From mboxrd@z Thu Jan 1 00:00:00 1970 From: Piotr =?utf-8?B?RGHFgmVr?= Subject: Re: Hammer vs Jewel librbd performance testing and git bisection results Date: Wed, 11 May 2016 15:35:04 +0200 Message-ID: <20160511133504.GB6902@predictor> References: Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from predictor.org.pl ([185.5.97.54]:45329 "EHLO predictor.org.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932543AbcEKNdA (ORCPT ); Wed, 11 May 2016 09:33:00 -0400 Content-Disposition: inline In-Reply-To: Sender: ceph-devel-owner@vger.kernel.org List-ID: To: ceph-devel@vger.kernel.org On Wed, May 11, 2016 at 08:21:18AM -0500, Mark Nelson wrote: > Hi Guys, >=20 > [..] > The gist of this is that Jewel is faster than Hammer for many random > workloads (Read, Write, and Mixed). There is one specific case > where performance degrades significantly: 64-128k sequential reads. > We couldn't find anything obviously wrong with these tests, so we > spent some time running git bisects between hammer and jewel with > the NVMe test configuration (these tests were faster to setup/run > than the HDD setup). We tested about 45 different commits with > anywhere from 1-5 samples depending on how confident the results > looked: >=20 > https://docs.google.com/spreadsheets/d/1hbsyNM5pr-ZwBuR7lqnphEd-4kQUi= d0C9eRyta3ohOA/edit?usp=3Dsharing >=20 > There are several commits of interest that have a noticeable effect > on 128K sequential read performance: >=20 > [..] > 2) https://github.com/ceph/ceph/commit/c474ee42 >=20 > This commit had a very large impact, reducing performance by another = 20-25%. https://github.com/ceph/ceph/commit/c474ee42#diff-254555dde8dcfb7fb9087= 91ab8214b92R318 I would check if temporarily forcing unique_lock_name() to return its a= rg (or other constant) would change things. If so, probably a more efficie= nt way to construct unique lock name may be in order. --=20 Piotr Da=C5=82ek branch@predictor.org.pl http://blog.predictor.org.pl -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html