Increasing CallSignalHandlerNumber/RtpHandlerNumber/disabling logging may give better results. ----- Original Message ----- From: "Sergio Lenzi" <enigma@xxxxxxxxxxxxxxxxx> Sent: Tuesday, August 17, 2004 9:50 AM > On Tuesday 17 August 2004 02:46, Budiman Indra wrote: > > Dear Sergio, > > > > What I meant is I run gnugk on Dual Pentium 850 on Redhat Linux Desktop Ed with 2 G Ram memory. Thus with a call generation of 120 calls simultanoes session and callgenerator running on the same machine are using up to 90 % CPU Usage. > > > > However, If you dont mind sharing , how do i see if the system is runnig out of cpu cycles > > or is woking on interrupts ? > > > > Thank you and regards, > > > hummmm..... > in reality you have the gnugk running on 850mhz... that is each cpu can process only on > 850mhz... and I do not know (guarantee) that the LInux kernel can fire threads > on both cpus... if gnugk was multitask than it is easy to see what is happening > I use the "top" command... it shows the various process running on the 2 cpus... > Please note that a pentium 4 HT has 2 cpus on the same chip... > for example: > ==============one cpu cyrix/via Samuel C3============= > last pid: 66865; load averages: 0.23, 0.17, 0.14 up 15+01:28:47 04:41:43 > 90 processes: 4 running, 86 sleeping > CPU states: 13.2% user, 0.0% nice, 2.6% system, 0.0% interrupt, 84.2% idle > Mem: 215M Active, 33M Inact, 105M Wired, 15M Cache, 47M Buf, 816K Free > Swap: 800M Total, 417M Used, 382M Free, 52% Inuse > > PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND > 7140 root 76 0 192M 35364K RUN 602:22 9.57% 9.57% XFree86 > 66862 lenzi 96 0 32720K 19792K RUN 0:02 8.72% 5.18% kdeinit > 61307 lenzi 96 0 55848K 42940K select 4:15 1.61% 1.61% kmail > 81041 lenzi 96 0 34124K 12580K select 98:57 0.88% 0.88% kdeinit > 479 root 96 0 1264K 68K select 20:11 0.15% 0.15% moused > 7241 lenzi 96 0 37544K 12252K select 104:58 0.10% 0.10% kdeinit > 7245 lenzi 76 0 32560K 6700K RUN 17:00 0.10% 0.10% kdeinit > 7227 lenzi 96 0 49868K 15280K select 290:21 0.05% 0.05% kdeinit > 70621 lenzi 96 0 21148K 3380K select 64:16 0.00% 0.00% artsd > 7255 lenzi 96 0 31236K 10408K select 59:10 0.00% 0.00% kdeinit > 7247 lenzi 96 0 48420K 17644K select 46:06 0.00% 0.00% kdeinit > 276 root 96 0 1552K 408K select 26:15 0.00% 0.00% natd > 14404 lenzi 96 0 31348K 6744K select 18:44 0.00% 0.00% kdeinit > 7251 lenzi 96 0 30216K 11068K select 16:58 0.00% 0.00% kdeinit > 81058 lenzi 96 0 30692K 6328K select 12:31 0.00% 0.00% kdeinit > 7221 lenzi 96 0 23232K 2456K select 11:58 0.00% 0.00% kdeinit > 726 root 96 0 3800K 836K select 3:59 0.00% 0.00% nmbd > > =====================2 cpu or HT ============================ > last pid: 98724; load averages: 0.00, 0.00, 0.00 up 7+06:00:22 04:42:52 > 39 processes: 3 running, 36 sleeping > CPU states: 0.0% user, 0.0% nice, 0.0% system, 0.0% interrupt, 100% idle > Mem: 58M Active, 295M Inact, 84M Wired, 22M Cache, 60M Buf, 34M Free > Swap: 2000M Total, 564K Used, 1999M Free > > PID USERNAME PRI NICE SIZE RES STATE C TIME WCPU CPU COMMAND > 98722 root 8 0 1668K 1216K wait 1 0:00 1.35% 0.24% login > 611 root 96 0 18512K 4932K select 0 2:12 0.00% 0.00% gnugk > 11796 squid 96 0 16436K 14564K select 0 0:46 0.00% 0.00% squid > 28962 root 96 0 3604K 2176K select 0 0:46 0.00% 0.00% ppp > 23270 root 96 0 3636K 2024K select 0 0:23 0.00% 0.00% nmbd > 14404 root 96 0 3556K 2032K select 0 0:12 0.00% 0.00% sendmail > 313 root 96 0 1348K 820K select 0 0:11 0.00% 0.00% syslogd > 591 pgsql 96 0 91956K 2696K select 1 0:08 0.00% 0.00% postgres > 672 pgsql 96 0 5604K 2060K select 1 0:03 0.00% 0.00% pg_autov > 369 root 96 0 2652K 2272K select 0 0:02 0.00% 0.00% ypbind > 559 root 8 0 1388K 960K nanslp 0 0:02 0.00% 0.00% cron > 658 root 96 0 1496K 1104K select 0 0:02 0.00% 0.00% inetd > 356 root 96 0 2064K 1672K select 0 0:02 0.00% 0.00% ypserv > 594 pgsql 96 0 8756K 2616K select 0 0:02 0.00% 0.00% postgres > 324 bind 96 0 2860K 2148K select 0 0:01 0.00% 0.00% named > 97397 root 76 0 3604K 2224K RUN 0 0:01 0.00% 0.00% ppp > 593 pgsql 96 0 9744K 3548K select 1 0:01 0.00% 0.00% postgres > > ================================================================ > note that in the 2 cpu model, thre is a small column C between the columns > STATE and TIME... > in my case (FreeBSD) the OS splits load between the tasks... some tasks > are in cpu 0, others are in cpu 1. > > Note also that gnugk is in cpu 0 and it does not split between cpus... > so in this case is better... and cheaper to put gnugk on a faster AMD cpu > like an athlon xp 2.8ghz... than in a dual 850... > > > On the other side, postgres (that is multitask) runs better on the dual.. > > Sergio ------------------------------------------------------- SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media 100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33 Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift. http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285 _______________________________________________________ List: Openh323gk-users@xxxxxxxxxxxxxxxxxxxxx Archive: http://sourceforge.net/mailarchive/forum.php?forum_id=8549 Homepage: http://www.gnugk.org/
