Poor-man DRP with two nodes

[skn666]

Hi,

I have a node 1 that I want to back up in DRP every day (not enough performance/money for BCP)

I created a Node 2 with the idea that if Node 1 fails, I manually "switch" to Node 2 (just changing DNS)...

I first thought about a PBS, but from what I see, it requires at least, 3 machines...

Are there other (simple) solutions?
Replication ? clone ?

Thanks for you help...

 

[Chris]

Hi,

I have a node 1 that I want to back up in DRP every day (not enough performance/money for BCP)

I created a Node 2 with the idea that if Node 1 fails, I manually "switch" to Node 2 (just changing DNS)...

if you plan to cluster these 2 nodes, please follow the advice to also add an external voter as described here [0].

I first thought about a PBS, but from what I see, it requires at least, 3 machines...

Yes, the recommended way is to setup PBS on a standalone host for better fault tolerance.

Are there other (simple) solutions?
Replication ? clone ?

If you have ZFS as a storage, you could use storage replication to duplicate the guests from one node to the other [1].


[0] https://pve.proxmox.com/pve-docs/pve-admin-guide.html#_corosync_external_vote_support
[1] https://pve.proxmox.com/pve-docs/pve-admin-guide.html#chapter_pvesr

 

[Johannes S]

One Addition: You could use the PBS as the external voter in the cluster