Losing Docker bridge network(cbr0) IP address after restarting docker service

[asreedharan]

Hi,
     Whenever I restart docker service in any of the Windows worker node, I'm losing the bridge network interface(L2 bridge - cbr0) IP address. So it breaks the communication between the node and container. Every time I have to reboot the node to retain that IP. Any help appreciated.

[PedroGCPSupport]

Hi, please could you explain where are these Windows worker node running? In addition of that, a brief explanation of what are you doing would be nice to understand the scenario. There is a StackOverflow post with some answers that could pinpoint the root cause or it's more related to internal networking configuration.

[asreedharan]

The worker node is running in a GKE on-prem user cluser in VMWare.

[asreedharan]

Google support says, its a known issue in GKE on-prem, the workaround is to restart flannel service after the docker restart. There will be a fix in the upcoming versions.