ArgoCD Troubleshooting

This guide provides solutions for common ArgoCD issues and troubleshooting steps for maintaining a healthy ArgoCD deployment.

Common Issues and Solutions

Sync Failures

Issue: Application Won’t Sync

Symptoms:

• Application shows “OutOfSync” status
• Sync operations fail or time out

Troubleshooting Steps:

1. Check for sync errors in the status:

   kubectl get application <app-name> -n argocd -o jsonpath='{.status.conditions}' | jq

2. Check resource constraints:

   kubectl top pod -n argocd

3. Examine application controller logs:

   kubectl logs -n argocd -l app.kubernetes.io/name=argocd-application-controller -c argocd-application-controller

Solutions:

• Force a refresh of the application:

  kubectl patch application <app-name> -n argocd --type=merge -p '{"metadata":{"annotations":{"argocd.argoproj.io/refresh":"hard"}}}'

• Restart the application controller:

  kubectl rollout restart deployment argocd-application-controller -n argocd

Resource Issues

Issue: Resource is “Progressing” Forever

Symptoms:

• Resources stay in “Progressing” health status
• Deployments never reach ready state

Troubleshooting Steps:

1. Check the health status details:

   kubectl get application <app-name> -n argocd -o jsonpath='{.status.resources}' | jq 'map(select(.health.status=="Progressing"))'

2. Check the problematic resource directly:

   kubectl describe <resource-kind> <resource-name> -n <resource-namespace>

Solutions:

• Check for resource constraints in target namespace
• Verify the image exists and is accessible
• Check for pod errors using kubectl describe pod

Issue: Resources Out of Sync but No Changes

Symptoms:

• Application shows “OutOfSync” but Git comparison shows no differences

Solutions:

• Enable resource tracking via labels:

  kubectl patch application <app-name> -n argocd --type=merge -p '{"spec":{"syncPolicy":{"syncOptions":["CreateNamespace=true","RespectIgnoreDifferences=true"]}}}'

• Add ignore differences for fields that change frequently:

  kubectl patch application <app-name> -n argocd --type=merge -p '{"spec":{"ignoreDifferences":[{"group":"apps","kind":"Deployment","jsonPointers":["/spec/replicas"]}]}}'

ApplicationSet Issues

Issue: ApplicationSet Not Generating Applications

Symptoms:

• No applications are created from an ApplicationSet
• Only some of the expected applications are created

Troubleshooting Steps:

1. Check ApplicationSet controller logs:

   kubectl logs -n argocd -l app.kubernetes.io/name=argocd-applicationset-controller

2. Verify the ApplicationSet spec:

   kubectl get applicationset <appset-name> -n argocd -o yaml

Solutions:

• Restart the ApplicationSet controller:

  kubectl rollout restart deployment argocd-applicationset-controller -n argocd

• Check if your generator (git, list, cluster, etc.) is correctly configured

• Verify Git repository access and credentials

Authentication and Access Issues

Issue: Repository Connection Issues

Symptoms:

• Unable to connect to Git repository
• “connection refused” or SSL errors

Troubleshooting Steps:

1. Check repository secret:

   kubectl get secret <repo-secret> -n argocd -o yaml

2. Test connection using ArgoCD CLI:

   argocd repo list

Solutions:

• Update repository credentials
• Check network policies that might be blocking outbound connections
• Verify SSL certificates for private repositories

ArgoCD Component Diagnostics

Component Health Check

Run a quick diagnostic on all ArgoCD components:

kubectl get pods -n argocd
kubectl top pod -n argocd
kubectl describe statefulset,deployment -n argocd

Fixing Stuck ArgoCD Components

If a component is in a bad state:

# Force restart the component
kubectl rollout restart deployment <component-name> -n argocd

# For example, to restart the API server:
kubectl rollout restart deployment argocd-server -n argocd

Checking ArgoCD Server Health

# Check server status
kubectl exec -it $(kubectl get pods -n argocd -l app.kubernetes.io/name=argocd-server -o name | head -n 1) -n argocd -- argocd admin server health

Recovering from Serious Issues

Dealing with Corrupted Applications

1. Export the corrupted application:

   kubectl get application <app-name> -n argocd -o yaml > corrupted-app.yaml

2. Delete the application without cascade:

   kubectl delete application <app-name> -n argocd --cascade=false

3. Edit the exported YAML to fix issues, then reapply:

   kubectl apply -f fixed-app.yaml

Recovering from Database Issues

If the ArgoCD Redis database is corrupted:

# Backup before attempting recovery
kubectl get all,secrets,configmaps,applications,appprojects -n argocd -o yaml > argocd-backup.yaml

# Delete and recreate Redis
kubectl delete pod -n argocd -l app.kubernetes.io/name=argocd-redis

Performance Troubleshooting

Addressing High CPU/Memory Usage

# Check resource usage
kubectl top pod -n argocd

# Scale up application controller if needed
kubectl scale deployment argocd-application-controller -n argocd --replicas=2

# Adjust application controller processing parallelism
kubectl edit configmap argocd-cmd-params-cm -n argocd
# Add --application-controller-operation-processors=10

Optimizing for Large Clusters

For clusters with many applications:

# Set higher limits for the controller
kubectl patch deployment argocd-application-controller -n argocd --type=json -p='[{"op":"replace","path":"/spec/template/spec/containers/0/resources/limits","value":{"cpu":"2","memory":"4Gi"}}]'

See Also

• ArgoCD Application Management
• Namespace Management
• Working with Finalizers