[
https://issues.apache.org/jira/browse/STREAMPIPES-264?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17245887#comment-17245887
]
Patrick Wiener commented on STREAMPIPES-264:
--------------------------------------------
I can reproduce your problem on my laptop (Docker for Mac, k8s). I guess the
problem occurs because we do not explicitly manage the persistent volumes
claims with a suitable storage driver. There are quite some options in k8s (see
types of persistent volumes:
[https://kubernetes.io/docs/concepts/storage/persistent-volumes/|https://kubernetes.io/docs/concepts/storage/persistent-volumes/)])
I see two possibilities to fix this issue:
# one is for local deployment (Docker for Mac)
# one is for cluster deployment (production.
For *local deployment* we could use statical created persistent volumes (using
local host volume). In my case, it worked on Docker for Mac using one of the
available file share directories (e.g. /tmp --> however here I faced a Docker
for Mac bug because on my Mac /tmp is actually /private/tmp, thus I had to use
/private/tmp as my mount point in the helm charts).
For *cluster deployment* it kind of depends on how the k8s cluster is setup,
what kind of storage plugin is used etc. Here, we would not want to have
statically assigned persistent volumes on node level.
However, we could provide our helm chart with the necessary configuration to
work with statically assigned host volumes and point out in the README.md to
make the additional changes for cluster deployment.
> Cannot login after k8s restart
> ------------------------------
>
> Key: STREAMPIPES-264
> URL: https://issues.apache.org/jira/browse/STREAMPIPES-264
> Project: StreamPipes
> Issue Type: Bug
> Components: Installer
> Affects Versions: 0.67.0
> Environment: k8s
> Reporter: Kiryl Batsiukov
> Assignee: Patrick Wiener
> Priority: High
>
> after installed from helm chart to k8s (on a Docker Desktop k8s instance),
> cannot login into the streampipes web UI after a cluster restart.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
