From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?utf-8?q?=C5=81ukasz_Ole=C5=9B?= Subject: Re: [Scst-devel] Tuning scst windows and sw raid5 Date: Fri, 22 Apr 2011 16:35:37 +0200 Message-ID: <201104221635.38092.lukasz.oles@open-e.com> References: <201104221608.50195.lukasz.oles@open-e.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: Sender: linux-raid-owner@vger.kernel.org To: Chris Weiss Cc: scst-devel@lists.sourceforge.net, linux-raid@vger.kernel.org List-Id: linux-raid.ids On Friday 22 of April 2011 16:19:48 you wrote: > On Fri, Apr 22, 2011 at 9:08 AM, =C5=81ukasz Ole=C5=9B wrote: > > Hi, > >=20 > > I'm trying to tune Software Raid5, SCST and Windows Initiator to ge= t best > > read/write performance. >=20 > My first recommendation is to stick with blockio due to the raid5 > write hole possibility when using software raid. >=20 > what windows version, and if not server2008, did you align the > partitions? (there's a 2003 and a 2008 R2) 2008 R2 > dd is going to use the initiator OS's cache, make sure your test file > size is 2x the ram in the initiator, if not the target too. On initiator I have 2GB RAM and I'm copying 20GB file. > Also xfs is much better at dealing with large files than ntfs. Best > apple to apples test is going to be on raw unformated partitions What this test(testing on raw partitions) can tell me? In the end I hav= e to=20 use filesystem. > I'm not familiar with BST5. I typically use iometer for both OS's. I will try it. -- Lukasz Oles -- To unsubscribe from this list: send the line "unsubscribe linux-raid" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html