-2

I'm installing bare metal one node Kubernetes cluster on virtual machine using Ansible playbook As virtualization host I'm using ESXi 8.0 and Hyper-V (on Windows 11), OS is RHEL 8.4, calico is network plugin that i use

Problem is that after virtual machine reboot on ESXi 8.0 some of Kubernetes pods not changing Ready status to 1/1 it's always 0/1 but on Hyper-V there no problem at all, all pods are 1/1

NAMESPACE NAME READY STATUS RESTARTS AGE calico-apiserver calico-apiserver-558c58dd84-cgmvt 1/1 Running 4 (<invalid> ago) 6d3h calico-apiserver calico-apiserver-558c58dd84-r4x6z 1/1 Running 4 (<invalid> ago) 6d3h calico-system calico-kube-controllers-cc8f5d649-tnplf 1/1 Running 4 (<invalid> ago) 6d3h calico-system calico-node-cfhhs 1/1 Running 4 (<invalid> ago) 6d3h calico-system calico-typha-596b4b6967-5bv8t 1/1 Running 6 (<invalid> ago) 6d3h calico-system csi-node-driver-9x2bs 2/2 Running 8 (<invalid> ago) 6d3h default nfs-subdir-external-provisioner-6c5b8fff7-shkcg 1/1 Running 4 (<invalid> ago) 6d3h kube-system coredns-565d847f94-mqdjs 1/1 Running 4 (<invalid> ago) 6d3h kube-system coredns-565d847f94-s5wgl 1/1 Running 4 (<invalid> ago) 6d3h kube-system etcd-localhost.localdomain 0/1 Running 4 (<invalid> ago) 32m kube-system kube-apiserver-localhost.localdomain 0/1 CreateContainerError 4 (27m ago) 30m kube-system kube-controller-manager-localhost.localdomain 0/1 Running 4 (<invalid> ago) 6d3h kube-system kube-proxy-gvwmq 1/1 Running 4 (<invalid> ago) 6d3h kube-system kube-scheduler-localhost.localdomain 0/1 Running 4 (<invalid> ago) 6d3h tigera-operator tigera-operator-7b594b484b-pd8p6 1/1 Running 10 (24m ago) 6d3h

Some one facing similar issue ? How fix that issue on ESXi 8.0 so all pods will be 1/1

Trying other ESXi versions (6.5, 6.8, 7.0) and on all versions have same error

old_timer
  • 69,149
  • 8
  • 89
  • 168
MeAngryOne
  • 1
  • 1
  • You need to post the logs of your pods as well to understand where is the issue. If you just delete a pod it should recreated by k8s. Try it and if the new pod run successfully then do it to all pods if still stuck in 0/1 then post your logs here. Remember delete pod not deployment – tauqeerahmad24 Aug 23 '23 at 12:45

1 Answers1

0

It's was not Kubernetes issue. it's hardware/bios issue

Problem was solved by installing ESXi using Dell Lifecycle Controller OS Deployment

MeAngryOne
  • 1
  • 1