Tunnelbroker deployments on prod sometimes failed due to The target group tunnelbroker-grpc-tg does not have an associated load balancer.
This was due to disabled LB listener for gRPC, so the target group had no listeners.
Detaching the ECS Service from adding Tunnelbroker to this blind target group resolved the issue.
Added appropriate comments
Details
Terraform apply succeeded on prod after previously failing with the message. Opened AWS Console and made sure Tunnelbroker is not registered to the gRPC target group.
Diff Detail
- Repository
- rCOMM Comm
- Lint
No Lint Coverage - Unit
No Test Coverage
Event Timeline
services/terraform/remote/service_tunnelbroker.tf | ||
---|---|---|
190 | Just double checking my understanding. Because there won't be any listeners, essentially this for_each will result in no `load_balancer" blocks actually being deployed. Why do we still need this block in the first place? Testing purposes? |
services/terraform/remote/service_tunnelbroker.tf | ||
---|---|---|
190 | We want it for staging, and yes, for "testing purposes".
Yes, exactly. No load_balancer blocks on prod, and one on staging. |