I had rancher setup and working around a month ago, and I had to leave for a month. When I came back, rancher was constantly restarting and the longest it lasted before restarting again was around 11 seconds, while it normally restarted after 7 seconds from what I could tell. I reinstalled Ubuntu Server and set Rancher back up with the persistent data config, yet it still crashes constantly. Any ideas about why that might be happening?
If you’re using Ubuntu, have a look in your logs to see if the system is failing trying to set nf_conntrack_max - I found that I couldn’t get Rancher to come up due to this issue on first install.
I was able to resolve by doing
sudo sysctl -w net/netfilter/nf_conntrack_max=327680 then manually stopping and starting the Rancher container.
See this GitHub issue - seems it’s an upstream K8s issue that’s unresolved.