Closed floriangrundig closed 5 years ago
@floriangrundig In the agent log, does the error message show that both connecting to acs
and tcs
had issues or you only see the error message for tcs
?
@floriangrundig In the agent log, does the error message show that both connecting to
acs
andtcs
had issues or you only see the error message fortcs
?
@yumex93 👇
2019-08-19T07:24:31Z [INFO] Done waiting; reconnecting to ACS
2019-08-19T07:24:51Z [ERROR] acs: unable to discover poll endpoint, err: RequestError: send request failed
caused by: Post https://ecs.eu-west-1.amazonaws.com/: dial tcp: lookup ecs.eu-west-1.amazonaws.com on 10.25.6.35:53: read udp 10.102.22.188:56563->10.25.6.35:53: i/o timeout
2019-08-19T07:24:51Z [INFO] Reconnecting to ACS in: 2m15.692563724s
As my colleague showed - both are affected...
Hi @floriangrundig ,
Can you enable debug logs for amazon ecs agent and send the debug logs to utsa@amazon.com?
closing this issue since we havent received debug level logs. if you're able to repro with debug logging, please package the logs using the log collector and send the logs to ecs-agent-external at amazon dot com, and feel free to reopen the issue then.
Hi all,
Summary
ECS-Agent logs show error 'unable to discover poll endpoint' and probably as consequence the container instance doesn't start/update new ecs tasks:
Description
Seems to be similar to #676.
Agent version 1.29.1 Docker version 18.06.1-ce
When running
we get the following output
The cluster consists of three container instances - a fresh registered (terminated old and autoscaling created new) container instance doesn't have this issue ...