From owner-freebsd-bugs Sun Jul 29 15:20: 5 2001 Delivered-To: freebsd-bugs@hub.freebsd.org Received: from freefall.freebsd.org (freefall.freebsd.org [216.136.204.21]) by hub.freebsd.org (Postfix) with ESMTP id 568DB37B403 for ; Sun, 29 Jul 2001 15:20:01 -0700 (PDT) (envelope-from gnats@FreeBSD.org) Received: (from gnats@localhost) by freefall.freebsd.org (8.11.4/8.11.4) id f6TMK1a13636; Sun, 29 Jul 2001 15:20:01 -0700 (PDT) (envelope-from gnats) Received: from freefall.freebsd.org (freefall.freebsd.org [216.136.204.21]) by hub.freebsd.org (Postfix) with ESMTP id 27F4437B401 for ; Sun, 29 Jul 2001 15:18:18 -0700 (PDT) (envelope-from nobody@FreeBSD.org) Received: (from nobody@localhost) by freefall.freebsd.org (8.11.4/8.11.4) id f6TMIIo13213; Sun, 29 Jul 2001 15:18:18 -0700 (PDT) (envelope-from nobody) Message-Id: <200107292218.f6TMIIo13213@freefall.freebsd.org> Date: Sun, 29 Jul 2001 15:18:18 -0700 (PDT) From: Dirk Zoller To: freebsd-gnats-submit@FreeBSD.org X-Send-Pr-Version: www-1.0 Subject: kern/29306: Network or kernel related performance problem. Sender: owner-freebsd-bugs@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org >Number: 29306 >Category: kern >Synopsis: Network or kernel related performance problem. >Confidential: no >Severity: serious >Priority: high >Responsible: freebsd-bugs >State: open >Quarter: >Keywords: >Date-Required: >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Sun Jul 29 15:20:01 PDT 2001 >Closed-Date: >Last-Modified: >Originator: Dirk Zoller >Release: 4.3-RELEASE >Organization: Sol-3 GmbH&Co KG >Environment: FreeBSD europa.sol-3.de 4.3-RELEASE FreeBSD 4.3-RELEASE #0: Sat Apr 21 10:54:49 GMT 2001 jkh@narf.osd.bsdi.com:/usr/src/sys/compile/GENERIC i386 >Description: We're building a big client-server system. We have a test program for our fundamental message passing library. The test program can act as server or client. To run a little torture test on a system we start a number of clients simultaneously. The clients send a lot of random sized messages and the server echoes these. When I run 10 clients on FreeBSD the whole system becomes very sluggish while `top' reports the system was mostly idle. The torture test takes extremely long and even when it's finished, the system needs a while to recover from the trauma and return to normal performance. With Linux/NetBSD/WindowsNT I can run the same torture test with 50 clients with no problems. These systems stay responsive while top reports they are busy. The whole torture test finishes in the expected time. I'd say, FreeBSD has a significant kernel problem here. >How-To-Repeat: Please contact me for the test program that I will provide happily. Run my test program on a freshly installed vanilla 4.3 FreeBSD system using the precompiled GENERIC kernel. >Fix: Sorry, I'm not a kernel hacker. >Release-Note: >Audit-Trail: >Unformatted: To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-bugs" in the body of the message