Verify a deployment
Read deployment status and logs, and know what a healthy deployment looks like.
A successful deploy is not just "no red banner". This page explains the signals Notploy exposes so you can confirm a release works and diagnose it when it does not.
Deployment status
Each deployment moves through a lifecycle. The important states:
| Status | Meaning |
|---|---|
queued | Waiting for a worker. A long queue usually means many builds are in flight. |
building | Producing the image from the source. |
running | Container released and healthy. |
error | The build or release failed — open the logs. |
Queued deployments can be cancelled; those already running cannot.
Where to look
- Deployments tab — the history of the last deployments for the service, with the build log of each. When a deploy fails, the log names the step that broke.
- Logs tab — real-time runtime output from the running container(s). This is where you see application errors after a successful build.
- Monitoring tab — CPU, memory, disk and network for the service. Metrics only update while the page is open.
- Server view — for host-level health (disk pressure, Docker usage).
What "healthy" means
A deployment is healthy when:
- its status is
running, - the container repeatedly passes its health check (if configured), and
- the domain, if any, serves requests.
If the build succeeds but the app does not answer, the problem has moved from build to runtime — check the container port and environment variables before redeploying.
Failure triage
| Symptom | First thing to check |
|---|---|
| Build fails | Deployment log — the failing step and its stderr. |
| Deploy runs, app unreachable | Container port vs. domain container port. |
| App crashes on start | Runtime logs — usually a missing env var or bad config. |
| Works locally, fails here | Missing environment variables or build-type mismatch. |
| Intermittent 502 | Health checks, resource limits, or a crash loop. |
Recovering
- Roll back to a previous working deployment — see Rollbacks.
- Redeploy after fixing configuration.
- Investigate with Troubleshooting.