After K8s Pod is deleted with K8s controller replica=0, the IPv6 address on VF won't be deleted (sometimes it need more than 10min)

[scopher128]

What happened?

After K8s Pod is deleted with K8s controller replica=0, the IPv6 address on VF won't be released immediately. (The IP address is added within Pod container) and then change the replica=1, host kernel report the DAD error. After initial debugging, the IP address will be kept and removed slowly, sometimes it need more than 10min. So if we want to configure the same IP address on the new startup Pod, there is the DAD error reported.

What did you expect to happen?

After K8s Pod is deleted with K8s controller replica=0, the IPv6 address on VF shall be released immediately, so that the new startup the Pod can use the same IP address.

What are the minimal steps needed to reproduce the bug?

1. create one network based on VF device pool;
2. create one Pod with replica=1 with the network, there is one container to configure the IP address for the interface;
3. delete Pod with K8s controller replica=0 and check the Pod is removed
4. create Pod with K8s controller replica=1
5. demsg can monitor the DAD message.

Anything else we need to know?

Component Versions

Please fill in the below table with the version numbers of components used.

Component	Version
SR-IOV Network Device Plugin	<Input Version Here>
SR-IOV CNI Plugin	<Input Version Here>
Multus	<Input Version Here>
Kubernetes	<Input Version Here>
OS	<Input Version Here>

Config Files

Config file locations may be config dependent.

Device pool config file location (Try '/etc/pcidp/config.json')

Multus config (Try '/etc/cni/multus/net.d')

CNI config (Try '/etc/cni/net.d/')

Kubernetes deployment type ( Bare Metal, Kubeadm etc.)

Kubeconfig file

SR-IOV Network Custom Resource Definition

Logs

SR-IOV Network Device Plugin Logs (use kubectl logs $PODNAME)

Multus logs (If enabled. Try '/var/log/multus.log' )

Kubelet logs (journalctl -u kubelet)

[killianmuldoon]

This probably isn't an issue with the SRIOV device plugin. Can you offer some more detail on this? What IPAM plugin are you using to provide the address? What NIC are you creating the VF from? What driver and version does it use? What VF driver are you using?

[zshi-redhat]

@scopher128 we you still hitting this issue? if yes, how does the ipv6 address be allocated to VF? could you share more details?

[martinkennelly]

@scopher128 any update?