Proxmox 3.2: Why is the cluster not working?

Question

Clustering

Raised a cluster of 2 nodes, they even somehow gathered into a cluster, but in the web interfaces the neighbor is always red + there is no quorum, you have to restart like this: /etc/init.d/cman restart
and in another console, while waiting for the quorum:
pvecm e 1
after that, the node is put into operation, but in fact without a quorum. A different set of machines is spinning on both, so it’s not possible to simply clear 1 node. The method from version 2 (migrate machines + config file) doesn't work, because /etc/pve is not a real fs, vzmigrate from the console doesn't work (it can't write the config - see above), it just doesn't work from the web.
Up to the heap, it is not possible to enter the transition to unicast in cluster.conf, when the cluster is restarted, it resets the config to default.
Quite to the heap - corosync eats 60-70% of memory from 64 gig, you have to kill it and restart the cluster services.
Who works closely with 3.x, can you tell me something?

Answer 1

When you edit cluster.conf, you must change the configuration version:
Read here: pve.proxmox.com/wiki/Two-Node_High_Availability_Cl...
Dmitry.