helm-dasbhoard pod crashing due to resource usage

[chitudorin]

I have deployed helm-dashboard using the helm chart and after adding a repository and trying to install a helm chart (while browsing different versions of that helm chart to see the manifest) the pod immediately spikes in resource usage and dies. I removed the memory limits once and I noticed that it spiked up to 8GB of memory usage.

I am running a latest-version RKE2 cluster installed using the tarball method.

# kubectl version
Client Version: v1.28.7
Kustomize Version: v5.0.4-0.20230601165947-6ce0bf390ce3
Server Version: v1.27.12+rke2r1

Hardware info:

CPU: Intel Core i7-7700HQ
RAM: 32GB
Disk Usage: 50% out of 240GB

After install the usage is normal:

# kubectl -n helm-dashboard top pods
NAME                              CPU(cores)   MEMORY(bytes)   
helm-dashboard-64495d4658-kdxmq   169m         39Mi

Adding the bitnami helm charts (https://charts.bitnami.com/bitnami) and attempting to install a helm chart, browsing different versions for install I see the memory spike:

# kubectl -n helm-dashboard top pods
NAME                              CPU(cores)   MEMORY(bytes)   
helm-dashboard-64495d4658-kdxmq   877m         546Mi

And it eventually crashes:

# kubectl -n helm-dashboard top pods
error: Metrics not available for pod helm-dashboard/helm-dashboard-64495d4658-kdxmq, age: 4m41.264431643s

# kubectl -n helm-dashboard describe pod helm-dashboard-64495d4658-kdxmq
Events:
  Type     Reason                  Age                  From                     Message
  ----     ------                  ----                 ----                     -------
  Warning  FailedScheduling        5m15s                default-scheduler        0/1 nodes are available: pod has unbound immediate PersistentVolumeClaims. preemption: 0/1 nodes are available: 1 Preemption is not helpful for scheduling..
  Warning  FailedScheduling        4m37s                default-scheduler        0/1 nodes are available: pod has unbound immediate PersistentVolumeClaims. preemption: 0/1 nodes are available: 1 Preemption is not helpful for scheduling..
  Normal   Scheduled               4m35s                default-scheduler        Successfully assigned helm-dashboard/helm-dashboard-64495d4658-kdxmq to nirod
  Normal   SuccessfulAttachVolume  4m24s                attachdetach-controller  AttachVolume.Attach succeeded for volume "pvc-3e6974d4-aba1-4298-8f94-c52dbeac8200"
  Normal   Pulling                 4m23s                kubelet                  Pulling image "komodorio/helm-dashboard:1.3.3"
  Normal   Pulled                  3m45s                kubelet                  Successfully pulled image "komodorio/helm-dashboard:1.3.3" in 38.062571434s (38.062606493s including waiting)
  Warning  Unhealthy               3m43s                kubelet                  Liveness probe failed: Get "http://10.42.0.49:8080/status": dial tcp 10.42.0.49:8080: connect: connection refused
  Normal   Created                 38s (x2 over 3m45s)  kubelet                  Created container helm-dashboard
  Normal   Started                 38s (x2 over 3m45s)  kubelet                  Started container helm-dashboard
  Normal   Pulled                  38s                  kubelet                  Container image "komodorio/helm-dashboard:1.3.3" already present on machine
  Warning  Unhealthy               36s (x5 over 3m44s)  kubelet                  Readiness probe failed: Get "http://10.42.0.49:8080/status": dial tcp 10.42.0.49:8080: connect: connection refused

I tried to remove the resource limits and the pod reached 8GB of memory usage at one point: