Errors
Deploy errors
When a deploy stops, the cabinet shows an error code and a short reason. What each code means and how to fix it.
When a deploy fails, the cabinet says why at the top and shows the most useful line of the log under it. Many errors also end the log with a Tip: line that says exactly what to do. The Details link in the error block leads to one of these pages.
By code
| Code | What happened | Page |
|---|---|---|
no_source, source_gone, clone_failed, no_branch, no_access, too_big |
We could not get the code | Could not get the code |
install_failed |
Dependencies did not install | Install failed |
build_failed, no_build_script |
The build command ended with an error | Build failed |
no_output, no_site, ssr_output |
The build passed, but there is nothing to serve | No site found |
no_start |
No start command | No start command |
start_failed, predeploy_failed |
Built, but did not start | The app did not start |
timed_out |
The build went past ten minutes | Time ran out |
library |
A library, not an app | This is a library |
no_capacity |
The server is full | We start it again ourselves; nothing needed from you |
internal |
An error on our side | Try again; if it repeats, write to us |
Where to look first
- The end of the log. The error itself is usually in the last 10–20 lines.
- The Tip line. When there is one, it is nearly always the fix.
- On your own machine. If
npm run buildorpip install -r requirements.txtfails for you too, the error is in the code, not on our side.