Bug 1658588

I have couple of questions here

- In 3.11, can you provide descheduler-cronjob description? - Our docs explicitly tell to run descheduler as job https://docs.openshift.com/container-platform/3.11/admin_guide/scheduling/descheduler.html (Unrelated question)
- When ran for 3.11, was the log-level increased?(You can find that from description again)

(In reply to ravig from comment #1)
> I have couple of questions here
> 
> - In 3.11, can you provide descheduler-cronjob description? - Our docs
> explicitly tell to run descheduler as job
> https://docs.openshift.com/container-platform/3.11/admin_guide/scheduling/
> descheduler.html (Unrelated question)

Yeah, in 3.11 the cronjob generate job to run descheduler at fixed period. We create cronjob by pass parameters to ansible job as following
openshift_ansible_vars:
  openshift_descheduler_install: true
  openshift_descheduler_strategies_dict: "{'remove_duplicates': True, 'remove_pods_violating_inter_pod_anti_affinity': True, 'low_node_utilization': True}"
  openshift_descheduler_cronjob_node_selector: ""
  openshift_descheduler_cronjob_schedule: "*/2 * * * *"

descheduler-cronjob description:
[root@qe-minmli-gcemaster-etcd-1 ~]# oc get cronjob descheduler-cronjob -o yaml 
apiVersion: batch/v1beta1
kind: CronJob
metadata:
  creationTimestamp: 2018-12-18T02:18:33Z
  name: descheduler-cronjob
  namespace: openshift-descheduler
  resourceVersion: "46315"
  selfLink: /apis/batch/v1beta1/namespaces/openshift-descheduler/cronjobs/descheduler-cronjob
  uid: 380feecd-026b-11e9-84f0-42010af00003
spec:
  concurrencyPolicy: Allow
  failedJobsHistoryLimit: 1
  jobTemplate:
    metadata:
      creationTimestamp: null
    spec:
      template:
        metadata:
          annotations:
            scheduler.alpha.kubernetes.io/critical-pod: ""
          creationTimestamp: null
          name: descheduler-cronjob
        spec:
          containers:
          - args:
            - --policy-config-file=/policy-dir/policy.yaml
            - --v=5
            - --dry-run
            command:
            - /bin/descheduler
            image: registry.reg-aws.openshift.com:443/openshift3/ose-descheduler:v3.11
            imagePullPolicy: IfNotPresent
            name: descheduler
            resources: {}
            terminationMessagePath: /dev/termination-log
            terminationMessagePolicy: File
            volumeMounts:
            - mountPath: /policy-dir
              name: policy-volume
          dnsPolicy: ClusterFirst
          priorityClassName: system-cluster-critical
          restartPolicy: Never
          schedulerName: default-scheduler
          securityContext: {}
          serviceAccount: descheduler-sa
          serviceAccountName: descheduler-sa
          terminationGracePeriodSeconds: 30
          volumes:
          - configMap:
              defaultMode: 420
              name: descheduler-policy-configmap
            name: policy-volume
  schedule: '*/2 * * * *'
  successfulJobsHistoryLimit: 3
  suspend: false
status:
  lastScheduleTime: 2018-12-18T07:02:00Z

configmap description:
[root@qe-minmli-gcemaster-etcd-1 ~]# oc get cm descheduler-policy-configmap -o yaml 
apiVersion: v1
data:
  policy.yaml: |
    apiVersion: "descheduler/v1alpha1"
    kind: "DeschedulerPolicy"
    strategies:
      "RemoveDuplicates":
         enabled: true
      "RemovePodsViolatingInterPodAntiAffinity":
         enabled: true
      "LowNodeUtilization":
         enabled: true
         params:
           nodeResourceUtilizationThresholds:
             thresholds:
               "cpu" : 40
               "memory": 40
               "pods": 40
             targetThresholds:
               "cpu" : 70
               "memory": 70
               "pods": 70
kind: ConfigMap
metadata:
  creationTimestamp: 2018-12-18T02:18:30Z
  name: descheduler-policy-configmap
  namespace: openshift-descheduler
  resourceVersion: "4390"
  selfLink: /api/v1/namespaces/openshift-descheduler/configmaps/descheduler-policy-configmap
  uid: 36175f3f-026b-11e9-84f0-42010af00003



> - When ran for 3.11, was the log-level increased?(You can find that from
> description again)
NO, DEBUG_LOGLEVEL=5 (/etc/origin/master/master.env )

Version-Release number of selected component:
oc v3.11.57
kubernetes v1.11.0+d4cacc0
features: Basic-Auth GSSAPI Kerberos SPNEGO

Server https://qe-minmli-gcemaster-etcd-1:8443
openshift v3.11.57
kubernetes v1.11.0+d4cacc0

Thanks for the output MinLi. Actually, the following section of job config is causing the increased logging:

 - args:
     - --policy-config-file=/policy-dir/policy.yaml
     - --v=5

While I don't mind increasing the logging level, I have an inflight PR https://github.com/openshift/descheduler-operator/pull/36, through which you can pass --v(or log-level) option for descheduler binary from the CR.

So, either way, we need to update docs to show that we are creating a cron job ibstead of a job. Once the PR merges, we should be able to add a flag which shows the log values.

https://github.com/openshift/descheduler-operator/pull/38

This PR should have fixed the issue.

verified!

Version info:
oc v4.0.0-alpha.0+85a0623-808
kubernetes v1.11.0+85a0623
features: Basic-Auth GSSAPI Kerberos SPNEGO

Server https://wsun-qe-api.origin-ci-int-aws.dev.rhcloud.com:6443
kubernetes v1.11.0+85a0623

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2019:0758