So, we have a 2-node cluster in our primary data center and a "mirror" of that in our secondary data center. I understand I can use a VM in the cluster on the secondary data center to serve as the quorom for the cluster in the primary data center. The IP ranges of the physical servers and the VMs are totally different / separate between the 2 data centers.
What are the issues with using a VM qdevice on cluster 2 to serve as the quorom for the nodes in cluster 1 (and vice-versa)?
I can't think of any which is why I am keen to go down this route and not have to purchase a couple of Raspberry Pi's just to maintain quorom.
The only possibility is when the link between the 2 data centers go down (they're within the same campus so a fibre cut between the 2 buildings is a possibility) but that should still allow the 2 nodes within the same data center to achieve quorom and that would be the only time we can't update/reboot any nodes within each data center.
Is there anything else we may be missing?
What are the issues with using a VM qdevice on cluster 2 to serve as the quorom for the nodes in cluster 1 (and vice-versa)?
I can't think of any which is why I am keen to go down this route and not have to purchase a couple of Raspberry Pi's just to maintain quorom.
The only possibility is when the link between the 2 data centers go down (they're within the same campus so a fibre cut between the 2 buildings is a possibility) but that should still allow the 2 nodes within the same data center to achieve quorom and that would be the only time we can't update/reboot any nodes within each data center.
Is there anything else we may be missing?
