Hello!
Occasionally we’ll see a problem where you can’t execute shell or grab logs from any container on a specific host. Recently, I noticed that a container coming up was never getting its DNS entries from rancher (resolv.conf was never being updated). Even though Rancher was marking this container as “running” (post networking).
I restarted the rancher-agent on the affected host and everything returned to normal.
There were no errors in the rancher-agent, or the rancher-agent-instance handling 500/4500. No apparent errors in /var/log/rancher/agent.log
We’re on rancher 0.39
Thanks!
Topper