Move physical hard disk from one proxmox node other proxmox node

[easpem]

Hi, first of all thanks for any help to this newbie

Im looking for information about moving a physical hard drive from one proxmox server (node1) to another proxmox server (node2), both same version - same hardware except hard drive space. (node1 has been freezing so im looking to move everything to node2. and later decomisson the failing hardware)

i've already migrated 3 Containers and 1 VM. to node2. The Second VM i'd like to move was installed on a dedicated hard disk and i was wondering what's needed to install it on the second node as i'm missing how to tell the new vm(node2) the name of the image thats already on the dedicated disk. (there is no enough space on node 2 to migrate this second VM =( )


Thanks!

 

[LnxBil]

The answer depends on what the exact setup on your node 1 is for the VM with the dedicated harddisk. What was the motivation for this setup in the first place? Normally, virtualization abstracts this away.

 

[easpem]

Hello LnxBill

I thought putting a new dedicated disk for this big VM would be a good way to go preformance wise. Now i know not so much.

Node1 conist of 2 hard disk. One with proxmox, 3 CT's and one VM, a second HD for a single VM /dev/sdb usage: LVM, Content: Disk image, Container

I did a little home work and found the config files of VM's are stored on /etc/pve/qemu-server

This is the config of my VM

bootdisk: scsi0
cores: 3
memory: 3072
name: UCS-Bachi
net0: virtio=E6:2D:F7:55:BD:BC,bridge=vmbr0,firewall=1
numa: 0
onboot: 1
ostype: l26
scsi0: Univention:vm-101-disk-0,size=700G
scsihw: virtio-scsi-pci
smbios1: uuid=c05166ce-b2df-4148-8514-493b63e8c982
sockets: 1
vmgenid: 2ec1d3d0-ecfa-47e2-ab99-15c11b801c67

I created a new VM on node2 with same config, and wonder if pasting this config to the VM on Node2 config file would be a step in the right direction.

Thanks

 

[LnxBil]

I did a little home work and found the config files of VM's are stored on /etc/pve/qemu-server

You need to create the same storage on the other node and mount also the disk equivalently as on node1. The VM configuration is one step, other steps include the files /etc/pve/storage.cfg and maybe also /etc/fstab.

Node1 conist of 2 hard disk. One with proxmox, 3 CT's and one VM, a second HD for a single VM /dev/sdb usage: LVM, Content: Disk image, Container

Okay, so no redundancy on the disk front. Normally, you get more performance by combining multiple disks logically in a RAID. All except RAID0 is used on servers to increase performance and redundancy and is the industry standard for server for many decades. If you only run your setup at home, single disks are fine, but if you have production data on PVE, please move to a raided setup.

 

[easpem]

Given your advice ill get a couple of new disk to set them up in RAID . this would be a 2TB array 2x 1TB disks. for this one VM.

Could you give me an advice on which storage type would be ok for a the 8 GB RAM the system has. i read ZFS needs al least 1 GB per terabyte.

I leave 2 GB RAM for Proxmox 3GB for the VM itself, 1GB for one CT . The other VM's and CT's only run ocassionaly. Would this be too tight?

Thanks.

 

[LnxBil]

Would this be too tight?

No, memory is not an issue per se, but as it is used as caching, you might run into performance "problems" (whereas the word "problem" is a bit strong). You will - as with every system - have reduced performance if you cache is not big enough, but that is common.