I just installed 3x2TB WD reds in my 2.x OMV system. I put them all together in a raid5 config that is building now as we speak.. But for killing time purpose I started reading about OMV i.c.w ZFS and to be honest it appeals to me as well. What I want to ask is; What is the better choice? Is there a performance difference between the two? And if I want to use ZFS do I need to delete the raid config and start a new pool?
Lots of questions in one post. I hope somebody can answer them...
I would say in general, raid 5 with an ext4 filesystem is going to be faster than zfs because zfs is a CoW (copy-on-write) filesystem. BUT, zfs is giving you additional features/protection. The are both good options. Try zfs if you like the features. Delete the current raid config first.
+1 to ZFS.
more secure, same speed as Ext4 for copy on 1GB LAN ( LAN is the bottleneck)
+1 for ZFS.
Just remember to read up on ZFS limitations if you plan to expand.
Personally I use ZFS, and it works great, but I do hope to migrate to BTRFS soon, as the expanding of BTRFS raid is more appealing for homeusers (add one or two drives at the time, and change raidlevels on the fly if needed).
+1 for ZFS.
Just remember to read up on ZFS limitations if you plan to expand.
Personally I use ZFS, and it works great, but I do hope to migrate to BTRFS soon, as the expanding of BTRFS raid is more appealing for homeusers (add one or two drives at…
ZFS is capable of this! All you have to do is enable the "autoexpand" flag for your zpool! The major limitation IMO of ZFS for home users is its overhead RAM requirement. The general rule of thumb is 8 GB minimum as a base, the +1GB for every 1TB RAW. Now of course this is a general rule of thumb. For most home uses, 16 GB is perfectly fine. ECC is also highly recommended for ZFS to prevent corruption, but there is alot of argument in favor of just doing a weekly reboot to combat this, which should be no issue for home users and is easily automated.
I used ZFS on FreeNAS for years before I got sick of FreeBSD's goofy jail permission system. So I switched to MDADM with EXT4 in OMV and never looked back. For what I use it for, it is great. And I utilize Crashplan as a headless service that runs nightly backups of my data.
ZFS is capable of this! All you have to do is enable the "autoexpand" flag for your zpool! The major limitation IMO of ZFS for home users is its overhead RAM requirement. The general rule of thumb is 8 GB minimum as a base, the +1GB for every 1TB RAW. Now of course this is a general rule of thumb. For most home uses, 16 GB is perfectly fine. ECC is also highly recommended for ZFS to prevent corruption, but there is alot of argument in favor of just doing a weekly reboot to combat this, which should be no issue for home users and is easily automated.
I used ZFS on FreeNAS for years before I got sick of FreeBSD's goofy jail permission system. So I switched to MDADM with EXT4 in OMV and never looked back. For what I use it for, it is great. And I utilize Crashplan as a headless service that runs nightly backups of my data.
ZFS is not able to add single disk in raidz/raidz2. You can add a single disk as a VDEV but if this disk breaks, then whole pool is dead.
Autoexpand is described here: http://docs.oracle.com/cd/E192…819-5461/githb/index.html , it gives the option to utilize the whole new disk if replacing a smaller drive.
ZFS is not able to add single disk in raidz/raidz2. You can add a single disk as a VDEV but if this disk breaks, then whole pool is dead.
Autoexpand is described here: docs.oracle.com/cd/E19253-01/819-5461/githb/index.html , it gives the option to utilize the whole new disk if replacing a smaller drive.
You are correct. I re-read the original post and mixed up what he was asking. Auto expand is for swapping out disks for larger ones. Thanks for catching that!
Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!
