[Bug] - Readiness probe fails, database driver unknown about data-center-helm-charts

[errcode1202]

Hi @ondrejbilcik,

From the info you have supplied I cannot see anything immediately obvious that would cause this. For some more detail would it be possible to supply the configured values.yaml you are using for this deployment?

[bianchi2]

@ondrejbilcik also a more detailed log will help, I wonder if jdbc url is picked up.

[errcode1202]

@ondrejbilcik also a more detailed log will help, I wonder if jdbc url is picked up.

Yes this was my thought too, is it possible the url is malformed.

[ondrejbilcik]

Hi,
Sorry, somehow missed the notifications about your replies.
Attaching the values.file and a bigger log dump. Thanks
values.txt
logs.txt

[bianchi2]

@ondrejbilcik thanks for sharing the files. What I noticed in the logs is:

INFO:root:/var/atlassian/application-data/jira/dbconfig.xml exists; skipping.

This is how Jira entrypoint works - if the dbconfig.xml exists, it's not overwritten. Perhaps, you passed some malformed value the first time you installed the Helm chart?

I'd suggest delete helm release, make sure PVCs are deleted as well.

OR you can exec into jira container and remove /var/atlassian/application-data/jira/dbconfig.xml file, kill jira pod and let the entrypoint re-generate the file. It'd be great if you can cat this file and share it as well (we may see what's wrong there)

@errcode1202 may have more info on whether or not overwriting dbconfig.xml is possible in the latest versions - we did have a discussion around that recently.

[ondrejbilcik]

@bianchi2 removing the file and killing the pod worked. Thanks a bunch.
dbconfig.txt

[bianchi2]

@ondrejbilcik awesome! Yes, currently, if you install the chart and pass malformed db related values, dbconfig.xml is messed and not overwritten when you run helm upgrade. This will be fixed as far as I know.