[Beowulf] Consumer vs. Enterprise Hard Drives in Clusters
Ellis Wilson
xclski at yahoo.com
Fri Feb 13 14:50:47 PST 2009
Peter Kjellstrom wrote:
> On Thursday 29 January 2009, Bill Broadley wrote:
>> Erwan Velu wrote:
>>> Bill Broadley wrote:
>>>> I found my actual results from a WD consumer 320GB drive (single
>>>> platter) with the drive on the bench vs inside a 1U node:
>>>>
>>>> http://cse.ucdavis.edu/bill/wd320-2MB-points.png
>>>>
>>>> http://cse.ucdavis.edu/bill/wd320-2MB-points-vibration.png
>>>>
>>>> Once the WD320 (a consumer drive) was replaced with a seagate es2 (raid
>>>> edition drive):
>>>>
>>>> http://cse.ucdavis.edu/bill/seagate-es.png
>>> Hey Bill,
>>> Are we sure the only difference between both cases are only the
>>> vibrations ?
>> Nothing else changed.
>>
>>> Do you use the same disk controller in all cases ?
>> Same physical node, not just the same kind of node.
>>
>>> Does the workload is exactly the same ?
>> Yes, an idle node with a PXE installed image runs my benchmark.
>
> Interesting data indeed. Can you tell us what kind of node this was? How many
> other drives in it? Did the fans spin at normal speed?
>
> /Peter
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf
Sorry for reviving this thread...I realize it has been dead for two
weeks now but my schedule forces me to read backburner these until I
have a chance to parse through a hundred every two or three weeks.
Anyhow, my question is a little less specific than Peter's, but
basically I'm wondering whether if those of us who don't rack mount (and
thus position all our drives vertically parallel) should expect the same
degradation in performance. For instance, given a standard Beowulf
cluster of desktops with only about one drive a piece sitting on
concrete I would suspect we shouldn't see much depreciation in
performance. This of course assumes given larger desktops we've
intelligently swapped out any crappy 80mm fans for slower (and in my
mind less vibrating) 120mm fans in some form of rubberized grommet.
Ellis
More information about the Beowulf
mailing list