Hi All,
I am facing an issue with Consul v1.12.6 (deployed in EKS as a StatefulSet) where peer communication frequently gets disrupted, causing the associated services to be deregistered from the service list. It would be helpful if someone could suggest a permanent solution to mitigate this problem. Please find the relevant logs below for further insight.
Logs:
2025-09-08T11:43:19.829Z [WARN] agent: Check is now critical: check=talaria-0:http
2025-09-08T11:43:19.834Z [INFO] agent: Synced check: check=talaria-0:http
2025-09-08T11:43:44.830Z [WARN] agent: Check is now critical: check=talaria-0:http
2025-09-08T11:44:04.843Z [INFO] agent: Synced check: check=talaria-0:http
2025-09-08T11:49:00.358Z [WARN] agent: Check is now critical: check=talaria-0:http
2025-09-08T11:49:00.365Z [INFO] agent: Synced check: check=talaria-0:http
2025-09-08T11:49:25.372Z [WARN] agent: Check is now critical: check=talaria-0:http
2025-09-08T11:49:50.375Z [WARN] agent: Check is now critical: check=talaria-0:http
2025-09-08T11:50:11.391Z [ERROR] agent.server.memberlist.lan: memberlist: Push/Pull with xmidt-consul-server-1 failed: dial tcp [2a05:d014:a36:4309:9181::7]:8301: i/o timeout
2025-09-08T11:50:15.378Z [WARN] agent: Check is now critical: check=talaria-0:http
2025-09-08T11:50:27.972Z [INFO] agent: Deregistered service: service=talaria-service
2025-09-08T11:50:27.973Z [INFO] agent: deregistered service with critical health due to exceeding health check’s ‘deregister_critical_service_after’ timeout: service=talaria-service check=talaria-0:http timeout=1m0s
Does upgrading it to the latest version will solve the problem? An early response would be highly appreciated.
Thanks,
Parm