes to this bug.