Implement health check for background listener and database; add List…

[kali-el]

Summary

Make the indexer fail safely when the background event listener crashes so the API does not continue serving stale data.

What this PR does

Graceful shutdown on listener failure

• Adds shared listener health state so the HTTP API and background listener can report a common health status.
• Makes the event listener stop with a fatal error when polling fails unexpectedly (for example, due to RPC rate limiting or database connectivity issues).
• Triggers graceful shutdown of the HTTP server and database pool when the listener exits with an error, avoiding a stuck-but-healthy service.

Health endpoint improvements

• Updates the /health endpoint to run a DB connectivity check and verify the listener is still active.
• Returns 503 Service Unavailable when the listener is dead or the database is unreachable, making unhealthy instances visible to load balancers and operators.

Files touched

• indexer/main.go
• indexer/api/handlers.go
• indexer/api/health.go
• indexer/api/router.go
• indexer/listener/soroban.go
• indexer/api/router_test.go

Tests

• Added regression tests covering the health endpoint for healthy and degraded states.
• Verified with:

cd indexer
go test ./api ./listener

Verification

• The indexer API tests and listener tests pass locally.
• The /health endpoint now returns 200 when the listener and DB are healthy and 503 when the listener has stopped or the DB check fails.
  closes indexer: initiate graceful shutdown on listener loop failure #31

[vercel]

@kali-el is attempting to deploy a commit to the K1NGD4VID Team on Vercel.

A member of the Team first needs to authorize it.

[drips-wave]

@kali-el Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits