Troubleshoot containers
Diagnose provisioning, deployment, routing, and runtime configuration problems.
Start with the container's Overview, Deployments, and Runtime Log. An accepted request can still be followed by a deployment or runtime failure.
No regions or plans are available
Refresh the creation options and select a region again. Plans are region-specific. If the lists remain empty or the selected option is no longer available, contact support rather than repeatedly submitting the same request.
Creation is rejected because of billing
Check that the project has an active billing account with positive, unexpired credits. Correct the billing configuration before retrying creation.
The image cannot be pulled or the process exits
Inspect the deployment log for the image name, tag, or digest involved. Confirm that the deployment server can access the image. For Artifact Registry, select an image belonging to the current project.
Check that command and entrypoint overrides name an executable present in the image, that the working directory exists, and that required environment variables are set. See Runtime settings.
The application runs but its URL does not respond
Confirm the process's listening port in its configuration or logs. Match that port to Internal HTTP port on Domain & DNS. For custom domains, confirm the CNAME points to the displayed target and run Verify connection after DNS propagation.
Check whether Firewall Policy allows the client's current public IP, including any VPN or outbound gateway. Inspect domain statuses for pending routing or verification work.
An environment or image change has no effect
Confirm that you saved the edit. Restart after environment-variable or editable health check changes. Redeploy after image changes. General-page edits must be saved or discarded before lifecycle actions can run.
The health check fails
For HTTP checks, confirm that the image contains curl or wget and that the configured endpoint responds inside the container. For command checks, confirm the executable exists and returns a successful exit code. Review startup timing and inspect runtime output for application failures.
An operation is already in progress
Wait for the existing operation and refresh the resource status. Do not submit repeated lifecycle requests while one is pending. If the state remains unchanged, contact support with the resource ID, operation or deployment details, and failure time.