[BUG] [Gitops] source-controller of Flux has the CrashLoopBackOff status

[selmison]

Describe the bug
When using this Tutorial to deploy apps using GitOps in the AKS, the app sample running normally and the walk-through in this tutorial is success.

But when take a look in the details on the pods of flux inside of AKS cluster, I notice the container related to source-controller has the CrashLoopBackOff status and it is in a restart loop every 5 min.

I am using Azure as network-plugin and Calico as network policy.

The behavior is the same, even though I disable readiness/liveness probe is disable

To Reproduce
Steps to reproduce the behavior:
1.

az k8s-configuration flux create -g flux-demo-rg \
-c flux-demo-arc \
-n cluster-config \
--namespace cluster-config \
-t connectedClusters \
--scope cluster \
-u https://github.com/Azure/gitops-flux2-kustomize-helm-mt \
--branch main  \
--kustomization name=infra path=./infrastructure prune=true \
--kustomization name=apps path=./apps/staging prune=true dependsOn=\["infra"\]

2. run kubectl get pod -n flux-system -w and wait for about 5 min

3. run kubectl get pod -n flux-system -w and wait for about 5 min:

Note: the log in this pod does not show any error.

Expected behavior
The source-controller deploy keep healthy.

Environment (please complete the following information):

• CLI Version 2.37
• Kubernetes version: 1.24.9
• CLI Extension version:
  • azure-devops: 0.26.0
  • k8s-extension: 1.4.0
  • k8s-configuration: 1.7.0

[ghost]

Action required from @Azure/aks-pm

[ghost]

Issue needing attention of @Azure/aks-leads

[ghost]

Issue needing attention of @Azure/aks-leads

[carvido1]

Hello @selmison .

Can you try to get the events produced by the source controller? To do so you can use

kubectl describe deployment/source-controller -n flux-system
kubectl describe pod/source-controller... -n flux-system

There will be a section for events that can add some context here. On the screenshot you have shared we can appreciate that the pod has been killed once due to OOM.

In addition if you are monitoring your cluster, maybe you can present some metrics for the memory on the particular pod.

[selmison]

Hello, the root of this issue was related to fluxcd/source-controller#929. My repo had a big size