Skip to content
LyftData Docs (Beta)

Logs and Issues

Use Logs for historical queries, Problems/Issues for “what’s broken right now”, and Messages when you need a live view. This page describes how these UI surfaces work together and how to troubleshoot common gaps. For job-driven notification patterns (including message actions with log-event: true), see Notifications & Alerts.

Logs

Open Observe → Logs (or visit /logs) to query historical log records emitted by the server and workers.

Common filters:

  • Time range: widen it first if you see “no logs”.
  • Severity: focus on Warning/Error when triaging failures.
  • Worker and Job: narrow to the component you’re investigating.

You can also fetch worker logs via the CLI for quick spot-checks:

Terminal window
lyftdata workers logs <worker-id> --limit 200

Problems / Issues

Open Observe → Problems (or use the Issues panel on a job) to see actionable warnings and errors that need attention. Problems/Issues are designed for triage:

  • Start with the most recent Error entries.
  • Use the job/worker links on each issue to jump directly to context.
  • Pivot into Logs to see surrounding events and repeated failures.

Messages (live)

Open Observe → Messages for a live stream of job and worker activity. It’s most useful when:

  • you are deploying or stopping jobs and want immediate feedback,
  • you are debugging intermittent warnings,
  • you need to correlate worker status changes with job failures.

See Messages for usage and filtering tips.

Troubleshooting

  • No logs or issues show up: widen the time range, confirm the worker is online, and verify your token has read access.
  • Live views look stale: long-lived HTTP connections can be interrupted by reverse proxies, VPNs, and corporate gateways; try a direct connection to the server or adjust proxy settings.