Update monitoring guide as dependencies are outdated

[nicolopicasso]

New versions of Kubernetes can't run the charts anymore.

Upgrading charts version is not enough to make it work since prometheus StatefulSet crashes (kube-prometheus-stack:57.2.1)

Tested on AWS-powered cluster created with eksctl folllowing AWS setup guide.

[J12934]

Hi @nicolopicasso
thank you for pointing this out 🙌

Could you open up a PR to increment the version numbers?

[rtfmkiesel]

@J12934 Just updating the numbers is not enough. As OP said, the monitoring charts are broken right now.

From my testing:

• kube-prometheus-stack 57.2.1 works
  • web UI accessible
  • Juiceshop panel loads but has no data obv.
• promtail 6.5.15 works
  • pod starts fine, didn't do more tests
• loki 6.5.0 does not work:

Error: template: loki/templates/write/statefulset-write.yaml:50:28: executing "loki/templates/write/statefulset-write.yaml" at <include (print .Template.BasePath "/config.yaml")
...
error calling include: template: loki/templates/_helpers.tpl:228:19: executing "loki.commonStorageConfig" at <$.Values.loki.storage.bucketNames.chunks>: nil pointer evaluating interface {}.chunks

[ontanj]

any progress on this?

[PandaFood]

A Loki config is needed for the new version. I've made a PR for bumping the versions and adding the config #258