CLI 'argocd admin cluster stats' doesn't work with round-robin

[zeusal]

Checklist:

• [x ] I've searched in the docs and FAQ for my answer: https://bit.ly/argocd-faq.
• [ x] I've included steps to reproduce the bug.
• [ x] I've pasted the output of argocd version.

Describe the bug

I do not see the statistics correctly when the cluster uses the round robin algorithm.

The application controller works correctly with round robin (at least I can see round robin in logs.) but argocd cli admin cluster stats not.

To Reproduce

• Install argocd
• Config argocd with 3 replicas in application controllers and algorithm of sharding: round-robin
• Login with argocd cli into cluster argocd
• Run: argocd admin cluster stats

Values:

crds:
  install: true
  keep: true

global:
  logging:
    level: debug
  params:
    controller.sharding.algorithm: "round-robin"

controller:
  replicas: 3

Expected behavior

I hope to see the values of shards correctly.

Screenshots

Application controller log:

Version

argocd: v2.9.3+6eba5be
  BuildDate: 2023-12-01T23:24:09Z
  GitCommit: 6eba5be864b7e031871ed7698f5233336dfe75c7
  GitTreeState: clean
  GoVersion: go1.21.4
  Compiler: gc
  Platform: linux/amd64
argocd-server: v2.9.3+6eba5be
  BuildDate: 2023-12-01T23:05:50Z
  GitCommit: 6eba5be864b7e031871ed7698f5233336dfe75c7
  GitTreeState: clean
  GoVersion: go1.21.3
  Compiler: gc
  Platform: linux/amd64
  Kustomize Version: v5.2.1 2023-10-19T20:13:51Z
  Helm Version: v3.13.2+g2a2fb3b
  Kubectl Version: v0.24.2
  Jsonnet Version: v0.20.0

Logs argocd cli

INFO[0000] Starting configmap/secret informers
INFO[0000] Configmap/secret informer synced
INFO[0000] Using filter function:  legacy
INFO[0000] Cluster with uid: d7b3dddc-3fa9-4794-8275-720a0c36e206 will be processed by shard 0
SERVER                          SHARD  CONNECTION  NAMESPACES COUNT  APPS COUNT  RESOURCES COUNT
https://kubernetes.default.svc  0                  0                 0           0

I have been see the code of argocd and I can find the following lines: https://github.com/argoproj/argo-cd/blob/v2.9.3/cmd/argocd/commands/admin/cluster.go#L125C77-L125C101 common.DefaultShardingAlgorithm = LegacyShardingAlgorithm LegacyShardingAlgorithm = "legacy" https://github.com/argoproj/argo-cd/blob/v2.9.3/common/common.go#L118C5-L118C5

does this mean that "admin cluster stats" doesn't work with any other algorithm than "legacy"?

Also, when I have configured the algorithm as legacy, I can see the same information (zero shard, zero apps).

[vitorscassiano]

I got the same error.

[zkck]

Hello, I would like to work on this issue.

[zkck]

Hello all,

It seems that this is no longer an issue with the current versions of the ArgoCD CLI.

I setup ArgoCD locally with the provided values and added a sample app. I built ArgoCD from source for the CLI commands.

~/argo-cd (master)> alias argocd "go run github.com/argoproj/argo-cd/v2/cmd"
~/argo-cd (master)> argocd version
argocd: v99.99.99+unknown
  BuildDate: 1970-01-01T00:00:00Z
  GitCommit:
  GitTreeState:
  GoVersion: go1.22.7
  Compiler: gc
  Platform: linux/amd64
argocd-server: v2.12.3+6b9cd82
  BuildDate: 2024-08-27T11:57:48Z
  GitCommit: 6b9cd828c6e9807398869ad5ac44efd2c28422d6
  GitTreeState: clean
  GoVersion: go1.22.4
  Compiler: gc
  Platform: linux/amd64
  Kustomize Version: v5.4.2 2024-05-22T15:19:38Z
  Helm Version: v3.15.2+g1a500d5
  Kubectl Version: v0.29.6
  Jsonnet Version: v0.20.0

I then compared the output of argocd admin cluster stats at the latest commit of master (0573ed79) and at the version specified in this issue, v2.9.3.

~/argo-cd ((0573ed79))> alias argocd "go run github.com/argoproj/argo-cd/v2/cmd"
~/argo-cd ((0573ed79))> argocd admin cluster stats
SERVER                          SHARD  CONNECTION  NAMESPACES COUNT  APPS COUNT  RESOURCES COUNT
https://kubernetes.default.svc  0      Successful  1                 1           367
~/argo-cd ((v2.9.3))> argocd admin cluster stats
SERVER                          SHARD  CONNECTION  NAMESPACES COUNT  APPS COUNT  RESOURCES COUNT
https://kubernetes.default.svc  0                  1                 0           0

It seems the latest commit on master now picks up the resources in the shard correctly.

Please let me know if I missed something, otherwise I think we can close the issue.