Expect table output from Lua script, not nil

[samtbg]

Checklist:

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

Describe the bug

Since upgrading to 2.14 I'm seeing an error in our logs each time a resource (pod, service etc.) is clicked on and viewed in the UI; there aren't any issues in functionality/errors that pop up in the UI though.

To Reproduce

Install the latest version of ArgoCD (via Helm chart), deploy an application/application-set and click on a resource in the UI, then check the server logs for the following error message.

Version

argocd-server: v2.14.2+ad27246

Logs

time="2025-02-28T11:18:28Z" level=error msg="finished unary call with code Unknown" error="error getting available actions: error executing Lua discovery script: expect table output from Lua script, not nil" grpc.code=Unknown grpc.method=ListResourceActions grpc.service=application.ApplicationService grpc.start_time="2025-02-28T11:18:28Z" grpc.time_ms=14.726 span.kind=server system=grpc

[andrii-korotkov-verkada]

Can you share the argocd-cm and application's resources manifests, please?

[samtbg]

Yep sure!

Redacted argocd-cm yaml

apiVersion: v1
data:
  accounts.deployuser: apiKey, login
  admin.enabled: "true"
  application.instanceLabelKey: argocd.argoproj.io/instance
  application.sync.impersonation.enabled: "false"
  dex.config: |
    connectors:
      - config:
          redirectURI: <REDIRECT_URI>
          issuer: https://accounts.google.com
          clientID: <CLIENT_ID>.apps.googleusercontent.com
          clientSecret: <CLIENT_SECRET>
          serviceAccountFilePath: /tmp/oidc/googleAuth.json
          adminEmail: <ADMIN_EMAIL>
          fetchTransitiveGroupMembership: True
        type: google
        id: google
        name: Google
  exec.enabled: "true"
  server.rbac.log.enforce.enable: "false"
  statusbadge.enabled: "false"
  timeout.hard.reconciliation: 0s
  timeout.reconciliation: 10s
  url: https://<URI>
kind: ConfigMap
metadata:
  annotations:
    meta.helm.sh/release-name: argo-cd
    meta.helm.sh/release-namespace: argo-cd
  creationTimestamp: "2022-12-15T13:49:41Z"
  labels:
    app.kubernetes.io/component: server
    app.kubernetes.io/instance: argo-cd
    app.kubernetes.io/managed-by: Helm
    app.kubernetes.io/name: argocd-cm
    app.kubernetes.io/part-of: argocd
    app.kubernetes.io/version: v2.14.2
    helm.sh/chart: argo-cd-7.8.2
  name: argocd-cm
  namespace: argo-cd
  resourceVersion: "1157035672"
  uid: 59cc0234-63b0-43ef-b2b1-af4c9c2b1e74

And this a version of the applicationset we apply

apiVersion: argoproj.io/v1alpha1
kind: ApplicationSet
metadata:
  name: appname
  namespace: argo-cd
spec:
  generators:
  - list:
      elements:
{{ toYaml $.Values.clusters | indent 8 }} 
  template:
    metadata:
      name: '{{ printf "{{colour}}" }}-appname'
      namespace: argo-cd
      finalizers:
        - resources-finalizer.argocd.argoproj.io
      labels:
        name: appname
      annotations:
        recipients.argocd-notifications.argoproj.io: grafana:unknown|succeeded|failed|{{ printf "{{cluster}}" }}|appname
        notifications.argoproj.io/subscribe.on-sync-status-unknown.grafana: unknown|{{ printf "{{cluster}}" }}|appname
        notifications.argoproj.io/subscribe.on-deployed.grafana: succeeded|{{ printf "{{cluster}}" }}|appname
        notifications.argoproj.io/subscribe.on-sync-failed.grafana: failed|{{ printf "{{cluster}}" }}|appname
    spec:
      project: project
      source:
        repoURL: http://<INTERNAL K8S DNS FOR GITLAB>
        targetRevision: main
        path: environments/project
      destination:
        namespace: namespace
        name: {{ printf "{{cluster}}" | quote }}
      syncPolicy:
        automated:
          prune: true
          selfHeal: true
          allowEmpty: false
        syncOptions:
          - CreateNamespace=true
          - PrunePropagationPolicy=foreground
        retry:
          limit: 3
          backoff:
            duration: 5s
            factor: 2
            maxDuration: 3m
      revisionHistoryLimit: 1

Thanks

[AM07]

We are also facing same issue since v2.13.5

[andrii-korotkov-verkada]

So no custom healhchecks with lua defined. There's been some fix for lua nil stuff. It would be great to determine which object causes this. Do you have regular pods, or have them as a part of some customized deployment?

[samtbg]

As in the Argo pods or application pods? Both just regular, Argo's installed via the Helm chart and application pods are using the standard Deployment k8s api.

Thanks

[crenshaw-dev]

Lon Nix pointed out on Slack that I likely introduced this regression while trying to improve the error message when a script was missing. Working on a fix.