cncf / cncf/cnf-testbed

Issue with Kubespray v2.14.0 (likely related to Multus configuration)

Open
#366 4 comments 0 reactions 1 assignee Claimed by @michaelspedersen View on GitHub
bug k8s
Dominant language
Shell
Stars
165
Forks
43
PR merge metrics
No merged PRs in 30d

Description

Link to issue with details: https://github.com/intel/multus-cni/issues/561

CNF Testbed (Kubespray) runs without errors, but inspecting the cluster afterwards show the following:
```
$ kubectl get all --all-namespaces
NAMESPACE NAME READY STATUS RESTARTS AGE
kube-system pod/calico-kube-controllers-b5c94f8f8-2j5fd 1/1 Running 0 43m
kube-system pod/calico-node-b4kmp 1/1 Running 0 45m
kube-system pod/calico-node-cm4x4 1/1 Running 0 45m
kube-system pod/coredns-dff8fc7d-xps6q 0/1 ContainerCreating 0 42m
kube-system pod/dns-autoscaler-6fcd794dd8-88mb4 0/1 ContainerCreating 0 42m
kube-system pod/kube-apiserver-node0 1/1 Running 0 49m
kube-system pod/kube-controller-manager-node0 1/1 Running 0 49m
kube-system pod/kube-multus-ds-amd64-b6dvr 1/1 Running 0 44m
kube-system pod/kube-multus-ds-amd64-pdpdn 1/1 Running 0 44m
kube-system pod/kube-proxy-57vbc 1/1 Running 0 49m
kube-system pod/kube-proxy-x89hc 1/1 Running 0 46m
kube-system pod/kube-scheduler-node0 1/1 Running 0 49m
kube-system pod/kubernetes-dashboard-667c4c65f8-8474b 0/1 ContainerCreating 0 42m
kube-system pod/kubernetes-metrics-scraper-54fbb4d595-df2ns 0/1 ContainerCreating 0 42m
kube-system pod/nginx-proxy-node1 1/1 Running 0 45m
kube-system pod/nodelocaldns-859qq 1/1 Running 0 42m
kube-system pod/nodelocaldns-gmhdm 1/1 Running 0 42m
```

Issue seems to be Kubespray deploying Multus with `--cni-version=0.4.0`, which should be supported, but it looks like it isn't.

**Workaround:**

With the cluster deployed, update the Multus daemonset:
```
$ kubectl edit ds kube-multus-ds-amd64 -n kube-system
```
Update the arguments to use `--cni-version=0.3.1` instead of `--cni-version=0.4.0` as shown below:
```
- args:
- --cni-conf-dir=/host/etc/cni/net.d
- --cni-bin-dir=/host/opt/cni/bin
- --multus-conf-file=auto
- --multus-kubeconfig-file-host=/etc/cni/net.d/multus.d/multus.kubeconfig
- --cni-version=0.3.1
```
Save the file and wait for the update to propagate. After that all of the above pods should end up in state "Running"

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.