VM/docker interface problem with LACP and VLAN's

[vincentvriends]

I configured my unraid server to bind eth0 and eth1 together with LACP to the switch. I also added VLAN's to separate some docker containers and VM's.

When bridging is disabled all docker containers work fine with their assigned bond0.x interface/IP but the VM can only use virbr0.

When bridging is enabled the VM can use the br0.x interfaces but all dockers are stopped. When I try to start them again I get the error "No such container".

When I disable the bridging again, all containers work like expected.

 

How can I fix this issue and make docker and VM's work with the LACP bonding and the VLAN interfaces created with this bond?

 

I'm running unraid NVIDIA version 6.8.3

Attached you can find the diagnostics file.

microserver-diagnostics-20200416-1728.zip

Edited April 16, 2020 by vincentvriends

[vincentvriends]

My networking settings:

 

 

[Chimestrike]

Can I ask if you managed to find a fix for this as I'm currently suffering with the same issue after re-configuring my LACP correctly?

 

[vincentvriends]

On 6/8/2020 at 8:35 PM, Chimestrike said:

Can I ask if you managed to find a fix for this as I'm currently suffering with the same issue after re-configuring my LACP correctly?

 

No, I never found a fix for it... right now I have 2 servers: an unraid server and an ESXi server

[bonienl]

This is perfectly doable. My test server has a LACP bond with 4 members and bridging enabled.

It runs both Docker and VMs.

 

[vincentvriends]

On 7/28/2020 at 8:08 PM, bonienl said:

This is perfectly doable. My test server has a LACP bond with 4 members and bridging enabled.

It runs both Docker and VMs.

 

Can you please share your network settings and maybe the network settings of a VM?