AIOps

 View Only
Expand all | Collapse all

The status of Topology Manager containers changes to CrashLoopBackOff.

  • 1.  The status of Topology Manager containers changes to CrashLoopBackOff.

    Posted Mon July 25, 2022 06:40 AM
    Hello Team!
    On Openshift V 4.8, I have a fresh installation of NOI v1.6.4 with user-provisioned infrastructure and dynamic storage.
    The pods listed below consistently change their status from "Running" to "CrashLoopBackOff" and crash.

    evtmanager-ibm-hdm-analytics-dev-inferenceservice-655c999clqbpt Running 0/1 23 134m.

    evtmanager-spark-master-5bf58784f6-h7cvm 19 134m 0/1 CrashLoopBackOff

    evtmanager-spark-slave-88c94f97b-dps72 Running 19 134m 0/1

    Running 20 134m evtmanager-spark-slave-88c94f97b-wrs5x 0/1

    evtmanager-topology-layout-5689d6d648-9m9k2 Running 0/1 27 133m

    evtmanager-topology-search-7f4589c856-t42z5 25 131m 0/1 CrashLoopBackOff


    Since I can't search for any CI in the topology viewer, I'm more interested in topology-search and topology-layout.

    I appreciate your help.

    Thanks

    ------------------------------
    Jeepshunters dotcom
    ------------------------------


  • 2.  RE: The status of Topology Manager containers changes to CrashLoopBackOff.

    Posted Tue July 26, 2022 02:31 AM
    Check these pods from openshift UI, you will find that the CPU is increasing then it reach the defined limit in pod configuration then the pod crash. you need to change pod configuration and increase the cpu limit defined.

    ------------------------------
    Mohamed Hassan
    Fingerprint Consultancy
    ------------------------------



  • 3.  RE: The status of Topology Manager containers changes to CrashLoopBackOff.

    Posted Tue July 26, 2022 03:38 AM

    You need to establish why those pods are in CLBO

    Use the OpenShift UI -- Workloads --- pods  ( change the project to your namespace )

    Check the Logs and Events tab of the pods (some pods contain multiple containers so you may need to switch log streams to see dependent processes)

    They may give an indication as to why the pods are not stable



    ------------------------------
    john postoyko
    IBM
    London
    ------------------------------