Skip to main content

Logs & Diagnostics

Agent Logs​

A default install runs three pods in the opsworker namespace: the agent, kubernetes-mcp-server, and grafana-mcp. The commands below target the agent using the label selector app.kubernetes.io/name=opsworker-agent.

Viewing Logs​

# Current logs
kubectl logs -n opsworker-agent -l app.kubernetes.io/name=opsworker-agent

# Last 100 lines
kubectl logs -n opsworker-agent -l app.kubernetes.io/name=opsworker-agent --tail=100

# Follow logs in real time
kubectl logs -n opsworker-agent -l app.kubernetes.io/name=opsworker-agent -f

# Previous container logs (if pod restarted)
kubectl logs -n opsworker-agent -l app.kubernetes.io/name=opsworker-agent --previous

To inspect the MCP sidecar pods, target them by name:

kubectl logs -n opsworker-agent -l app.kubernetes.io/name=kubernetes-mcp-server
kubectl logs -n opsworker-agent -l app.kubernetes.io/name=grafana-mcp

Log Levels​

LevelMeaning
INFONormal operation: connection established, commands executed
WARNRecoverable issues: temporary connectivity loss, retry
ERRORAction needed: authentication failure, persistent connectivity issues

Common Log Messages​

MessageMeaning
"Connected to SQS"Agent successfully connected to OpsWorker
"Polling for commands"Normal operation: waiting for investigation commands
"Executing command"Agent is gathering data for an investigation
"Command completed"Data collection successful
"Connection error"Cannot reach SQS: check network/proxy
"Authentication failed"Invalid cluster token

Investigation Logs​

In the OpsWorker portal, each investigation has a conversation log showing:

  • Which AI agents were invoked
  • What data was requested from the cluster
  • The AI's reasoning process
  • How the root cause was identified

View these in the investigation detail page.

Cloud-Side Logs​

OpsWorker's cloud infrastructure logs are managed by the OpsWorker team. For issues that require cloud-side investigation, contact OpsWorker support.

Next Steps​