Prior to upgrading Docker to 1.9 (RHEL channel) when I looked at /var/log/messages I saw plenty of messages with swarm querying the hosts:
Mar 27 03:51:29 dockerhost1 docker: time=“2016-03-27T03:51:29.688520673+01:00” level=info msg="GET /v1.15/containers/json?all=1&size=0"
Mar 27 03:51:29 dockerhost1 docker: time=“2016-03-27T03:51:29.691176216+01:00” level=info msg=“GET /v1.15/images/json?all=1”
Since I upgraded to 1.9 I’ve only just noticed that i’m getting different errors:
Apr 26 20:37:40 dockerhost1 journal: time=“2016-04-26T20:37:40.000217160+01:00” level=info msg="{Action=networks, LoginUID unknown, PID unknown}"
Apr 26 20:37:40 dockerhost1 journal: time=“2016-04-26T20:37:40.002614556+01:00” level=warning msg="tcpconn is not a struct"
Apr 26 20:37:40 dockerhost1 journal: time=“2016-04-26T20:37:40.002643253+01:00” level=error msg=“Unable to read peer creds and server socket address: bad file descriptor”
I’ve ensured the swarm images are all fully up to date now however i’m still getting these errors even after a full restart of the manager servers and the other servers also.
Has anyone got any advice on where I can check for any further information or got any ideas to try?
Overall it doesn’t seem to be stopping the swarm from managing the cluster but these errors are somewhat strange.
Thansk