Article View: borland.public.cppbuilder.nativeapi
Article #11973Poor thread performance on multi-CPU machines
From: Arnie Mauer
Date: Fri, 09 Jan 2004 19:50
Date: Fri, 09 Jan 2004 19:50
14 lines
1467 bytes
1467 bytes
Sorry if this is a bit vague, but was wondering if anyone else has run into it. We have a service app that handles input requests from a TCP/IP port on one end, does some business logic, sends a request to a DB server, gets info and sends it out again. Both the front and backend of the app uses thread pools and I/O completion is used for the comm end. The threads use critical sections where necessary. It turns out that running the app on a single CPU desktop achieves higher throughput than running on a 2 or 4 CPU server. Does anyone have any idea why this might be? Thanks, - Arnie
Technical Details
Message-ID:
<3fff05ce$1@newsgroups.borland.com>Path:
arc1-ca.pugleaf.net!archiver1.google.com!news1.google.com!news2.google.com!fu-berlin.de!news.linkpendium.com!newsfeeds.ihug.co.nz!ihug.co.nz!newsgroups.borland.com!not-for-mailShow Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 10cc67,6aabc16a59635b8c,start X-Google-Attributes: gid10cc67,public X-Google-ArrivalTime: 2004-01-09 12:09:12 PST From: "Arnie Mauer" <xxx@nowhere.net> Newsgroups: borland.public.cppbuilder.nativeapi Subject: Poor thread performance on multi-CPU machines Date: Fri, 9 Jan 2004 14:50:33 -0500 Lines: 16 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1158 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1165 NNTP-Posting-Host: 24.151.21.222 Message-ID: <3fff05ce$1@newsgroups.borland.com> X-Trace: newsgroups.borland.com 1073677774 24.151.21.222 (9 Jan 2004 11:49:34 -0700) Path: arc1-ca.pugleaf.net!archiver1.google.com!news1.google.com!news2.google.com!fu-berlin.de!news.linkpendium.com!newsfeeds.ihug.co.nz!ihug.co.nz!newsgroups.borland.com!not-for-mail