[Beowulf] Big storage
gerry.creager at tamu.edu
Wed Apr 16 06:24:43 PDT 2008
We've had two fail rather randomly. The failures did cause disk
corruption but it wasn't an undetected/undetectable sort. They started
throwing errors to syslog, then fell over and stopped accessing disks.
Bruce Allen wrote:
> Hi Gerry,
> So far the only problem we have had is with one Areca card that had a
> bad 2GB memory module. This generated lots of (correctable) single bit
> errors but eventually caused real problems. Could you say something
> about the reliability issues you have seen?
> On Wed, 16 Apr 2008, Gerry Creager wrote:
>> We've used AoE (CoRAID hardware) with pretty good success (modulo one
>> RAID shelf fire that was caused by a manufacturing defect and dealt
>> with promptly by CoRAID). We've had some reliability issues with
>> Areca cards but no data corruption on the systems we've built that way.
>> Bruce Allen wrote:
>>> Hi Xavier,
>>>>>>> PPS: We've also been doing some experiments with putting
>>>>>>> OpenSolaris+ZFS on some of our generic (Supermicro + Areca)
>>>>>>> 16-disk RAID systems, which were originally intended to run Linux.
>>>>>> I think that DESY proved some data corruption with such
>>>>>> configuration, so they switched to OpenSolaris+ZFS.
>>>>> I'm confused. I am also talking about OpenSolaris+ZFS. What did
>>>>> DESY try, and what did they switch to?
>>>> Sorry, I am indeed not clear. As far as I know, DESY found data
>>>> corruption using Linux and Areca cards. They moved from linux to
>>>> OpenSolaris and ZFS, avoiding other corruption. This has been
>>>> discussed in HEPiX storage workgroup. However, I can not speak on
>>>> their behalf at all. I'll try to get you in touch with someone more
>>>> aware of this issue, as my statements lack of figures.
>>> I think that would be very interesting to the entire Beowulf mailing
>>> list, so please suggest that they respond to the entire group, not
>>> just to me personally. Here is an LKML thread about silent data
>>> So far we have not seen any signs of data corruption on Linux+Areca
>>> systems (and our data files carry both internal and external
>>> checksums, so we would be sensitive to this).
>>> Beowulf mailing list, Beowulf at beowulf.org
>>> To change your subscription (digest mode or unsubscribe) visit
Gerry Creager -- gerry.creager at tamu.edu
Texas Mesonet -- AATLT, Texas A&M University
Cell: 979.229.5301 Office: 979.458.4020 FAX: 979.862.3983
Office: 1700 Research Parkway Ste 160, TAMU, College Station, TX 77843
More information about the Beowulf