From: Christian Ehrhardt Subject: Re: [PATCH 0/8] Reduce latencies and improve overall reclaim efficiency v2 Date: Wed, 03 Nov 2010 11:50:35 +0100 Message-ID: <4CD13E7B.5090804@linux.vnet.ibm.com> References: <1284553671-31574-1-git-send-email-mel@csn.ul.ie> <4CB721A1.4010508@linux.vnet.ibm.com> <20101018135535.GC30667@csn.ul.ie> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Andrew Morton , linux-mm@kvack.org, linux-fsdevel@vger.kernel.org, Linux Kernel List , Johannes Weiner , Minchan Kim , Wu Fengguang , KAMEZAWA Hiroyuki , KOSAKI Motohiro To: Mel Gorman Return-path: Received: from mtagate7.uk.ibm.com ([194.196.100.167]:54416 "EHLO mtagate7.uk.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754344Ab0KCKuo (ORCPT ); Wed, 3 Nov 2010 06:50:44 -0400 In-Reply-To: <20101018135535.GC30667@csn.ul.ie> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On 10/18/2010 03:55 PM, Mel Gorman wrote: > On Thu, Oct 14, 2010 at 05:28:33PM +0200, Christian Ehrhardt wrote: >=20 >> Seing the patches Mel sent a few weeks ago I realized that this seri= es >> might be at least partially related to my reports in 1Q 2010 - so I = ran my >> testcase on a few kernels to provide you with some more backing data= =2E >=20 > Thanks very much for revisiting this. >=20 >> Results are always the average of three iozone runs as it is known t= o be somewhat noisy - especially when affected by the issue I try to sh= ow here. >> As discussed in detail in older threads the setup uses 16 disks and = scales the number of concurrent iozone processes. >> Processes are evenly distributed so that it always is one process pe= r disk. >> In the past we reported 40% to 80% degradation for the sequential re= ad case based on 2.6.32 which can still be seen. >> What we found was that the allocations for page cache with GFP_COLD = flag loop a long time between try_to_free, get_page, reclaim as free ma= kes some progress and due to that GFP_COLD allocations can loop and ret= ry. >> In addition my case had no writes at all, which forced congestion_wa= it to wait the full timeout all the time. >> >> Kernel (git) 4 8 16 deviation #= 16 case comment >> linux-2.6.30 902694 1396073 1892624 = base base >> linux-2.6.32 752008 990425 932938 = -50.7% impact as reported in 1Q 2010 >> linux-2.6.35 63532 71573 64083 = -96.6% got even worse >> linux-2.6.35.6 176485 174442 212102 = -88.8% fixes useful, but still far away >> linux-2.6.36-rc4-trace 119683 188997 187012 = -90.1% still bad >> linux-2.6.36-rc4-fix 884431 1114073 1470659 = -22.3% Mels fixes help a lot! >> [...] > If all goes according to plan, > kernel 2.6.37-rc1 will be of interest. Thanks again. Here a measurement with 2.6.37-rc1 as confirmation of progress: linux-2.6.37-rc1 876588 1161876 1643430 = -13.1% even better than 2.6.36-fix That means 2.6.37-rc1 really shows what we hoped for. And it eventually even turned out a little bit better than 2.6.36 + you= r fixes. =20 --=20 Gr=FCsse / regards, Christian Ehrhardt IBM Linux Technology Center, System z Linux Performance=20 -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel= " in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html