Adding the bridge network to corosync

[IroesStrongarm]

After troubleshooting various network related issues last week, I found the root cause of my problem. I won't go into the whole thing, but it has lead me to decide to add a dedicated switch just for my main corosync and migration network. That said all my nodes have two NICs. I would like to incorporate the NIC I use for management and VMs as a backup corosync network. This would have been practically plug and play if I did it at the start, but I didn't. I've read the documentation here https://pve.proxmox.com/wiki/Cluster_Manager#pvecm_redundancy. What I want to know is, as you can see in my /etc/network/interfaces, my VM network is attached to vmbr0. When editing my corosync.cfg, would I still make ring1_addr: 192.168.1.8 for this node or does it get called differently since it's bound to that virtual bridge?

auto lo
iface lo inet loopback

iface enp9s0f0 inet manual

auto vmbr0
iface vmbr0 inet static
        address 192.168.1.8/24
        gateway 192.168.1.1
        bridge-ports enp9s0f0
        bridge-stp off
        bridge-fd 0

iface enx026662e2a00e inet manual

auto enp9s0f1
iface enp9s0f1 inet static
        address 192.168.2.6/24

 

[shanreich]

It would still be 192.168.1.8

 

[IroesStrongarm]

It would still be 192.168.1.8

Thank you for this confirmation. I assumed it was like this based on the documentation but couldn't find through any research whether it being bound to the virtual bridge complicated this.

Would it be okay if I asked another quick question to you regarding the setup of a migration network that seems to differ from the documentation and the GUI or should I post a separate thread for that question?

 

[shanreich]

Yes, go ahead - although it might be better to create a new thread so it is easier to find for future reference.

 

[IroesStrongarm]

Yes, go ahead - although it might be better to create a new thread so it is easier to find for future reference.

I reviewed the documentation here about the migration network https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_migration_network

At the bottom it references the setup in the datacenter.cfg file that it refences the IP of the subnet network as a whole, ie:

migration: secure,network=192.168.2.0/24

By setting it in the GUI however it sets that config file to say

migration: network=192.168.2.6/24,type=secure

This is obviously referencing this one specific node for all nodes in this config instead of just the subnet. Is it still correct this way?

 

[shanreich]

Yes, the CIDR suffix /24 indicates in this case that the last digit gets masked - it means that it uses IP addresses ranging from 192.168.2.0 - 192.168.2.255

 

[IroesStrongarm]

Yes, the CIDR suffix /24 indicates in this case that the last digit gets masked - it means that it uses IP addresses ranging from 192.168.2.0 - 192.168.2.255

Got it, so even having the 192.168.2.6/24 is perfectly acceptable due to the /24? Sorry for essentially repeating you, just want to be certain for my own understanding and education. Thank you

 

[shanreich]

Yes, exactly!

 

[IroesStrongarm]

Yes, exactly!

Excellent! Thank you very much for taking the time. It's always good to learn more!