I have 3 node cluster where I defined resource r0. The device in 3 equal configs on all 3 nodes is /dev/drbd0. 2 nodes of the cluster running drbd-reactor and 3rd node is just a diskfull witness in order to satisfy quorum of 2.
Now I have to define another similar cluster. So I decided to provision 2 extra VMs and keep 3rd node of 1st cluster as a diskfull “witness” and define additional resource r1. Obviously, in that 2nd cluster I am going to define a resource r1 with /dev/drbd0 device. And of course on 3rd disk full node which is shared among 2 clusters it’s gonna be an r1 resource (because r0 already exist, with its own underlaying partition indeed), the question is: is it okay to call a device on 2nd cluster /dev/drbd0 and only on 3rd node (where /dev/drbd0 already used for 1st cluster) it’s gonna be /dev/drbd1. Is it fine or there must be exactly the same device name per resource per cluster? Also I am confused with node-id, in the first cluster I am using node-id 0, 1 and 2. Is it file to also use node-id 0, 1 and 2 for the second resource? At first look it could be a stupid question, but it’s a time when I would like to avoid surprises. Thank you.
