0
votes

I have created a Kubernetes cluster in my PC (master + 2 nodes). Version is 1.13.2. Afterwards, I installed Istio using the guidelines in https://istio.io/docs/setup/kubernetes/download-release and https://istio.io/docs/setup/kubernetes/quick-start (version is 1.0.5).

After checking the status of the pods, I see that the istio-pilot pod is always in status Pending, while some other pods fail as well. After several iterations, the result is always the same.

enter image description here

Initially the Master node has 2 cores/4 GB RAM and I increased it to 5 cores / 6 GB RAM, but the result is still the same. There are also 2 Worker nodes with 2 cores / 2 GB RAM each.

Any ideas what the problem could be?

1
Are you sure that the cluster is fully operational? Are all the system pods running correctly in kube-system namespace?Paul Annetts
Can you please share the full output from 'kubectl get pod/<servicegraph-...> -n <your-istio-namespace> -o yaml'. What installation option exactly you have picked up from Istio`s website ?Nepomucen
it's old issue, try kubectl logs <pilot-pod-name> -n istio-system, seen this behaviour due to insufficient memory. Similar question but for minikube: stackoverflow.com/questions/52519668/…Sumit Murari
Look at kubectl describe <failing pod>. This should include information on why the pods are failing. Since the failures are cluster-wide, it is likely insufficient resources, like @murarisumit suggests.Isaiah Snell-Feikema

1 Answers

0
votes

most likely the pods that are failing could be stateful pods. can you check those manifests and update the volume to emptyDir