|operations/deployment-charts||master||+1 -1||Bump eventstreams chart version.|
|operations/deployment-charts||master||+29 -332||charts:eventstreams bump common_templates and standardize labels|
|Resolved||Jelto||T292390 Upgrade all deployment charts to use the latest version of common_templates|
|Resolved||gmodena||T310721 eventstreams chart should use latest common_templates|
@akosiaris, good news, Gabriele is working on this!!!
@Jelto @JMeybohm, it seems the upgrade to common_templates means that this will be a totally new k8s deployment, as the names of the resources have changed. @gmodena and I need some help doing this deployment, as a helmfile apply won't work. I think we need to delete the deployment and reinstall it.
We can test this in staging, but when we do this for any prod services, we should probably depool at the traffic routing later so we don't break any active usages. @gmodena will sync with you on this after we are all back from our respective offsites, (so early October :) ).
BTW, since we merged the helm chart changes, eventstreams is currently undeployable. We rarely deploy it anyway, so I don't think this is a problem, but we will need to revert if we want to do any eventstreams changes.
Hi, I'll be your SRE support for today, and will handle de/repooling, destroying the old chart, and applying the new.
For now, I'll check if destroy/apply works correctly in staging and report back.
Destroy/apply done in staging:
# helmfile -e staging status helmfile.yaml: basePath=. Getting status production NAME: production LAST DEPLOYED: Wed Oct 12 10:01:41 2022 NAMESPACE: eventstreams STATUS: deployed REVISION: 1 NOTES: Thank you for installing eventstreams. [...] # helm3 -n eventstreams history production REVISION UPDATED STATUS CHART APP VERSION DESCRIPTION 1 Wed Oct 12 10:01:41 2022 deployed eventstreams-0.5.0 Install complete # kubectl get pods -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINE SS GATES eventstreams-production-6f4fd9bc49-6kwgx 2/2 Running 0 10m 10.64.75.225 kubestage1004.eqiad.wmnet <none> <none>
eventstreams-internal is still used?
I am not sure! I'd imagine folks use it, as it is a really nice GUI to view any of our streams, which is nice for quick testing and debugging. However, the ssh tunnel needed to connect to it might be a barrier enough that causes no one to use it. It is just as easy for me to use a CLI client, so I don't use it much.
For now, we should probably just redeploy it too.
I suppose could consider removing it in another task.