Alright so here's what I've been tackling...
We have 10 of our servers on Proxmox hosts, it looks like this:
Host 1: 5 servers
Host 2: 5 servers
Host 3: Backup, before we bought Arcserve
Recently, we decommissioned Host 3, but wanted to upgrade our Proxmox version as it was still running 4.4
So I've been attempting to play musical chairs with the servers (as I don't have a license). I've been trying to transfer the VMs on H2, to H3 (after upgrading the software on H3).
I did this, for the most part, by backing up the VMs to a seperate LVM directory I created on their respective local host (just using Mode:Stop; lzo (fast)) and keeping the VMs shut down until brought up on the new server to avoid conflict in LDAP.
I transferred them using the scp command from the shell of the original host (H2) with the following syntax:
scp /mnt/backups/dump/VM-XXX.vma.lzo root@10.10.10.10:/mnt/backups/dump/
4/5 of these VMs transferred successfully, then restored successfully on the new machine/PM version.
The issue is the 5th.
It's a database server that contains more than 4TB of data, uncompressed.
Every time I attempt to transfer it (approximately 3 times during the downtime I scheduled), it partially transfers, then the shell itself times out (at least that's what it seems like).
My boss suggested that instead of backing up locally, I could just backup to the new box, then restore from there. They are all on the same local network, but it doesn't seem like I can cluster them to do that.
TL;DR
I'm attempting to transfer a large VM via scp to another Host (Version 4.4-1 to Version 5.3-9) and the shell seems to be timing out, less than halfway through. Is there a better way to do this?
We have 10 of our servers on Proxmox hosts, it looks like this:
Host 1: 5 servers
Host 2: 5 servers
Host 3: Backup, before we bought Arcserve
Recently, we decommissioned Host 3, but wanted to upgrade our Proxmox version as it was still running 4.4
So I've been attempting to play musical chairs with the servers (as I don't have a license). I've been trying to transfer the VMs on H2, to H3 (after upgrading the software on H3).
I did this, for the most part, by backing up the VMs to a seperate LVM directory I created on their respective local host (just using Mode:Stop; lzo (fast)) and keeping the VMs shut down until brought up on the new server to avoid conflict in LDAP.
I transferred them using the scp command from the shell of the original host (H2) with the following syntax:
scp /mnt/backups/dump/VM-XXX.vma.lzo root@10.10.10.10:/mnt/backups/dump/
4/5 of these VMs transferred successfully, then restored successfully on the new machine/PM version.
The issue is the 5th.
It's a database server that contains more than 4TB of data, uncompressed.
Every time I attempt to transfer it (approximately 3 times during the downtime I scheduled), it partially transfers, then the shell itself times out (at least that's what it seems like).
My boss suggested that instead of backing up locally, I could just backup to the new box, then restore from there. They are all on the same local network, but it doesn't seem like I can cluster them to do that.
TL;DR
I'm attempting to transfer a large VM via scp to another Host (Version 4.4-1 to Version 5.3-9) and the shell seems to be timing out, less than halfway through. Is there a better way to do this?