Skip to main content

Deployment logs and limitations

Last updated on

Harness deployment logging has the following limitations:

Log size limits

The following table shows the log size limits for different Harness modules:

Limit TypeCI ModuleOther Modules (CD, CCM, etc.)
Total log size per step5MB25MB
Maximum size per log line70KB25KB

Log truncation behavior

  • When logs exceed the size limit, Harness truncates the oldest log lines and retains the newest ones.
  • You can export full CI logs to an external cache. For information about CI log limits and exporting CI logs, go to Truncated execution logs.

Viewing limits

  • Line limit: The Harness log service has a limit of 5000 lines. Logs rendered in the Harness UI are truncated from the top if the logs exceed the 5000 line limit.
note

If you need logs more than 5000 lines, contact Harness Support

  • Step runtime limit: Steps exceeding a 5-hour runtime may experience log visibility issues. We are working to address this limitation.

Real-time log display behavior

When running pipelines, you may notice differences between real-time log display and the complete logs available after step completion:

Initial display limitation

  • During step execution, the initial log display shows approximately 10,000 lines in real-time
  • The complete log becomes visible after refreshing the page once the step is complete
  • This behavior is by design and does not affect the execution of your steps

Log batching optimization

The log system uses batching to optimize log storage and display:

  • The 5,000 line limit applies at the step level per request
  • Log batching allows multiple log lines within each message, effectively increasing the total number of lines stored
  • Logs are flushed every second from the delegate to the log service
  • The batch size depends on how many logs the step produces within that flush interval
  • If a task produces more log lines than the configured limit within 100ms, older logs from that batch may be truncated

Expected behavior

  • Real-time monitoring: During execution, you might see fewer lines than actually generated, or notice a mismatch between log line numbers and displayed content
  • Post-execution: The full log output becomes available after step completion and page refresh
  • Log line count: The total number of visible log lines can vary depending on the log generation rate and batching behavior

Best practices

To work effectively with log display behavior:

  • For critical log analysis, wait for step completion and refresh the page to view complete logs
  • Consider breaking down steps that generate excessive logs into smaller tasks
  • Add identifiers or markers to log lines to make them easily searchable
  • If you notice missing logs during real-time monitoring, refresh the page after step completion to view the full output