We currently have a Primary node, which resides in the US within a Datacenter, but we are going to be using this for all serious bandwidth and resource intensive websites, and through a configuration of Virtualmin + Webmin, will be setup as a sort of pseudo-cluster, using Virtualmins Cluster Modules.
Anyways, on to the issue:
We also have a business line setup locally, with three servers. Here are their specs:
Intel P4 2. 4 GHz, 1 GB RAM, 110 GB SATA, Ubuntu 12. 04*
AMD 1. 3 GHz, 512 MB RAM, 20 GB IDE
P3 Xeon 800mhz (dual physical processors), 1 GB RAM, 3 * 25 GB Raid Configuration (one in use for host operating system).
The first machine is currently IN USE and is serving virtual hosts off a sub-domain.
My question is this: How can I integrate the Secondary node (which will be the Primary node per say, in this smaller configuration...) which is currently in use, into the cluster configuration w/ the other two servers for:
- Sharing Resources
- Redundancy (HA?)
- NFS /w the two Raid Disks
Without having the FORMAT the secondary node, and start fresh moving all my services in to a DRBD network drive or something similar, and than restoring all active virtualmin's Virtual hosts.
The idea is that I want minimal downtime to people currently being served from server2.mywebsite.com, and from what I understand, all services need to be on a NFS so that they can be mounted on demand and accessed from the other machine taking over (i.e. heartbeat + DRBD Config.)
My issue is that I already have all these services installed to their default directory structure:
How can I most easily setup this NFS and HA system, move all my desires services to this new drive, and do it with minimal down time, and without breaking Virtualmin and everything else on my server?