Views:

Read pod logs to verify successful runs and identify failed or skipped chats.

The collector emits one structured JSON object per log line. A healthy run emits exactly one list_activities_complete event (showing chats_discovered and the scan window), one chat_succeeded event per processed chat, and possibly chat_skipped or exhausted_retries events for individual failures.

Procedure

  1. Find the most recent job created by the CronJob and tail its logs:
    LATEST_JOB=$(kubectl -n "$NS" get jobs --sort-by=.metadata.creationTimestamp \
  -o jsonpath='{.items[-1:].metadata.name}')

kubectl -n "$NS" logs --selector=job-name="$LATEST_JOB" --tail=200
  2. Filter the output to specific event types with jq:
    kubectl -n "$NS" logs --selector=job-name="$LATEST_JOB" \
  | jq -c 'select(.event=="chat_succeeded" or .event=="chat_skipped"
    or .event=="exhausted_retries" or .event=="batch_dropped")'
    Note
    Note
    A pod that reports chunks_sent: 0 for a chat had no in-window text to scan — this is not a fault. A configuration_error or batch_dropped event means the run failed; see Troubleshoot Kubernetes collector connectivity issues.