Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

When deciding whether to use LVM volume groups or ZFS pools in configuring new file servers, what needs to be considered? Is there a "better" choice in the realm of multi-purpose file servers? Unlike this previous question, I don't want to layer the technologies.


  • RHEL / CentOS 6 x64 servers
  • many available, identical DAS and SAN LUNs


I am personally quite familiar with LVM, so am comfortable using it if it is the better option. However, ZFS looks pretty promising, and learning new technology is always good.

Given that we want to be able to share-out a fairly large store (multiple TB) to different departments, and they need to be able to access them over both CIFS and NFS, should we use ZFS or LVM on for the underlying pool manager?

I know that using a product like FreeNAS is possible, but for a variety of reasons, I need to be able to roll-out onto "typical" Linux servers.

share|improve this question
In short, ZFS filesystems are the way to go... I'll follow up later. – ewwhite Apr 17 '14 at 16:27
I've been using ZFS for a few years now. The most compelling argument in its' favor (in my opinion) is this (from here) showing mean time to data loss. Anyway, that's why I am running raid-z3. – Elliott Frisch Apr 17 '14 at 16:42
Regarding FreeNAS: this might be a nice solution, but as many NAS-like solutions, there are quite a few things hidden (ie a small swap partition at the beginning of each disks, ...). You might learn a thing or two playing with the beast, but if you know what you are doing, then you should do it all by yourself. – Ouki Apr 17 '14 at 23:39
I stumbled into your question and have to strongly disagree with your accepted answer--not because ZFS is a poor choice, but because the example given is a textbook study in "what not to do with ZFS" – STW Nov 26 '14 at 19:58
@warren it's because the answer really should come with a large warning that it could very well end in catastrophic data loss. ZFS's redundancy against hardware failures and data corruption are both undermined in the answer's approach--he's using ZFS solely as a volume manager (which is fine). ZFS is a successor to RAID; running ZFS on RAID results in the all the risks of RAID and none of the benefits of ZFS. – STW Nov 26 '14 at 22:31
up vote 10 down vote accepted

I use both, but prefer ZFS. ZFS on Linux has been very good to me, but isn't the "fix all" for every situation.

A typical server will look like this:

(Remember, I usually use hardware RAID and mostly use ZFS as a flexible volume manager)

  • Hardware RAID with a logical volume comprised of underlying disks. That array will be carved into a small OS volume (presented as a block device), then partitioned (/,/usr,/var and such).
  • The remaining space will present another block device to be used as a ZFS zpool.
Smart Array P420i in Slot 0 (Embedded)    (sn: 001438029619AC0)

   array A (SAS, Unused Space: 1238353  MB)

      logicaldrive 1 (72.0 GB, RAID 1+0, OK)
      logicaldrive 2 (800.0 GB, RAID 1+0, OK)

      physicaldrive 1I:2:1 (port 1I:box 2:bay 1, SAS, 900.1 GB, OK)
      physicaldrive 1I:2:2 (port 1I:box 2:bay 2, SAS, 900.1 GB, OK)
      physicaldrive 1I:2:3 (port 1I:box 2:bay 3, SAS, 900.1 GB, OK)
      physicaldrive 1I:2:4 (port 1I:box 2:bay 4, SAS, 900.1 GB, OK)

I then take the zpool and create the additional ZFS filesystems (or mountpoints) and zvols as necessary.

# zpool list -v vol1
vol1   796G   245G   551G    30%  1.00x  ONLINE  -
  wwn-0x600508b1001c4d9ea960806c1becebeb   796G   245G   551G         -

And the filesystems...

# zfs list
vol1                245G   539G   136K  /vol1
vol1/images         245G   539G   245G  /images
vol1/mdmarra        100G   539G   100G  /growlr_pix

So, using ZFS for data partitions is extremely nice because it allows you to address a pool of data, set quotas and manage attributes at mountpoint granularity. LVM still requires dealing with filesystem tools and is a bit more rigid.

share|improve this answer
using ZFS on RAID is discouraged and partly defeats the benefits of ZFS: – STW Nov 26 '14 at 19:07
Your example above appears to completely undermine the benefits of ZFS. Having four physical disks as a single RAID logical disk in a zpool means that you don't gain ZFS's data integrity, your RAID controller will prevent ZFS from monitoring and responding to disk health, and you're sacrificing memory for effectively no gains. Your setup is an ideal candidate for LVM--not ZFS. Quotas, attributes (not supported by ZFS on Linux), and features such as compression can be had without ZFS. – STW Nov 26 '14 at 19:56
@STW you're not in a position to inform me of what works or doesn't in my environments. Take it somewhere else. – ewwhite Nov 26 '14 at 21:55
@ewwhite I would just note that anyone using this setup is risking catastrophic data loss. If you really only want to use ZFS for volume management that's fine, but it's only a small part of ZFS's functionality and loses the safeguards. – STW Nov 26 '14 at 22:25
You're relatively safe if you're running RAID10 and have your zpool configured to copies=2. However in the event of a disk or controller failure you're more likely to end up with a complete loss of your pool than if you had the equivalent setup purely in ZFS. – STW Nov 26 '14 at 22:48

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.