[infinispan-issues] [JBoss JIRA] (ISPN-8424) Cluster does not recover after killing a pod on OpenShift

[Galder Zamarreño (JIRA)]

    [ https://issues.jboss.org/browse/ISPN-8424?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13480035#comment-13480035 ] 

Galder Zamarreño commented on ISPN-8424:
----------------------------------------

I've recreated the issue with TRACE on org.infinispan:

https://www.dropbox.com/s/2lz45t3p1tml3oy/ISPN-8424_1020_1.zip?dl=0

It seems to appear only when the remote continuous query is happening. If only injecting data, it seems to work fine.

> Cluster does not recover after killing a pod on OpenShift
> ---------------------------------------------------------
>
>                 Key: ISPN-8424
>                 URL: https://issues.jboss.org/browse/ISPN-8424
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Core, JMX, reporting and management
>    Affects Versions: 9.1.0.Final
>            Reporter: Galder Zamarreño
>         Attachments: datagrid-3-kg593.log, datagrid-3-tzh6h.log
>
>
> To replicate issue, follow instructions in [here|https://github.com/infinispan-demos/streaming-data-kubernetes/blob/master/live-coding/basel-one-17.md] without changing the code.
> Once you have injector working and you connect from the dashboard, do the following:
> {code}
> $ docker ps
> ...
> // find id for one of the infinispan server nodes
> ...
> $ docker rm -f <container_id>
> {code}
> The container will be killed but the other nodes do not manage to recover the cluster and install the new formation. It seems to enter an endless loop... See attached logs.
> I suspect the cluster stats requests that appear in the logs come from the management console.
> I don't yet have TRACE logs but I'll try to replicate with TRACE later.



--
This message was sent by Atlassian JIRA
(v7.5.0#75005)