Logs
Every request, error and job in your project leaves a trace. The console gathers that stream, scopes it to your project and makes it readable without server access.
Read the stream
The Logs screen shows your project’s recent window, newest first. Each line carries its timestamp, the service it came from, its level and the message.
- Filter by service — API, Auth, Storage, Realtime, Functions — to isolate the layer involved.
- Filter by level (error, warn, info, debug) to keep only what matters during an incident.
- Search free text (an identifier, a table name, an error message) to find one specific request.
Watch it live
Live mode refreshes the window automatically — handy during a deployment or a test campaign. Otherwise the view stays on the displayed window and reloads on demand.
Diagnose an incident
The stream is scoped by project identifier: you only see your own. To tie a client error to its cause, compare the timestamps.
- Reproduce the action in your application, then read the window again: the service and level point to the layer at fault.
- A failing request leaves a line on the API side; if it never appears, the problem happens earlier, before it reaches your project.
- An Auth error spike often follows an email campaign or an over-strict policy: check the policies before calling it an outage.
Availability and retention
The stream covers a sliding window: beyond it, lines are no longer readable — export whatever you need. If collection is momentarily unavailable, the screen says so instead of showing an empty or misleading stream.