Skip to main content
Runtime logs show what a live app is doing. Build logs show why a deploy failed. Reach for floo doctor first. floo doctor accounts --app my-app and floo doctor managed-services --app my-app return one verdict with the evidence behind it, and answer posture questions in one round trip instead of correlating the gateway, request logs, and config by hand.

​
Runtime logs

Use floo logs query for a bounded result and floo logs tail for a continuous stream.
Runtime logs are the right tool when the build succeeded but the live app is unhealthy, noisy, or returning errors. For gateway request history filtered by app key, use GET /v1/apps/{app_id}/requests?api_key_id={key_id} with an operator credential. See App API keys and request history for time filters, pagination, and retention limits.

​
Build logs

Use deploy logs for what happened during build and rollout.
Build logs are the right tool when a deploy never goes live or fails before traffic reaches your app.

​
Live deploy watch

Use floo deploys watch when a deploy is in flight and you want the fastest end-to-end view:
This is the quickest way to confirm whether a push-triggered deploy is progressing, blocked, or already finished. For scripts and agents, floo deploys status --app my-app --json returns the same answer as phase booleans without streaming build logs. host_bound: false means the image built and the service is serving on its internal URL, but the gateway has not bound the public hostname yet: it deployed, but users cannot reach it.

​
A practical debug loop

  1. Find the deploy you care about.
  2. Read build logs if the deploy failed before going live.
  3. Read runtime logs if the deploy succeeded but the app is broken.
  4. Fix config or env vars.
  5. Deploy again or let the next push trigger a redeploy.
Example:

​
For agents

Use JSON mode when an agent needs to parse the result:

​
Dashboard surface

The shipped app detail view exposes:
  • service logs panels
  • deploy history panels
  • service-level status and env views
The CLI remains the better surface for scripted filters, streaming, and copy-paste debugging loops.

Agent webhooks

Route failed deploys and repeating runtime errors to your agent or on-call workflow.

Releases and rollbacks

Recover quickly when the right move is shifting traffic back.