Helm chart 6.2.10 fails health check with no logs

Self Hosted Retool
1

Hello,

Upgrading from version 6.2.8 of the Helm chart to 6.2.10 fails.
The jobs-runner pods run fine but the actuall Retool pods fail to become ready.
kubectl logs shows nothing on the pods.

Running on AWS EKS 1.28

Any tips?

Thanks,
Sayadi

1 Like
2

Hi @Sayadi! Welcome to the community. :slightly_smiling_face:

My general recommendation when running into issues with an upgrade is to explicitly restart the jobs-runner service. Even if it doesn't look like the logs have anything relevant, I'd be interested in taking a look if you want to share them via DM!

3

Thanks @Darren

The previous 6.2.8 version worked fine for now. Will probably upgrade to the next release after 6.2.10 when it comes out.

And, what I meant by no logs was literally no logs, not nothing relevant; the pods produced nothing at all.

4

Ah got it. :thinking: Without any logs, it's hard to say what the issue really is. For what it's worth, the only significant change in the chart seems to be enabling the Java database connector by default:

image
image1866×334 43.7 KB

Maybe try disabling it? If that doesn't work, rolling back to v6.2.8 is fine. :+1:

I'll do some more digging on this end, as we would definitely prefer upgrades to go through seamlessly. Don't hesitate to reach out if you run into any issues in the future!