From: Niu Yawei Subject: Re: [PATCH] quota: remove dqptr_sem for scalability Date: Fri, 23 May 2014 13:22:13 +0800 Message-ID: <537EDB05.8030109@gmail.com> References: <537DD5BA.1050105@gmail.com> <537EC854.1060605@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=GB2312 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: yawei.niu@intel.com, andreas.dilger@intel.com, jack@suse.cz, lai.siyao@intel.com To: Eric Sandeen , linux-fsdevel@vger.kernel.org, linux-ext4@vger.kernel.org Return-path: In-Reply-To: <537EC854.1060605@redhat.com> Sender: linux-fsdevel-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org =D3=DA 2014/5/23 12:02, Eric Sandeen =D0=B4=B5=C0: > On 5/22/14, 5:47 AM, Niu Yawei wrote: >> There are several global locks in the VFS quota code which hurts >> performance a lot when quota accounting enabled, dqptr_sem is the ma= jor one. >> >> This patch tries to make the VFS quota code scalable with minimal ch= anges. >> >> Following tests (mdtest & dbench) were running over ext4 fs in a >> centos6.5 vm (8 cpus, 4G mem, kenrel: 3.15.0-rc5+), and the result s= hows >> the patch relieved the lock congestion a lot. >> > Just noticed this patch - FWIW, Lustre has a=20 > "quota-replace-dqptr-sem-sles11sp2.patch" that they apply: Yes, I'm Lustre developer and trying to push the patch upstream. :) > > http://git.whamcloud.com/?p=3Dfs/lustre-release.git;a=3Dblob;f=3Dlust= re/kernel_patches/patches/quota-replace-dqptr-sem-sles11sp2.patch;h=3Dc= 880dac83473f48cac96dc467ea76f64a74fe5dd;hb=3DHEAD > > which might be interesting if you're looking at this. > > (Or maybe it's doing the same thing; TBH I have not looked at > either patch, I just remembered that it existed...) > > -eric -- 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