What to do when it fails
Calm first steps when a seenode deploy fails: check logs, fix 502s and failed builds, and restore credits when a service is suspended.
Last updated September 18, 2026
Something failed. Start here before changing random settings.
1. Open Logs
On the service, open the Logs tab.
| Source | When to use it |
|---|---|
| Build | The deploy stopped while installing dependencies or compiling |
| Runtime | The build finished, but the process crashes or never listens |
| Request | The process is up, but HTTP responses look wrong |
Static sites have build (and request) logs; there is no runtime process. Search and switch between a live tail and a fixed range (up to 14 days).
2. Match the symptom
Build failed
The deploy stops in the build stage. Typical causes: missing requirements.txt / package.json / go.mod, wrong Root directory, or a start command that does not match the build output.
502 Bad Gateway
The public URL responds 502. Almost always the app is not listening on 0.0.0.0 at the Port configured on the service.
Service suspended (LIMITS)
The service shows as suspended because the workspace ran out of credits. Add credits from Buy more credits. Restore balance promptly — suspended services can be deleted if credits are not restored within about 7 days.
Tip (Prevent the next suspension)
Enable auto top-up (“Low credit prevention”) so seenode adds credits when your remaining balance is under about 9 days of usage.
3. Framework-specific errors
These belong with the framework guide. Use them only when the log names that stack:
Django DisallowedHost
Add your domain to ALLOWED_HOSTS.Phoenix SECRET_KEY_BASE
Set the production secret before boot.Elixir cacerts_get
Build against OTP 25+.More framework walkthroughs: Frameworks.
Still stuck
Email help@seenode.com with the service name and the relevant Log lines.