From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 1575 invoked by uid 1002); 9 Apr 2003 13:11:05 -0000 Mailing-List: contact gentoo-dev-help@gentoo.org; run by ezmlm Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-Id: Gentoo Linux mail X-BeenThere: gentoo-dev@gentoo.org Received: (qmail 10313 invoked from network); 9 Apr 2003 13:11:04 -0000 Date: Wed, 9 Apr 2003 15:11:03 +0200 From: Eske Christiansen To: gentoo-dev@gentoo.org Message-ID: <20030409131103.GA30445@diku.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.4i Subject: [gentoo-dev] preformance in gentoo X-Archives-Salt: de99c992-6adc-4c2b-b1ad-a5cc65a8a2d6 X-Archives-Hash: 9422949d3fa036a60869e2ba56fab99d Hi all I was building a new cluster of dell optiplexs n-series(P4, 512 Mb ram, 120Gb disk) to descide on which linux dist to use I ran a som tests to see which suited my needs (it was a test to see if a std gentoo install (not many CFLAGS) was better than minimum redhat install). I was very superised to learn that when running lmbench gentoo version 1.4rc3 (upto date, march=pentium3) was slower in system calls than redhat 8.0 and 9.0. infact tcp select was 5 times slower om gentoo. Does any know what redhat have done to there kernel? Is it a question on the module ratio in the kernel or du you have any ideers. I like gentoo but 5 times slower in tcp select must be fixed (if any can lead me in the right direction I will gladly look at it.). The some of test results (the change in hostname is dhcp all run on the same hardware) : gentoo: Processor, Processes - times in microseconds - smaller is better ------------------------------------------------------------------------------ Host OS Mhz null null open slct sig sig fork exec sh call I/O stat clos TCP inst hndl proc proc proc --------- ------------- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- amigos41. Linux 2.4.20- 2 0.64 0.89 2.84 4.26 28.3 0.93 3.27 155. 520. 2140 amigos41. Linux 2.4.20- 2 0.65 0.90 2.90 4.38 28.3 0.94 3.24 155. 524. 2156 amigos41. Linux 2.4.20- 2 0.64 0.90 2.83 4.30 27.6 0.94 3.28 153. 528. 2154 amigos41. Linux 2.4.20- 2 0.64 0.90 2.85 4.29 27.7 0.94 3.36 153. 545. 2153 amigos41. Linux 2.4.20- 2 0.64 0.90 2.86 4.32 27.7 0.94 3.28 155. 516. 2152 amigos41. Linux 2.4.20- 2 0.64 0.90 2.83 4.31 27.6 0.93 3.28 156. 524. 2156 amigos41. Linux 2.4.20- 2 0.64 0.90 2.83 4.30 28.3 0.94 3.28 156. 522. 2163 amigos41. Linux 2.4.20- 2 0.64 0.89 2.84 4.29 28.3 0.93 3.24 155. 516. 2149 amigos41. Linux 2.4.20- 2 0.64 0.90 2.84 4.33 28.3 0.93 3.25 154. 519. 2148 amigos41. Linux 2.4.20- 2 0.64 0.90 2.83 4.31 28.3 0.94 3.25 154. 529. 2152 redhat (redhat 8.0, the redhat 9.0 shows the same numbers): Processor, Processes - times in microseconds - smaller is better ------------------------------------------------------------------------------ Host OS Mhz null null open slct sig sig fork exec sh call I/O stat clos TCP inst hndl proc proc proc --------- ------------- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- amigos42. Linux 2.4.18- 2 0.55 0.62 1.56 2.32 6.34 0.92 3.16 119. 432. 2218 amigos42. Linux 2.4.18- 2 0.55 0.61 1.56 2.27 6.37 0.91 3.22 120. 428. 2213 amigos42. Linux 2.4.18- 2 0.55 0.61 1.57 2.31 6.33 0.91 3.13 122. 430. 2237 amigos42. Linux 2.4.18- 2 0.55 0.61 1.59 2.28 6.59 0.90 3.19 119. 441. 2214 amigos42. Linux 2.4.18- 2 0.55 0.63 1.55 2.28 6.65 0.91 3.17 121. 444. 2226 amigos42. Linux 2.4.18- 2 0.55 0.61 1.56 2.31 6.43 0.90 3.22 125. 431. 2223 amigos42. Linux 2.4.18- 2 0.55 0.61 1.58 2.29 6.64 0.90 3.25 121. 446. 2222 amigos42. Linux 2.4.18- 2 0.55 0.61 1.58 2.30 6.23 0.90 3.20 120. 431. 2215 amigos42. Linux 2.4.18- 2 0.55 0.61 1.56 2.26 6.24 0.90 3.13 123. 441. 2213 amigos42. Linux 2.4.18- 2 0.55 0.61 1.53 2.27 6.58 0.91 3.22 118. 434. 2216 yours Eske -- gentoo-dev@gentoo.org mailing list