Runtime and Deployment Troubleshooting

Diagnose unhealthy services, failed rollouts, routing, permissions, queues, databases, and uptime alerts.

Use Analytics or Services > Overview to identify the affected environment and service, then open the latest deployment, logs, metrics, and uptime checks.

Deployment Does Not Become Healthy

Confirm the application listens on the configured port and interface.
Request the exact health path and verify an accepted status code is returned.
Check startup, container, function, and pre-deploy logs.
Confirm required variables and secrets exist in that environment.
Check CPU, memory, desired capacity, and dependency connectivity.

Route or Domain Fails

Verify the service is included in the environment and has a successful deployment.
Check host, path, alternate routes, certificate, DNS, CDN, and WAF settings.
Test the health endpoint separately from the public application route.
Allow for DNS propagation, but confirm the expected record rather than waiting blindly.

Component Access Fails

Check both sides of the connection: the service needs the derived endpoint or resource variable and an appropriate access grant. Verify the deployed task or function role reflects the latest infrastructure configuration.

Queue or Event Handler Is Not Processing

Confirm the handler points to the correct service target and event or topic.
Check message age, visible message count, dead-letter behaviour, and consumer logs.
Make the handler idempotent before retrying or replaying events.

Database or Cache Connection Fails

Confirm the endpoint variable, port, credentials, encryption mode, network reachability, and environment. Avoid enabling public access as a diagnostic shortcut.

Restart or Roll Back

Restart only when configuration is correct and a fresh process can recover. Roll back when a previous release is known to be healthy and compatible with current database and infrastructure state. See Restart and Rollback.