Skip to main content
The Logs page is the dashboard viewer for everything your app writes to stdout and stderr. It is built to cut through noise quickly: a timeline to spot error clusters, drag-to-filter to zoom into a window, and collaboration features so you and your team (or the fal team) can share and debug the same log line. For how logging works under the hood — the different log scopes, what is captured, and how request-level logs are exposed — see Logging.

​
Timeline and Filtering

The logs viewer opens with a timeline histogram of log volume over time, making it easy to see when errors spiked or clustered. From there you can:
  • Drag to filter — select a region on the timeline to narrow all logs to that window, instead of typing exact timestamps.
  • Filter by severity — switch out of live mode and filter to errors (or any minimum level) to focus only on what matters.
  • Filter by identifier — narrow logs by runner ID, request ID, or version ID, or scope to a specific endpoint.

​
Investigating a Log Line

Click into any log entry for options that make debugging faster:
  • Show in context — when you are filtered to error-level logs, expand a single error to see the surrounding log lines and full traceback that led to it.
  • Share link — copy a link that opens directly to that exact log line, at that point in time and with the same filter applied, so a teammate lands on precisely what you are looking at.

​
Collaborating and Exporting

Debugging production apps is often a team effort, so the Logs page makes it easy to move logs out of the dashboard:
  • Copy multiple lines — bulk-select log lines and copy them (as CSV) to paste into Slack for a teammate or into an agent for analysis.
  • Export as CSV — export the full, filtered set of logs at once.
  • Logging — log scopes, what is captured, and request vs runner logs
  • Runner Analytics — logs filtered to a single runner, alongside telemetry and cold start stages
  • Log Drains — forward logs to external services (Datadog, Splunk, or any HTTP endpoint) in real time