hey everyone,
So rn I'm running the storage backend for my proxmox ve (8.2.4) box via zfs over iscsi.
Trunas core box provides the iscsi backend.
I have two storage pools on my truenas box:
HDD_Pool:
4x 8TB WD Reds running raidz2 ( atime 12 and 128k record size)
Nvme_Pool:
6x TB Samsung 990 Pros running raidz2 ( atime 12 and 128k record size)
I have added both of my truenas pools to proxmox via Thegrandwazoo 's tool as a zfs over iscsi target.
looking around I see a lot of different takes on what should the zvol blocksize for the vms be.
what would be the recommended block size for my current setup?
my use case is a mix of linux and windows VMs, and some VMs would be running mssql.
if I were to use the default 4k blocksize (of thegrandwazoo 's plugin) will there be any major performance hit to any of my vms?
what I understand currently is that with record size of 128k I can have zvols with block size UPTO 128k however I'm still not sure if its a wise Idea to have zvols with different blocksize under the same dataset.
So rn I'm running the storage backend for my proxmox ve (8.2.4) box via zfs over iscsi.
Trunas core box provides the iscsi backend.
I have two storage pools on my truenas box:
HDD_Pool:
4x 8TB WD Reds running raidz2 ( atime 12 and 128k record size)
Nvme_Pool:
6x TB Samsung 990 Pros running raidz2 ( atime 12 and 128k record size)
I have added both of my truenas pools to proxmox via Thegrandwazoo 's tool as a zfs over iscsi target.
looking around I see a lot of different takes on what should the zvol blocksize for the vms be.
what would be the recommended block size for my current setup?
my use case is a mix of linux and windows VMs, and some VMs would be running mssql.
if I were to use the default 4k blocksize (of thegrandwazoo 's plugin) will there be any major performance hit to any of my vms?
what I understand currently is that with record size of 128k I can have zvols with block size UPTO 128k however I'm still not sure if its a wise Idea to have zvols with different blocksize under the same dataset.