Re: [SAGE] SANs, Oracle, and RAID

[John Borwick <borwicjh@wfu.edu>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Doug Hughes wrote:
| On Fri, 7 May 2004, John Borwick wrote:
[snip]
|>So, here's what I haven't found anywhere else:
|>
|>* Should RAID levels be considered differently in a SAN environment than
|>~  in a local-disk environment?  That is, does the cost of a RAID 5 write
|>get mitigated by the SAN cache?
|
| to a very large extent, yes. benchmark it yourself. Most vendors will
| give you hardware for evaluation. I find iozone and bonnie++ to be
| excellent tools.

Great!  We're on the right path so far; iozone has already produced some
useful (and attractive) graphs that have helped us figure out the
machines' raw throughput.

|>* Can we get away with one huge RAID 5 array instead of 5-7 small RAID 1
|>arrays (as far as performance goes)?
|>
|
| quite possibly. 2GB/sec is a lot of bandwidth on a single channel. Many
| vendors will give you multi-channel capability.

I was thinking the fiberchannel wouldn't be the limiting factor :)  This
weekend I learned the definition of a SAN (from SAGE's book about
backups), and so realized that my question might not have been clear.

We *will* be using dual-channel connections for redundancy.

|>* How are SANs optimized for databases, in general?
|
| The SAN is just a transport fabric like SCSI: a big pipe (with extra
| tidbits like device-device connectivity for transfers).  dual 2GB
| cards are available even now. you'll want at least a 64 bit slot
| with 64mhz performance for one of those. PCI-X cards are the standard
| spec for the HBAs.

Ah.  We will have our (64 bit) cards in the 133mhz slots, and hopefully
the SAN fabric will work fine.

I guess I meant how the disks on the back-end are typically
configured/optimized for databases.  One person wrote me and mentioned
they have two RAID 5 arrays configured for a given database, with
certain files guaranteed to be on each array.

| More spindles is a good approach. Oracle likes to recommend using
| a 1MB interleave factor and stripe and mirror everything to better utilize
| all disks and eliminate hotspots (aka SAME - Stripe and Mirror
| Everything). Sun calls the technique Wide-Thin striping (white paper
| at www.sun.com/blueprints)

Thank you!  I found
http://otn.oracle.com/deploy/availability/pdf/oow2000_same.pdf , which
looks A LOT like a good guiding document.

| Are you optimizing for number of IO operations per second like a TPC
| or are you optimizing for sequential throughput? (or something in
| between). IOPS wants as many fast spindles as possible with minimal
| prefetching.
| Sequential throughput doesn't require as many spindles but likes higher
| RPM disks and lots of pre-fetching.

I will go to the DBAs and see if they can give us more specific
requirements about actual performance (in terms of random and sequential
reads/sec, writes/sec).

- --
~ John Borwick, Systems Administrator | work                 336 758 2507
~       Wake Forest University        | cell                 336 391 6623
~       Winston-Salem, NC, USA        | web  http://www.wfu.edu/~borwicjh
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (FreeBSD)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFAn3o5j7+4jlbWCHIRAsnTAJ0aQXHkjkGdx2qVOwsyb5ZwQjYNhACeOs08
DbzKkZWaV0NbJ9eqf5TQRbk=
=z1Fq
-----END PGP SIGNATURE-----