Checking Deployment Status¶
After deploying, you will want to know whether it succeeded, what is running, and how to read logs. AppMecca provides several commands for this.
The dashboard updates live — no refresh needed
On the dashboard, an app's status reacts in real time. After you click the Deploy button it stays busy and phases through Building → Deploying → Deployed (and is un-clickable while busy, so a stray second click can't queue a duplicate), and the deployment-preview screenshot appears the moment it's captured. You don't need to refresh the page to see progress.
Quick Status Check¶
Deployments are tracked as queue actions. mecca queue status takes the
action ID that mecca app deploy / mecca build trigger printed:
The ID can be a full UUID or a unique prefix of at least 8 characters.
Expected output
The key field is Status:
| Status | What It Means | What to Do |
|---|---|---|
queued |
Waiting for an agent to pick it up | Wait a few seconds, then check again |
in_progress |
An agent is working on it | Wait for it to finish |
ready_to_deploy |
Leader set up routing, agents deploying containers | Almost done, wait for completion |
deploy_completed |
Everything is running and healthy | Nothing -- your app is live |
deploy_failed |
Something went wrong | Run mecca queue logs <id> |
cancelled |
The action was cancelled | Re-queue it if you still want it |
Held back rather than failing?
If an action sits at queued for a long time, mecca queue diagnose <id>
explains why in plain language -- an admission gate, an ordering dependency,
or a fleet upgrade in progress.
Forgot the action ID?¶
List the recent actions in the environment and pick it out:
Expected output
View Deployment Detail¶
mecca queue status <id> also renders the placement plan -- which services
were assigned to which hosts. This is useful for verifying that replicas are
spread across different hosts and for diagnosing issues on a specific host.
For the deploy log itself:
View Running Services¶
To see what an app is currently running:
mecca app show takes the app name as a required argument. To see every app
in the environment, use mecca app list instead:
Expected output
View Container Logs¶
Container logs (stdout and stderr from your running application) are collected by
the platform log collector. mecca logs searches every log source -- agent,
deploy, build and container -- in one place. The service is selected with
-s/--service, not as a positional argument:
Expected output
Useful Options¶
# Container stdout/stderr only
mecca logs myapp -s web --type container -e dev
# Widen the time window (default is the last hour)
mecca logs myapp -s web --since 24h -e dev
# Errors only
mecca logs myapp --level error -e dev
# Full-text search
mecca logs myapp --search "timeout" -e dev
# Logs for one deployment action
mecca logs myapp --deployment f82c2671 -e dev
# Return more rows (default 100, max 1000)
mecca logs myapp -n 500 -e dev
mecca logs is a search, not a live tail
There is no -f/--follow flag. Each run returns a snapshot of matching
rows; re-run the command to see newer output.
Branch defaults to all branches
Logs are not branch-scoped, so mecca logs searches every branch by default.
Pass -b main to narrow it.
Not sure which services exist?¶
Run mecca app show myapp -- it lists the app's services and their URLs.
JSON Output¶
--json is a global flag and works on every command, for scripting and automation:
mecca queue status f82c2671 -e dev --json
mecca queue list -e dev --json
mecca app show myapp -e dev --json
mecca app list -e dev --json
Combining with jq
You can pipe JSON output to jq for filtering:
Diagnosing a Failed Deployment¶
When a deployment shows a *_failed status, here is how to investigate:
-
Check the action status for the error message:
-
Read the deploy log for the full execution detail:
-
Check container logs to see if the app started and then crashed:
-
Get the one-shot diagnostic dump if the cause is still unclear:
Common failure causes:
- Health check timeout -- your app did not respond at the configured health check path within
startup_timeout(default 120 s, max 600). Make sure the path is correct and raise the timeout if your app is slow to boot. - Container crash -- the container started but exited immediately. Check container logs for startup errors.
- Out of memory -- the container exceeded its
resources.memorylimit and was killed. Note the value needs a unit suffix (512m,1g) — a bare number means bytes. - Image pull failure -- the image could not be pulled. Verify the build completed successfully with
mecca build status <build-id>.
Next Steps¶
- Undeploying -- remove a running deployment
- Rolling Back -- revert to a previous version
- Services & Logs -- restart services, stop and start containers