Operations
Operational troubleshooting
Diagnose deployment and integration failures.
| Symptom | Check |
|---|---|
| Database connection error | Confirm DATABASE_URL takes precedence and has a valid host and port; check private-network DNS. |
| Migration mismatch | Run the migration-status and schema-validation commands against a disposable or approved database. |
| Worker not consuming jobs | Confirm REDIS_URL, queue names, worker deployment, and shared database access. |
| Provider failures | Inspect Alchemy/Chainstack credentials, health checks, failover configuration, and the RPC timeout. |
| Webhook signature failure | Verify the raw body, secret, timestamp, and the exact timestamp.body canonical string. |
| Dashboard status looks stale | Confirm the web process, Redis, payment listener, reconciler, and provider-webhook intake are all healthy. |
Do not clear or truncate a database as a first diagnostic step. The application's clear script requires an explicit execute flag precisely because it is destructive.
Related documentation
- Deployment — the release sequence these checks support.
- Environment variables — the configuration referenced above.
- Merchant troubleshooting — the merchant-facing symptom table.