Yes.  But remember my testes we MBS throughput (backups) not IO (database).
I think the results would be the similar, but I have not been able to
perform DB IOS tests.  Someone also just reminded me that we have a mixture
of SATA and FC in our 6140 arrays thus resulting in having to put the 6140
backend in 2GBs and not 4GBs mode.  I believe the newer SATA drives don't
impose this limitation.


The following is what I remember from a discussion with sun and may not be
entirely accurate.

The SAS backend on the 2540 allows 4 point to point connections per
controller (a switched backed). So two 3+1s one assigned on each controller
can transfer data to and from each drive simultaneously in the RAID set.  In
a Fibre Channel array with a backend arbitrated loop, you have to arbitrate
for each drive to talk to the controller and I believe that there are two
loops, allowing for only two drive to talk to the controller(s) at a time.
So depending on the RAID set size arbitration occurs a number of times with
a single RAID set.


--
mike cannon
[EMAIL PROTECTED]
864.650.2577 (cell)
864.656.3809 (office)

computing & information technology
340 computer court
anderson, sc 29625


> From: Justin Vassallo <[EMAIL PROTECTED]>
> Organization: Ixaris Systems Ltd
> Date: Tue, 8 Jul 2008 15:40:21 +0200
> To: <[email protected]>
> Subject: Re: [storage-discuss] 2x2540 FC vs 1x6140FC
> 
> Mike, Just to double check i am understanding correctly...
> 
> In your tests, the 6140 (with 4G of cache) was slower than the 2540 (w 1G
> cache)? 
> 
> _______________________________________________
> storage-discuss mailing list
> [email protected]
> http://mail.opensolaris.org/mailman/listinfo/storage-discuss

_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to