Skip to main content
Monitor more than container uptime. A running application can still fail to generate, process documents, or export presentations when a provider or storage dependency is unhealthy.

​
Monitor continuously

  • Container state and restart count
  • HTTP availability through the same route users access
  • CPU, memory, disk usage, and available space
  • Application and export logs
  • External database availability when configured
  • Provider authentication, latency, errors, and rate limits
  • Growth and backup status of /app_data

​
Functional health test

Periodically create a small presentation using a controlled prompt, confirm the task completes, open the result, and export it. This checks the application, text provider, storage, and export runtime together.

​
Useful diagnostics

Never send unsanitized logs to a public issue because they may contain internal URLs, filenames, or provider details.