From nobody@FreeBSD.org Tue Jul 18 02:42:44 2006 Return-Path: Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 21D9816A4DA for ; Tue, 18 Jul 2006 02:42:44 +0000 (UTC) (envelope-from nobody@FreeBSD.org) Received: from www.freebsd.org (www.freebsd.org [216.136.204.117]) by mx1.FreeBSD.org (Postfix) with ESMTP id A9AA243D4C for ; Tue, 18 Jul 2006 02:42:43 +0000 (GMT) (envelope-from nobody@FreeBSD.org) Received: from www.freebsd.org (localhost [127.0.0.1]) by www.freebsd.org (8.13.1/8.13.1) with ESMTP id k6I2ghI7051098 for ; Tue, 18 Jul 2006 02:42:43 GMT (envelope-from nobody@www.freebsd.org) Received: (from nobody@localhost) by www.freebsd.org (8.13.1/8.13.1/Submit) id k6I2ghLE051097; Tue, 18 Jul 2006 02:42:43 GMT (envelope-from nobody) Message-Id: <200607180242.k6I2ghLE051097@www.freebsd.org> Date: Tue, 18 Jul 2006 02:42:43 GMT From: Angus Wang To: freebsd-gnats-submit@FreeBSD.org Subject: Maybe the 6.0 Release is not much more than the 5.3 Release on file system. X-Send-Pr-Version: www-2.3 >Number: 100460 >Category: i386 >Synopsis: Maybe the 6.0 Release is not much more than the 5.3 Release on file system. >Confidential: no >Severity: serious >Priority: high >Responsible: freebsd-i386 >State: closed >Quarter: >Keywords: >Date-Required: >Class: maintainer-update >Submitter-Id: current-users >Arrival-Date: Tue Jul 18 02:50:13 GMT 2006 >Closed-Date: Sat Jul 29 07:49:22 GMT 2006 >Last-Modified: Sat Jul 29 07:49:22 GMT 2006 >Originator: Angus Wang >Release: 6.0 Realease & 5.3 Release >Organization: >Environment: The Server with two Intel 2.0 CPU,512M*4 Memory. >Description: I used the iozone3_257, bonnie++-1.03a and iobench to test the 6.0 Realease and 5.3 Release. The test contents local, sync, async, soft updates and sync+soft updates to load the filesystem, then I found the results is amazed me. When I used the iozone to test the performence of read, reread, write and rewrite. It shows the read curve is no difference between the two system. but in the write curve the 6.0's sometimes is bad than 5.3's. The other two soft tested shows the same resuls. Maybe there is something wrong during my testing, please explains for me. Thanks! >How-To-Repeat: ./iozone -A -n 4m -g 2G -R -b XXXX.xls ./tiobench.pl > XXXX.txt ./bonnie++ -u root -s 4G > XXX.txt >Fix: >Release-Note: >Audit-Trail: State-Changed-From-To: open->closed State-Changed-By: linimon State-Changed-When: Sat Jul 29 07:44:32 UTC 2006 State-Changed-Why: The 5.3 release does not have some of the improvements that were later made to 5.4 and 5.5. The 6.0 release is missing some improvements made in the 6.1 release. Kris Kenneway's tests, as presented at BSDCan, tended to indicate that the latest codebase in 6-STABLE was outperforming not only 5-STABLE but also 4-STABLE. Again, that was with the latest code as available at the time. In general issues like this should be discussed on freebsd-questions or freebsd-hackers. The PR database is best used to flag very specific bugs in very specific code. The PRs that are submitted with performance problems generally tend to get lost in the large quantity of the former. http://www.freebsd.org/cgi/query-pr.cgi?pr=100460 >Unformatted: