linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jim Schatzman <James.Schatzman@fulab.com>
To: linux-raid@vger.kernel.org
Subject: Re: HBA Adaptor advice
Date: Sun, 22 May 2011 20:11:54 -0600	[thread overview]
Message-ID: <20110523021234.1BAEAF9C03C@mail.fulab.com> (raw)

One more input re. using cheap drives. 

I have been running about 20 Western Digital "Green" and "Enterprise" drives (half are 1.5 TB and half are 2 TB) for several years in Raid-5 and Raid-6 configurations (all linux md). They are up 24x7. When they first came on the market, about 30% of my new drives failed within 3 months of operation (about equal fractions of Green and Enterprise). Overall, 50% of the drives eventually failed - 35% of the Green drives and 100% of the Enterprise drives. In the past 18 months, one has failed (an Enterprise drive). That drive was a warranty replacement for an earlier Enterprise drive failure.

My impression is that Western was having some quality control issues with the 2GB drives - both Green and Enterprise. This was very annoying. It appears that quality has improved. I never lost any data nor ever had to restore from backup because I was always able to replace the bad drive and rebuild the raid without difficulty that I could not get solved through this forum. 

My experience suggests that the WD Enterprise class drives were an unnecessary expense, at least as far as reliability is concerned.

Would I recommend cheap SATA drives for mission critical data?  Absolutely not. I wouldn't recommend  any  SATA drives. Go with the most  expensive SAS drives available. For that matter, I have loads of SCSI drives that are still going fine after 5 to 10 years of 365x24x7 operation.

If you are going to build a RAID from cheap drives, expect that part of your hardware savings will be compensated for by labor costs. Run smart checks often. Also, and I cannot emphasize this enough, make certain that everything attached to the RAID is plugged into a high quality UPS. Otherwise, you are just asking for a power spike to take out multiple drives and/or the controller and to lose data.

Jim


             reply	other threads:[~2011-05-23  2:11 UTC|newest]

Thread overview: 60+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-05-23  2:11 Jim Schatzman [this message]
2011-05-23  3:39 ` HBA Adaptor advice Tobias McNulty
2011-05-23 10:42   ` Ed W
  -- strict thread matches above, loose matches on Subject: below --
2011-05-23 11:14 HBA Adaptor Advice Ed W
2011-05-23 11:55 ` Joe Landman
2011-05-19 12:26 HBA Adaptor advice Ed W
2011-05-19 12:36 ` Roman Mamedov
2011-05-19 12:43   ` Mathias Burén
2011-05-19 14:06 ` Michael Sallaway
2011-05-19 19:10 ` Thomas Harold
2011-05-19 21:12   ` Rudy Zijlstra
2011-05-19 21:07 ` Brad Campbell
2011-05-20 20:58   ` Tobias McNulty
2011-05-20 21:23     ` Brad Campbell
2011-05-20  2:08 ` Andy Smith
2011-05-20  5:30   ` Stan Hoeppner
2011-05-21  9:52     ` Ed W
2011-05-20  7:33   ` Ed W
2011-05-20 10:21     ` Stan Hoeppner
2011-05-21 11:17       ` Ed W
2011-05-21 11:29         ` Rudy Zijlstra
2011-05-21 11:54           ` Ed W
2011-05-21 17:37             ` Leslie Rhorer
2011-05-22  9:41             ` Stan Hoeppner
2011-05-22 10:03               ` Rudy Zijlstra
2011-05-23  9:32                 ` Ed W
2011-05-21 17:05           ` Leslie Rhorer
2011-05-22  9:04         ` Stan Hoeppner
2011-05-22 10:09           ` Brad Campbell
2011-05-22 19:25             ` Stan Hoeppner
2011-05-22 20:57               ` Tobias McNulty
2011-05-22 21:13                 ` Johannes Truschnigg
2011-05-23  9:48                   ` Ed W
2011-05-23 10:44                     ` John Robinson
2011-05-22 23:19               ` Brad Campbell
2011-05-23  4:09                 ` Roman Mamedov
2011-05-23  5:54                   ` Brad Campbell
2011-05-23  6:08                     ` Roman Mamedov
2011-05-23 10:42                     ` Stan Hoeppner
2011-05-23 11:35                       ` David Brown
2011-05-23  6:54                 ` Stan Hoeppner
2011-05-23  7:23                   ` Brad Campbell
2011-05-22 23:44               ` Brad Campbell
2011-05-23  0:07                 ` Brad Campbell
2011-05-23  5:30                   ` Stefan /*St0fF*/ Hübner
2011-05-23 10:18                     ` Ed W
2011-05-23  9:58                 ` Stan Hoeppner
2011-05-23 10:33                   ` Ed W
2011-05-23 11:21                     ` Stan Hoeppner
2011-05-20 12:18     ` Joe Landman
2011-05-20 12:34       ` Roman Mamedov
2011-05-20 12:36         ` Mathias Burén
2011-05-20 12:48         ` Joe Landman
2011-05-20 13:21       ` Ed W
2011-05-20 14:23         ` Joe Landman
2011-05-20 20:01       ` Andy Smith
2011-05-20 20:12         ` Stan Hoeppner
2011-05-20 20:24         ` Drew
2011-05-20 20:58           ` Stan Hoeppner
     [not found]             ` <4DD7A100.2010807@wildgooses.com>
2011-05-22  8:13               ` Stan Hoeppner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20110523021234.1BAEAF9C03C@mail.fulab.com \
    --to=james.schatzman@fulab.com \
    --cc=linux-raid@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).