HelmReleases are Upgraded Without Any Valid Trigger

[mhamzarazzaq97]

Describe the bug

I have installed multiple HelmRelease Objects along with HelmRepository, GitRepository & Kustomization Objects. Few helm releases are getting upgraded without any reasons. Helm releases upgrade should be triggered only if there is a change in GitOps repository. There is no change in values of the helm releases, no commits in gitops repository, no manual reconcile or upgrade. I am not able to understand what is triggering helm controller to upgrade the helm releases.

Steps to reproduce

This behavior is random. For some helm releases it never happens, but for some of the releases it happens more often. One thing I observed is that, there is one HelmRelease with a pod exiting with OMM killed many times (once in an hour), it is upgrading that particular helm release very often. Screenshot is attached.

Expected behavior

HelmReleases should be upgraded only when there is some change in GitOps repository.

Screenshots and recordings

helm ls output. This helm release is upgraded 120 times although there is no change in gitrepository or values configs.

OS / Distro

macOS 14

Flux version

flux-2.0.0

Flux check

► checking prerequisites
✔ Kubernetes 1.29.3 >=1.28.0-0
► checking version in cluster
✔ distribution: flux-2.0.0
✔ bootstrapped: false
► checking controllers
✔ helm-controller: deployment ready
► docker.io/fluxcd/helm-controller:v0.35.0
✔ kustomize-controller: deployment ready
► docker.io/fluxcd/kustomize-controller:v1.0.0
✔ notification-controller: deployment ready
► docker.io/fluxcd/notification-controller:v1.0.0
✔ source-controller: deployment ready
► docker.io/fluxcd/source-controller:v1.0.0
► checking crds
✔ alerts.notification.toolkit.fluxcd.io/v1beta2
✔ buckets.source.toolkit.fluxcd.io/v1beta2
✔ gitrepositories.source.toolkit.fluxcd.io/v1
✔ helmcharts.source.toolkit.fluxcd.io/v1beta2
✔ helmreleases.helm.toolkit.fluxcd.io/v2beta1
✔ helmrepositories.source.toolkit.fluxcd.io/v1beta2
✔ kustomizations.kustomize.toolkit.fluxcd.io/v1
✔ ocirepositories.source.toolkit.fluxcd.io/v1beta2
✔ providers.notification.toolkit.fluxcd.io/v1beta2
✔ receivers.notification.toolkit.fluxcd.io/v1
✔ all checks passed

Git provider

No response

Container Registry provider

No response

Additional context

No response

Code of Conduct

• I agree to follow this project's Code of Conduct

[souleb]

So many things happened since v0.38. Please upgrade your Flux version and check whether it fixes your issue. There have been many changes on the helm-controller side.

[mhamzarazzaq97]

@souleb this is version of flux cli on my local machine, these are the details with latest flux cli. I am observing this issue on flux 2.0.0
► checking prerequisites
✔ Kubernetes 1.29.3 >=1.28.0-0
► checking version in cluster
✔ distribution: flux-2.0.0
✔ bootstrapped: false
► checking controllers
✔ helm-controller: deployment ready
► docker.io/fluxcd/helm-controller:v0.35.0
✔ kustomize-controller: deployment ready
► docker.io/fluxcd/kustomize-controller:v1.0.0
✔ notification-controller: deployment ready
► docker.io/fluxcd/notification-controller:v1.0.0
✔ source-controller: deployment ready
► docker.io/fluxcd/source-controller:v1.0.0
► checking crds
✔ alerts.notification.toolkit.fluxcd.io/v1beta2
✔ buckets.source.toolkit.fluxcd.io/v1beta2
✔ gitrepositories.source.toolkit.fluxcd.io/v1
✔ helmcharts.source.toolkit.fluxcd.io/v1beta2
✔ helmreleases.helm.toolkit.fluxcd.io/v2beta1
✔ helmrepositories.source.toolkit.fluxcd.io/v1beta2
✔ kustomizations.kustomize.toolkit.fluxcd.io/v1
✔ ocirepositories.source.toolkit.fluxcd.io/v1beta2
✔ providers.notification.toolkit.fluxcd.io/v1beta2
✔ receivers.notification.toolkit.fluxcd.io/v1
✔ all checks passed

[souleb]

There are memory improvements on the last version v2.3.0. Also increase your memory limits to avoid OOM on your pod.

See https://fluxcd.io/flux/installation/ for installation and upgrade procedures.