[jira] [Commented] (STREAMPIPES-358) pod restart freezes a pipeline (k8s)

["Dominik Riemer (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/STREAMPIPES-358?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17351309#comment-17351309 ] 

Dominik Riemer commented on STREAMPIPES-358:
--------------------------------------------

Hi Kiryl,

thanks for reporting this issue!

Although a full-fledge resolution to this bug is blocked by the implementation of STREAMPIPES-319 (improved service discovery), there is now a health check which checks the health of each pipeline element every 60 seconds and automatically re-invokes the pipeline element in case it is not available or was restarted.

There is now also an indicator in the pipeline view that shows the status of the pipeline. In case of failures, you can also force the stop of a pipeline, which will stop all elements that are available.

You can try the resolution using the Docker-Compose files from the installer's dev branch (I'd recommend to delete all volumes of the previous version before). Please tell me if the fix works or if you have any further problems! 

> pod restart freezes a pipeline (k8s)
> ------------------------------------
>
>                 Key: STREAMPIPES-358
>                 URL: https://issues.apache.org/jira/browse/STREAMPIPES-358
>             Project: StreamPipes
>          Issue Type: Bug
>          Components: Backend
>            Reporter: Kiryl Batsiukov
>            Assignee: Dominik Riemer
>            Priority: Major
>             Fix For: 0.68.0
>
>
> On a k8s deployment do the following:
>  
> 1) create and start a pipeline with couple of elements
> 2) restart a pod which hosts any of the active pipeline elements
> result: pipeline freezes (no new messages). pipeline cannot be stopped or restarted.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)