You know that feeling when you’re all set to start working on something, your plan is clear, and you’re ready to take action — but your homelab decides otherwise? That was me recently.
I urgently needed a few print screens for a presentation I was giving at the UK VMUG, but of course, that’s exactly when my environment decided to break. Time was short, and troubleshooting wasn’t really part of my agenda that day.
Normally, when I start vRealize Automation (vRA), my homelab gets a little noisy — fans spin up, and I know it’s alive. But this time, it was eerily silent. When I tried connecting to vRA, I got the classic:
“Hmm. We’re having trouble finding that site.”
So, I already knew something hadn’t started properly. When I checked the status of my nodes, I got an unfamiliar error message — one I’d never seen before. Time to dig in.
Getting More Details from kubectl
After some searching on our dear friend Google, I found that you can add extra parameters to your get nodes command to get more detailed troubleshooting output:
- k get nodes -v=10
From the output, I noticed it was performing a DNS lookup but returning an empty array. That led me to check the contents of my /etc/hosts file — and sure enough, at the bottom of the print screen, the entry was empty.
Fixing the Missing /etc/hosts Entry
There’s a VMware KB article that explains how to fix this by copying the /etc/hosts content from another appliance in your cluster (link). But of course, this being a lab environment, I only have a single node — no cluster to copy from. Just my luck again.
Fortunately, I knew that one of my customers had recently deployed a standard vRA setup, so I decided to borrow their /etc/hosts content and try it in my environment.
Here’s the content that saved me:
127.0.0.1 localhost
127.0.0.1 vra-k8s.local
###k8s service hostnames follow. *.svc.cluster.local are automatically updated
10.244.6.59 pgpool.prelude.svc.cluster.local
10.244.6.233 proxy-service.prelude.svc.cluster.local
10.244.5.31 identity-service.prelude.svc.cluster.local
10.244.4.230 docker-artifactory.prelude.svc.cluster.local
10.244.6.154 provisioning-service.prelude.svc.cluster.local
10.244.7.53 vco-service.prelude.svc.cluster.local
Once you’ve copied this into your /etc/hosts file, just run:systemctl restart kubelet
Then give it a few minutes for the services to start up — and you’re good to go! Homelab peace restored. 🌩️
Kristof's virtual life 