Re: Reasonable maximum package size ?
-----BEGIN PGP SIGNED MESSAGE-----
On 11 Jun 2007, at 9:22 pm, Josselin Mouette wrote:
You seem to strongly believe the cheap desktop hard disk is different
from the server hard disk. This is entirely wrong. Apart from 10k and
15k rpm disks, these are all strictly the same. Only the electronics
That's not true, unfortunately. They also have different design
criteria for duty cycles, and more stringent MTBF testing
requirements. There's been a lot of assertion in this thread,
without any real data, so this post provides links to some hard data
provided by disk manufacturers.
Here follows a posting I sent to the Bioclusters mailing list a few
months ago, which includes PDF documents from Seagate showing that
their desktop products are absolutely *not* equivalent to their
Practical experience of running a petabyte of storage arrays here is
what I'm basing my opinion on, not claims for device MTBF. Besides,
MTBF is highly dependent on duty cycle. MTBF figures are meaningless
if you don't also consider the duty cycle. You need to make sure the
spindles are designed for a 24/7 duty cycle. Cheap SATA drives
normally are not.
Anyway, you asked for figures, so here we are, from a spindle
manufacturer and (later) from Microsoft (yes, I know):
Their nearline fibrechannel drives have an MTBF of 1 million hours,
24/7 duty cycle, read/write. Their desktop SATA drives have an MTBF
of 600,000 hours, but that's for an 8x5 largely read-only duty
cycle. If you abuse them by running them 24x7 read-write, I dare say
it will be considerably less. But ignoring that, by my rough
estimate, a double drive failure causing data loss of your RAID5
array using desktop SATA drives will probably happen about four times
more frequently than using fibrechannel disks. Of course, you can
buy high MTBF SATA drives but (surprise!) they cost about the same as
the fibrechannel ones.
More details here about what Seagate put in their cheap drives vs.
the expensive ones:
There are also some sobering graphs in this joint presentation from
Seagate and Microsoft (apologies for the Powerpoint):
The graph showing the probability of second disk error during RAID5
rebuild on desktop and server drives is slightly scary even for
server drives, but positively terrifying for the cheap drives. This
of course, is why we use RAID6 and a hot spare in our large Lustre
filesystems. RAID5 is simply not reliable enough, using the SATA
drives which underly our SFS servers.
As many others have said in this thread, you get cheap or reliable.
You do not get both.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (Darwin)
-----END PGP SIGNATURE-----
The Wellcome Trust Sanger Institute is operated by Genome Research
Limited, a charity registered in England with number 1021457 and a
company registered in England with number 2742969, whose registered
office is 215 Euston Road, London, NW1 2BE.