Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751283AbWI3RbK (ORCPT ); Sat, 30 Sep 2006 13:31:10 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751317AbWI3RbK (ORCPT ); Sat, 30 Sep 2006 13:31:10 -0400 Received: from mga03.intel.com ([143.182.124.21]:21640 "EHLO mga03.intel.com") by vger.kernel.org with ESMTP id S1751283AbWI3RbI convert rfc822-to-8bit (ORCPT ); Sat, 30 Sep 2006 13:31:08 -0400 X-ExtLoop1: 1 X-IronPort-AV: i="4.09,240,1157353200"; d="scan'208"; a="125253466:sNHT25090009" X-MimeOLE: Produced By Microsoft Exchange V6.5 Content-class: urn:content-classes:message MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8BIT Subject: RE: Postal 56% waits for flock_lock_file_wait Date: Sat, 30 Sep 2006 21:26:40 +0400 Message-ID: X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: Postal 56% waits for flock_lock_file_wait Thread-Index: Acbkof3Lqy94qeO2QhKIq8bZeBW9DQAEGifw From: "Ananiev, Leonid I" To: "Trond Myklebust" Cc: "Linux Kernel Mailing List" X-OriginalArrivalTime: 30 Sep 2006 17:30:44.0438 (UTC) FILETIME=[28AF0F60:01C6E4B6] Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2251 Lines: 60 > On which filesystem were the above results obtained if it was not ext2? The default ext3 fs was used. > All the above results are telling you is that your test involves several > processes contending for the same lock, and so all of them barring the > one process that actually holds the lock are idle. Yes. It is flock_lock_file_wait. Leonid -----Original Message----- From: Trond Myklebust [mailto:trond.myklebust@fys.uio.no] Sent: Saturday, September 30, 2006 7:06 PM To: Ananiev, Leonid I Cc: Linux Kernel Mailing List Subject: Re: Postal 56% waits for flock_lock_file_wait On Sat, 2006-09-30 at 09:25 +0400, Ananiev, Leonid I wrote: > A benchmark > 'postal -p 16 localhost list_of_1000_users' > 56% of run time waits for flock_lock_file_wait; > Vmstat reports that 66% cpu is idle and vmstat bi+bo=3600 (far from > max). > Postfix server with FD_SETSIZE=2048 was used. > Similar results got for sendmail. > Wchan is counted by > while :; do > ps -o user,wchan=WIDE-WCHAN-COLUMN,comm; > sleep 1; > done | awk '/ postfix /{a[$2]++}END{for (i in a) print > a[i]"\t"i}' > If ext2 fs is used the Postal throughput is twice more and bi+bo by 50% > less while flock_lock_file_wait 60% still. On which filesystem were the above results obtained if it was not ext2? > Is flock_lock_file_wait considered as a performance limiting waiting for > similar applications in smp? All the above results are telling you is that your test involves several processes contending for the same lock, and so all of them barring the one process that actually holds the lock are idle. As for the throughput issue, that really depends on the filesystem you are measuring. For remote filesystems like NFS, locks can _really_ slow down performance because they are often required to flush all dirty data to disk prior to releasing the lock (so that it becomes visible to processes on other clients that might subsequently obtain the lock). Cheers, Trond - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/