Skip to content

Latest commit

 

History

History
41 lines (29 loc) · 2.23 KB

kind.md

File metadata and controls

41 lines (29 loc) · 2.23 KB
Raw

KIND

Setup ovn-kubernetes on KIND

You can follow this guide either with docker or podman

Load local image

You can load any local docker images in kind without pushing them online kind load docker-image my-custom-image-0 my-custom-image-1

Example for dev goflow-kube image on ovn cluster: kind load docker-image quay.io/netobserv/goflow-kube:dev --name ovn

Enable EphemeralContainers feature gate

Simply add EphemeralContainers: true in featureGates section of your kind cluster yaml configuration file

For ovn-kubernetes : $(go env GOPATH)/src/github.com/ovn-org/ovn-kubernetes/contrib/kind_ephemeral.yaml.j2

Remove + Create IPFIX configuration to speed up flow collection

This force OVS to resend templates to allow kube-renricher to get flow right away You need to execute ovs-vsctl -- clear Bridge br-int ipfix and ovs-vsctl -- set Bridge br-int ipfix=@i -- --id=@i create IPFIX targets=\"172.17.0.1:2055\" on any node/ovnkube-node-XXXXX of you ovn-kubernetes namespace.

This can be done on the first onvkube-node by running the following command:

kubectl get pods -n ovn-kubernetes -l name=ovnkube-node --no-headers=true | awk 'NR==1{print $1}'| xargs -I %  sh -c '
	\kubectl exec % -n ovn-kubernetes -c ovnkube-node -- ovs-vsctl -- clear Bridge br-int ipfix;
	\kubectl exec % -n ovn-kubernetes -c ovnkube-node -- ovs-vsctl -- set Bridge br-int ipfix=@i -- --id=@i create IPFIX  targets=\"172.17.0.1:2055\";'

Troubleshooting

[github.com/ovn-org/ovn-kubernetes](ovn-kubernetes repository) may be in a non stable state. If you have issues, feel free to checkout previous revision like commit ac5ce1c951ac3533eebed24ec75fb2ddfcb5adf1 (HEAD -> master) Date: Wed Oct 13 11:40:48 2021 -0500

If your kubectl respond connection refused or missing config, check your KUBECONFIG variable export KUBECONFIG=${HOME}/admin.conf

✗ Joining worker nodes 🚜 There are various reasons to get this error but you may restart your docker and kubelet services before trying any other fixes

You can use -wk 1 option to use only one worker and lower ressources usage