Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

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

I have various firewall/router/utility type boxes that do not require much if any storage for their primary purpose other than booting. But I do like to be able to put a more or less standard CentOS/RHEL install on them which takes a few gig. So I normally end up mirroring two small hard drives. Recently VARs have been proposing boxes with a single SSD like the Intel X-25M 40G on the theory that a single drive is simpler/takes up less space/draws less power (all important factors to me).

Any opinions on whether a single SSD is better than a pair of HDs?

share|improve this question
It's actually the "X-25M", correct? I know of no "Intel M25" SSD. – EEAA Apr 3 '11 at 2:33
Er...correct. X-25M. – Tracy Reed Apr 3 '11 at 5:15

What's your definition of "better"?

Will it be faster? Probably.

Will it require less power? Yes.

Will it be more reliable? Maybe not. Even if a single SSD proves to be more reliable than a single standard hard drive (which is up for debate), your data is still not protected from a single-drive failure like it would be in a RAID1 scenario. Additionally, data recovery methods for failed SSD drives are not anywhere close what they are for standard hard drives, so if you have an SSD fail, it's likely you can kiss that data goodbye unless you have it backed up, whereas with a standard hard drive, there are many methods (some expensive, some cheap) that you can employ to get data off of the drive.

share|improve this answer
+1 for: "data recovery methods for failed SSD drives are not anywhere close what they are for standard hard drives" thumbs up – 0xC0000022L Apr 3 '11 at 2:16
Also the 10,000 writes MTBF is a little concerning in general, and then of course there are garbage collection issues to look into as well. – Joseph Kern Apr 3 '11 at 2:23
@Joseph - agreed, of course people will get what's coming to them any time they implement consumer gear (like the above-mentioned SSD) in a server environment. High-end SLC SSDs are much more resilient and can stay at 100% of their rated IOPs for far more years than any sane person would keep them in production. – EEAA Apr 3 '11 at 2:30
I'm not so concerned about data recovery methods. I have never done data recovery on a physical HD anyway. Always have a backup. RAID isn't a backup method. Maybe I should get an SLC SSD instead. Or maybe I should mirror a pair of these M25s... – Tracy Reed Apr 3 '11 at 5:03
@TristanK - Tracy is right - RAID isn't backup. What is is, though, is a method to 1) reduce chances of needing to restore from backup, 2) improve disk performance and 3) keep the system up in case of a disk failure. – EEAA Apr 3 '11 at 6:35

Sure it can but the idea of replacing a system with redundancy with one that has a single point of failure surely can't be a good thing. If you feel absolutely compelled to replace the spinning drives with solid state at least keep the mirroring, which of course means two SSDs. Then, and only then, might it be called "better" in any sense that matters.

share|improve this answer

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.