01/06/2023, 10:06 PM
👋 Hi everyone! Any experience with Monitoring chart not working on v2.7.0? It seems that pushprox-k3s-server-client is using alpine 3.13.5, for which DNS options ndots:5 could be faulty. (That's my closest guess as far as I understand it yet) The symptoms are - monitoring in Grafana just can't display information about other nodes than the Master node of the cluster. And digging into it, apparently the pushprox-k3s-server-client just cannot resolve service names.
Even though resolv.conf has "search cattle-monitoring-system.svc.cluster.local svc.cluster.local cluster.local" it just doesn't search any of them and just fails. Other containers can perform the exact same nslookups, but the pushprox-k3s-server-client fails. I also found a seemingly old thread about a similar Alpine bug, but as far as I understand from my containers, the bug persists in the 3.13.5 that I have.