Debugging pod connection issues

Hey All,

I’m having some specific issues with various pods (running dgraph on kubernetes) not connecting after a fair amount of use which I outlined here Pod connection issues · Issue #2818 · dgraph-io/dgraph · GitHub. I was also wondering if there are any good general strategies for debugging (what appear to be) consensus issues - all I’ve been able to do is treat each node as a black box, and try to turn things on and off again.


We usually use LoadBalancer, you are using NodePort. There must be differences there. Check dgraph/contrib/config/kubernetes at master · dgraph-io/dgraph · GitHub

A NodePort service is the most primitive way to get external traffic directly to your service. NodePort, as the name implies, opens a specific port on all the Nodes (the VMs), and any traffic that is sent to this port is forwarded to the service.

Source: Kubernetes NodePort vs LoadBalancer vs Ingress? When should I use what?

So I guess the K8’s DNS works differently tho.

Great, thanks for the feedback. We’re working on debugging that now (it takes a while to recreate the issue). Are multi node clusters generally stable now? Earlier this year there were several connection errors and we were recommended to use single node clusters.

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.