Going Live Checklist
Before You Go Live
Going live with your Aelix integration is straightforward, but completing this checklist will prevent the most common production issues.
✅ Security
- All credentials stored in a secrets manager — not in code or config files
- Production client secret is different from sandbox secret
- IP allowlist configured for your production egress IPs
- OAuth scopes reviewed — request only what you need
- HTTPS enforced for all internal service-to-service calls
✅ Reliability
- Exponential back-off retry logic implemented for 429 and 5xx responses
- Idempotency keys used for all payment and write operations
- Token refresh logic tested — application does not crash on token expiry
- Circuit breaker in place to stop cascading failures if the API is unavailable
- Timeouts configured (recommended: 30s connection, 60s read)
✅ Observability
- All API responses logged with trace_id for debugging
- Alert configured when error rate exceeds 1% over 5 minutes
- Alert configured when API quota reaches 80%
- Webhook delivery monitoring set up in the portal
✅ Functional Testing
- Happy-path flows tested end-to-end in sandbox
- All error codes handled gracefully in the UI
- Webhook signature verification confirmed working
- Pagination tested with >100 records
- Concurrent request load test completed without errors
✅ Compliance
- Data handling reviewed against applicable regulations (GDPR, PCI-DSS)
- Audit log retention policy configured
- Customer-facing consent flows reviewed with legal
Ready? Switch to Production
Update your environment variable from sandbox to production and your base URL from sandbox-api.aelix.digitalapi.ai to api.aelix.digitalapi.ai. Start with a small percentage of live traffic and monitor closely for the first 24 hours.