status or stats reports your fps.
In TF2 the is a command which starts benchmark test, may be there are the
same commands in other servers.

2008/5/28 Chris Brunelle <[EMAIL PROTECTED]>:

> What would be an accurate way to report FPS while the system is under load?
>
>
>
> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of Nephyrin
> Zey
> Sent: Tuesday, May 27, 2008 11:06 PM
> To: Half-Life dedicated Linux server mailing list
> Subject: Re: [hlds_linux] Fwd: 1000 FPS........ but accelerated
>
> Will P4s do EM64T? :-P
>
> I've also noticed that the 'stats' command completely fails to report
> valid FPS when the system is under load - it seems to average out the
> time of the last frame over a second - if you run a windows server
> where you get the banner at the top of the server output, you'll
> notice it's different (and much more stable) than the stats command
> when under load.
>
> - Neph
>
> On Tue, May 27, 2008 at 8:24 PM, Gary Stanley <[EMAIL PROTECTED]>
> wrote:
> > At 08:46 AM 5/27/2008, =?ISO-8859-2?Q?Faustas_Bu=B9kevi=E8ius?= wrote:
> >>Judging from what I've read, I'm guessing my linux config (IIRC its
> >>2.6.16 with 1000HZ, yup, thats April 2006) is fundamentally flawed.
> >>Full hlds server (about 20 clients) uses about 30% cpu (p4 3.4 ghz
> >>with HT on) and about 12% ram (out of 1G) but fps is only stable when
> >>server is empty and only with sys_ticrate 1000 (or 10000) and
> >>-pingboost 2. As soon as several clients get onto server, number fps
> >>start fluctuating constantly in between about 100 and 600. One other
> >>~15 slot server is present, same config. Total cpu usage normally does
> >>not exceed 70%. I am not really doing anything as head admin at the
> >>moment, everything is fully automated and I haven't been on the server
> >>myself for a couple of months now, but I still can't understand where
> >>did I go wrong. Nor can I let go of the fact that I could have done
> >>better.
> >
> > Try using x64 linux. x64 uses something called a vsyscall, which is a
> > shared page to map gettimeofday() to userland, and it's going to
> > reduce the cost of timekeeping with a P4.
> > A P4 has horrible pipeline latency, and it executes syscalls about 2x
> > slower than a PIII.
> >
> > You could also make sure you are using the TSC for the timecounter
> > instead of ACPI.
> >
> >
> >
> > _______________________________________________
> > To unsubscribe, edit your list preferences, or view the list archives,
> please visit:
> > http://list.valvesoftware.com/mailman/listinfo/hlds_linux
> >
>
> _______________________________________________
> To unsubscribe, edit your list preferences, or view the list archives,
> please visit:
> http://list.valvesoftware.com/mailman/listinfo/hlds_linux
> No virus found in this incoming message.
> Checked by AVG.
> Version: 8.0.100 / Virus Database: 269.24.1/1469 - Release Date: 5/27/2008
> 1:25 PM
>
>
> _______________________________________________
> To unsubscribe, edit your list preferences, or view the list archives,
> please visit:
> http://list.valvesoftware.com/mailman/listinfo/hlds_linux
>
_______________________________________________
To unsubscribe, edit your list preferences, or view the list archives, please 
visit:
http://list.valvesoftware.com/mailman/listinfo/hlds_linux

Reply via email to