RE: [hlds_linux] Crazy Usage
How do you work that out ? -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Matthew Donnon Sent: 22 September 2003 10:54 To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Crazy Usage halving the CPU mhz is not an accurate way of representing a HT system the virtual CPU simply indicates the unused pipelines and math calc units. which would make it 1176 mhz Matt - Original Message - From: "James Sykes" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Monday, September 22, 2003 2:09 AM Subject: RE: [hlds_linux] Crazy Usage > Just lookin at a full 1.5 16 player server running de_dust2. > Note : this is on a dual xeon 2.4 with HT, no other servers running) > The CPU figures are of 1 of the 4 virtual processors. (eg. 1.2ghz) > TOTAL server usage is about 10-12% > > 49.3 > 63.4 > 62.2 > 57.2 > 54.2 > 59.6 > 57.9 > > 588mhz CPU. > > > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of raoul > bhatia > Sent: 21 September 2003 17:05 > To: [EMAIL PROTECTED] > Subject: Re: [hlds_linux] Crazy Usage > > would you care posting your 1.5 results with these two test machines > too? > > would be nice to see the actual differences there too. > > James Sykes wrote: > > > That's not the point. > > This difference is MUCH bigger in 1.6 that it was in 1.5 > > In 1.5 I saw 100-150mhz difference - no more. > > > ___ > 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 > ___ 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
Re: [hlds_linux] Crazy Usage
Matthew Donnon wrote: halving the CPU mhz is not an accurate way of representing a HT system the virtual CPU simply indicates the unused pipelines and math calc units. which would make it 1176 mhz Not to mention that he's referring to a Xeon, not even a regular P4. -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
halving the CPU mhz is not an accurate way of representing a HT system the virtual CPU simply indicates the unused pipelines and math calc units. which would make it 1176 mhz Matt - Original Message - From: "James Sykes" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Monday, September 22, 2003 2:09 AM Subject: RE: [hlds_linux] Crazy Usage > Just lookin at a full 1.5 16 player server running de_dust2. > Note : this is on a dual xeon 2.4 with HT, no other servers running) > The CPU figures are of 1 of the 4 virtual processors. (eg. 1.2ghz) > TOTAL server usage is about 10-12% > > 49.3 > 63.4 > 62.2 > 57.2 > 54.2 > 59.6 > 57.9 > > 588mhz CPU. > > > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of raoul > bhatia > Sent: 21 September 2003 17:05 > To: [EMAIL PROTECTED] > Subject: Re: [hlds_linux] Crazy Usage > > would you care posting your 1.5 results with these two test machines > too? > > would be nice to see the actual differences there too. > > James Sykes wrote: > > > That's not the point. > > This difference is MUCH bigger in 1.6 that it was in 1.5 > > In 1.5 I saw 100-150mhz difference - no more. > > > ___ > 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 > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
Unfortunatly My cs 1.6 steam 16 player server on a p4 with glibc2.2.5 on debian uses 40 - 60 % CPU So its not a glibc issue for me The kernels 2.4.22 with default HT - Original Message - From: "raoul bhatia" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Monday, September 22, 2003 6:13 AM Subject: Re: [hlds_linux] Crazy Usage > i get the feeling, that the glibc-2.2.x does something better. somehow, > most ppl using 2.2.x report normal cpu usage, while the others with > 2.3.x report high usage + lags. > > or am i wrong? > > best regards, > raoul bhatia > > bad ping wrote: > > >>can you support a little more information? > >>kernel, glibc, libpthread, ... version > >>would be nice to know. > >> > >>best regards, > >>raoul bhatia > > > > > > 2.4.18 kernel, glibc-2.2.5. I'm a tard for not mentioning the core software :p > > > > I just noticed that aztec and office both are still hard on the cpu. With 18 > > players, I'm seeing 10% average CPU usage on them, with short spikes to 15% :-/ > > Maps like dust and inferno however are still in the 2.5-5% range. > > > > I hated aztec anyway ;) > > > > > > > > > > ___ > > 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 > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
>James, > >Are you being ignorant again? The P4 _IS_ a marketing chip, > ...The Pentium 4 is a HORRIBLE design...Read any of the reviews even from >sites... Hate to interfere in what is clearly a bitter rivalry... You're confusing 3D graphics performance with application performance. I love AMD and I run an XP 2400 it in my XPC. But that 2.53 GHz P4 we paid $200 for is quite an engine. I've been lurking in this list for many months, and I havent seen CPU problems such as discussed recently since we got off the 2.2 kernel and Quad Xeon 550 2MBs in a Netfinity server. -BP ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
James Sykes wrote: Clearly blinded by some kind of AMD fanaticism. I like AMD, but at the end of the day Intel still make the better chips. Yeah, I'll drop that in the same category of unproven dogma as the whole "It's an established fact that Windows CPU reporting is very inaccurate" and "THOSE NUMBERS ARE IMPOSSIBLE!" and such... -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
Clearly blinded by some kind of AMD fanaticism. I like AMD, but at the end of the day Intel still make the better chips. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of James Couzens Sent: 22 September 2003 01:23 To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Crazy Usage James, Are you being ignorant again? The P4 _IS_ a marketing chip, and every word Eric said is more or less accurate to a T. The Pentium 4 is a HORRIBLE design which gets higher megahertz with a massive performance. In benchmarking half-life dedicated server under linux on pentium III's and P4's I found a 1.6 Ghz P4 to be equivalent to a P3 866. How do you explain that. Its pretty obvious. Read any of the reviews even from sites like Anandtech or Tom's Hardware (who won't be as brutally honest as they should be). I paid almost $700 for a P4 3.06 chip, which is fortunately a mistake I will never make again. Its processing power is equivalent to that of an AMD Barton XP2600 for which I paid a fraction of the cost ($260 to be precise). http://www.anandtech.com/showdoc.html?i=1360&p=17 "In spite of the 12% lead the 1.5GHz Pentium 4 took in the Quake III Arena benchmarks, the 1.2GHz Athlon on the AMD 760 platform manages to take a 5% lead over the 1.5GHz P4. This is the perfect example of how the Pentium 4 needs a higher clock speed in order to distance itself from the competition. At clock speeds close to that of the Athlon, without any SSE2 specific optimizations, the Pentium 4 will almost always come out under the Athlon." http://www.anandtech.com/showdoc.html?i=1360&p=18 http://www.anandtech.com/showdoc.html?i=1360&p=19 "And for the home/office user, the Pentium 4 would actually be a downgrade in many cases. " Whats also interesting is how their PIII (I love the p3, it was a great chip) Tulatin outperformed their P4 chips even into the 2Ghz range... Intel as Eric indicated, was selling the retarded public with the p4. The box should say "One Testicle, Inside". And this one is just for you: http://prodigy.redphive.org/images/asr-drinks.jpg Cheers, James - Original Message - From: "James Sykes" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Sunday, September 21, 2003 9:46 AM Subject: RE: [hlds_linux] Crazy Usage > Eric, > > Talking shit? However much I prefer P3's for a server platform - there's > no denying P4s performance. P4 a crappy design? I don't think so. > You're not a CPU architect are you? > I await your lovely thought out reply - where you attempt to prove your > point with information you read on GREATWEBSITE.COM.(powered by AMD of > course) > > > > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of Eric > (Deacon) > Sent: 21 September 2003 17:29 > To: [EMAIL PROTECTED] > Subject: Re: [hlds_linux] Crazy Usage > > James Sykes wrote: > > To sum up the total average CPU usage. > > The P4 was using : 1008Mhz > > The p3 was using : 576Mhz > > > > I also joined the servers to see the performance ingame - both were > > about equal.So we've got quite a large usage difference of 432Mhz! > > > > Anyone care to shed some light? > > The P4 is a marketing chip. It's not actually meant to give good > performance. Instead, it's meant to sound good to ignorant end-users. > Intel takes advantage of them by giving them a crappy design, but with > really high clock speeds, and then forces major manufacturers like Dell > to use nothing else. And if you want "good performance", you pay out > the nose for the highest-end CPU, the latest in inefficient design. > > -- > Eric (the Deacon remix) > > > ___ > 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 > ___ 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
RE: [hlds_linux] Crazy Usage
Valve might consider trying a different compiler? http://www.open-mag.com/754088105111.htm Even if they dont, I'd be interested to see the effect of a kernel compiled with ICC on HLDS performance. -EvilGrin -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of James Sykes Sent: Sunday, September 21, 2003 5:46 PM To: [EMAIL PROTECTED] Subject: RE: [hlds_linux] Crazy Usage Eric, Talking shit? However much I prefer P3's for a server platform - there's no denying P4s performance. P4 a crappy design? I don't think so. You're not a CPU architect are you? I await your lovely thought out reply - where you attempt to prove your point with information you read on GREATWEBSITE.COM.(powered by AMD of course) -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Eric (Deacon) Sent: 21 September 2003 17:29 To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Crazy Usage James Sykes wrote: > To sum up the total average CPU usage. > The P4 was using : 1008Mhz > The p3 was using : 576Mhz > > I also joined the servers to see the performance ingame - both were > about equal.So we've got quite a large usage difference of 432Mhz! > > Anyone care to shed some light? The P4 is a marketing chip. It's not actually meant to give good performance. Instead, it's meant to sound good to ignorant end-users. Intel takes advantage of them by giving them a crappy design, but with really high clock speeds, and then forces major manufacturers like Dell to use nothing else. And if you want "good performance", you pay out the nose for the highest-end CPU, the latest in inefficient design. --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.520 / Virus Database: 318 - Release Date: 18/09/2003 ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
Eric, Talking shit? However much I prefer P3's for a server platform - there's no denying P4s performance. P4 a crappy design? I don't think so. You're not a CPU architect are you? I await your lovely thought out reply - where you attempt to prove your point with information you read on GREATWEBSITE.COM.(powered by AMD of course) -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Eric (Deacon) Sent: 21 September 2003 17:29 To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Crazy Usage James Sykes wrote: > To sum up the total average CPU usage. > The P4 was using : 1008Mhz > The p3 was using : 576Mhz > > I also joined the servers to see the performance ingame - both were > about equal.So we've got quite a large usage difference of 432Mhz! > > Anyone care to shed some light? The P4 is a marketing chip. It's not actually meant to give good performance. Instead, it's meant to sound good to ignorant end-users. Intel takes advantage of them by giving them a crappy design, but with really high clock speeds, and then forces major manufacturers like Dell to use nothing else. And if you want "good performance", you pay out the nose for the highest-end CPU, the latest in inefficient design. -- Eric (the Deacon remix) ___ 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
RE: [hlds_linux] Crazy Usage
Just lookin at a full 1.5 16 player server running de_dust2. Note : this is on a dual xeon 2.4 with HT, no other servers running) The CPU figures are of 1 of the 4 virtual processors. (eg. 1.2ghz) TOTAL server usage is about 10-12% 49.3 63.4 62.2 57.2 54.2 59.6 57.9 588mhz CPU. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of raoul bhatia Sent: 21 September 2003 17:05 To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Crazy Usage would you care posting your 1.5 results with these two test machines too? would be nice to see the actual differences there too. James Sykes wrote: > That's not the point. > This difference is MUCH bigger in 1.6 that it was in 1.5 > In 1.5 I saw 100-150mhz difference - no more. ___ 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
Re: [hlds_linux] Crazy Usage
James Sykes wrote: To sum up the total average CPU usage. The P4 was using : 1008Mhz The p3 was using : 576Mhz I also joined the servers to see the performance ingame - both were about equal.So we've got quite a large usage difference of 432Mhz! Anyone care to shed some light? The P4 is a marketing chip. It's not actually meant to give good performance. Instead, it's meant to sound good to ignorant end-users. Intel takes advantage of them by giving them a crappy design, but with really high clock speeds, and then forces major manufacturers like Dell to use nothing else. And if you want "good performance", you pay out the nose for the highest-end CPU, the latest in inefficient design. -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
would you care posting your 1.5 results with these two test machines too? would be nice to see the actual differences there too. James Sykes wrote: That's not the point. This difference is MUCH bigger in 1.6 that it was in 1.5 In 1.5 I saw 100-150mhz difference - no more. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
That's not the point. This difference is MUCH bigger in 1.6 that it was in 1.5 In 1.5 I saw 100-150mhz difference - no more. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Sindre Sent: 21 September 2003 16:14 To: James Sykes; [EMAIL PROTECTED] Subject: RE: [hlds_linux] Crazy Usage P4 have sucky raw fpu performance compared to amd or p3. - Sindre >= Original Message From "James Sykes" <[EMAIL PROTECTED]> = >Just did some testing on a few machines here! >Both machines are running Linux 2.5.75, with HZ set to 100. >Both of the servers are with sys_ticrate 100. > >Each sample was taken every few seconds. >I originally did this for several minutes - I have just posted the >average readings. >No other HLDS were running when these were taken. > >model name : Intel(R) Pentium(R) 4 CPU 2.40GHz >bogomips: 4767.74 > >CPU InOut Uptime Users FPSPlayers >44.67 52.85 67.47 91283 90.92 14 >45.00 51.42 65.07 91283 100.47 14 >44.60 51.60 64.11 91283 90.92 14 >41.00 52.24 65.36 91283 100.02 14 >42.00 51.67 63.74 91283 90.74 14 >42.67 50.97 63.41 91283 83.49 14 >42.67 51.73 63.90 91283 90.92 14 >43.25 51.10 63.03 91283 91.12 14 >43.80 50.12 61.70 91283 83.35 14 > >model name : Intel(R) Pentium(R) III CPU family 1133MHz >bogomips: 2252.80 > >CPU InOut Uptime Users FPSPlayers >58.00 47.56 53.65 7175 71.42 14 >57.80 48.10 54.54 7175 99.91 14 >55.00 47.51 53.44 7175 77.22 14 >56.00 48.49 55.59 7175 69.40 14 >57.33 50.00 58.25 7175 78.48 14 >58.75 50.90 59.95 7175 62.32 14 >59.20 51.56 60.50 7175 75.45 14 >59.00 49.38 56.36 7175 68.58 14 >58.00 47.68 52.31 7175 90.97 14 >55.25 46.62 51.19 7175 81.67 14 > >To sum up the total average CPU usage. >The P4 was using : 1008Mhz >The p3 was using : 576Mhz > >I also joined the servers to see the performance ingame - both were >about equal.So we've got quite a large usage difference of 432Mhz! > >Anyone care to shed some light? I have always found our P3s to use less >cpu clock for clock against our P4 servers - however not such a great >deal - perhaps only 100mhz or so. > >James > > > > > > > > >___ >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 ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
P4 have sucky raw fpu performance compared to amd or p3. - Sindre >= Original Message From "James Sykes" <[EMAIL PROTECTED]> = >Just did some testing on a few machines here! >Both machines are running Linux 2.5.75, with HZ set to 100. >Both of the servers are with sys_ticrate 100. > >Each sample was taken every few seconds. >I originally did this for several minutes - I have just posted the >average readings. >No other HLDS were running when these were taken. > >model name : Intel(R) Pentium(R) 4 CPU 2.40GHz >bogomips: 4767.74 > >CPU InOut Uptime Users FPSPlayers >44.67 52.85 67.47 91283 90.92 14 >45.00 51.42 65.07 91283 100.47 14 >44.60 51.60 64.11 91283 90.92 14 >41.00 52.24 65.36 91283 100.02 14 >42.00 51.67 63.74 91283 90.74 14 >42.67 50.97 63.41 91283 83.49 14 >42.67 51.73 63.90 91283 90.92 14 >43.25 51.10 63.03 91283 91.12 14 >43.80 50.12 61.70 91283 83.35 14 > >model name : Intel(R) Pentium(R) III CPU family 1133MHz >bogomips: 2252.80 > >CPU InOut Uptime Users FPSPlayers >58.00 47.56 53.65 7175 71.42 14 >57.80 48.10 54.54 7175 99.91 14 >55.00 47.51 53.44 7175 77.22 14 >56.00 48.49 55.59 7175 69.40 14 >57.33 50.00 58.25 7175 78.48 14 >58.75 50.90 59.95 7175 62.32 14 >59.20 51.56 60.50 7175 75.45 14 >59.00 49.38 56.36 7175 68.58 14 >58.00 47.68 52.31 7175 90.97 14 >55.25 46.62 51.19 7175 81.67 14 > >To sum up the total average CPU usage. >The P4 was using : 1008Mhz >The p3 was using : 576Mhz > >I also joined the servers to see the performance ingame - both were >about equal.So we've got quite a large usage difference of 432Mhz! > >Anyone care to shed some light? I have always found our P3s to use less >cpu clock for clock against our P4 servers - however not such a great >deal - perhaps only 100mhz or so. > >James > > > > > > > > >___ >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
RE: [hlds_linux] Crazy Usage
Just did some testing on a few machines here! Both machines are running Linux 2.5.75, with HZ set to 100. Both of the servers are with sys_ticrate 100. Each sample was taken every few seconds. I originally did this for several minutes - I have just posted the average readings. No other HLDS were running when these were taken. model name : Intel(R) Pentium(R) 4 CPU 2.40GHz bogomips: 4767.74 CPU InOut Uptime Users FPSPlayers 44.67 52.85 67.47 91283 90.92 14 45.00 51.42 65.07 91283 100.47 14 44.60 51.60 64.11 91283 90.92 14 41.00 52.24 65.36 91283 100.02 14 42.00 51.67 63.74 91283 90.74 14 42.67 50.97 63.41 91283 83.49 14 42.67 51.73 63.90 91283 90.92 14 43.25 51.10 63.03 91283 91.12 14 43.80 50.12 61.70 91283 83.35 14 model name : Intel(R) Pentium(R) III CPU family 1133MHz bogomips: 2252.80 CPU InOut Uptime Users FPSPlayers 58.00 47.56 53.65 7175 71.42 14 57.80 48.10 54.54 7175 99.91 14 55.00 47.51 53.44 7175 77.22 14 56.00 48.49 55.59 7175 69.40 14 57.33 50.00 58.25 7175 78.48 14 58.75 50.90 59.95 7175 62.32 14 59.20 51.56 60.50 7175 75.45 14 59.00 49.38 56.36 7175 68.58 14 58.00 47.68 52.31 7175 90.97 14 55.25 46.62 51.19 7175 81.67 14 To sum up the total average CPU usage. The P4 was using : 1008Mhz The p3 was using : 576Mhz I also joined the servers to see the performance ingame - both were about equal.So we've got quite a large usage difference of 432Mhz! Anyone care to shed some light? I have always found our P3s to use less cpu clock for clock against our P4 servers - however not such a great deal - perhaps only 100mhz or so. James ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
Stan Bubrouski wrote: Again what the hell are you talking about? All the quote above says is that windows CPU usage reporting can be innacurate. I never said it could never reach 100% and I never made a single claim about Windows reporting more or less CPU usage than is true. Jesus. Wait...you say that it's "inaccurate", but then "I never made a single claim about Windows reporting more or less CPU usage than is true." Please explain these two diametrically opposed statements. No CPU usage reporting system is perfect, I don't see your argument here. It seems like you want to just keep touting your own horn. Yes, I'm "touting" my own horn. *sigh* Except that what you said was NOT that "No CPU usage reporting system is perfect." You said "It's an established fact that windows can be very inaccurate when reporting CPU usage." The point was that when he presented to you the fact that running with the kind of low CPU usage that he was seeing with 2.4.9 was not only *possible* in Win32, but that it did, in fact, directly correlated with the numbers he was seeing, showing that it *IS* indeed possible. However, you fire back that Win32 CPU reporting "can be very inaccurate", which means that you don't trust *IT*, either. The problem is, nobody else has seen whatever it is that you've seen to create your inherent distrust of 2.4.9 and Win32. Instead you put words in my mouth that I never said and start up an argument that you want me to now prove that Windows CPU usage isn't perfect. Hahaha...backpeddle much? > How so? You've never seen a process reach 99% CPU? ...as the entirety of your response. Thanks for the enlightenment. I'm not going to waste my time on every flame put forth by you. That's, unfortunately, the kind of response I was expecting. You're not a politician by trade, by any chance, are you? -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage [OT]
Daniel Stroven wrote: This is all I was trying to argue last night, but James and Daniel aren't even open to the possibility that the numbers are even remotely innacurate Yes, I know, I'm sorry, names often get mixed up my head as do the words I'm trying to get out, very frustrating :-/ you mean james and brian. I was the first to dispute the numbers and mention security issues with going back to such an old kernel. Yes very sorry, your completely right. I apologize. I just disagree in an agreeable manner. With testing and my opinions. Im not trying to insult anyone or get into an argument. My boxes all normally run 2.6.0-test kernels at 1000hz. Cpu usage is high. I think this is valve issue mostly, but it can't hurt to test and share results between us to find something that works as good as possible with the current usage issues. I totally agree. I didn't take issue with him testing or retesting, all I didn't agree with was that retesting a bogus value would produce verifyable results. -sb ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
James Sykes wrote: YOUR CPU USAGE IS INCORRECT. It is NOT POSSIBLE to have a FULL 16 PLAYER SERVER RUNNING AT 2% CPU USAGE. Guys, please bear in mind that this almeighty, mythical "CPU" does not exist. Instead, we have an insanely wide variety of both CPUs and their supporting cast. There's a whole lot that goes into system performance, and if you don't specify what kind of CPU you're dealing with, your blanket statements make no sense. -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
This is a multi-part message in MIME format. -- [ Picked text/plain from multipart/alternative ] Deacon what are the specs of your server?? - Original Message - From: Eric (Deacon) To: [EMAIL PROTECTED] Sent: Friday, September 19, 2003 10:08 AM Subject: Re: [hlds_linux] Crazy Usage James Sykes wrote: > I remember having a similar problems with HLDS running on windows. > All the servers would show less than 3% cpu usage - which of course was > complete rubbish. Just by looking at CPU temperature we could see it was > running very fast. (probably 80-90%) Haha! I don't think I've ever heard anyone accuse the Task Manager in Windows as reading lower CPU levels than are accurate! Awesome. I see all *kinds* of new things in this amazing dog and pony show. By the way, if your CPU were really running at 90%, youd' *know* it. Your server would be draagging. -- Eric (the Deacon remix) ___ 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
Re: [hlds_linux] Crazy Usage [OT]
Stan Bubrouski wrote: The fact remains that you cannot run 3 half-full or full HLDS at only 30% total. And that a 10/15 player server is only using 2% of CPU. Why not? -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
Stan Bubrouski wrote: Gotta stop encouraging this kid. This thread is convincing people to use outdated, broken, insecure kernels because his results are skewed by too low of a sampling rate. -sb First of all, stop referring to him as "this kid". You're being a complete asshole, arrogant and insulting in everything you say. Secondly, it's pretty ballsy to say that Linux is vastly inferior to Windows, that Bill Gates has had accurate CPU usage reporting while the entire opensource community has had its collective head up its ass for years. -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
James, > It's an established fact that windows can be very inaccurate when > reporting CPU usage. I myself have also seen this happen. Another blanket statement? Can we stop just making statements and not backing them up? To me, thats not established at all, in fact, I've never read or come across anything that would leave me to disbelieve a windows machine when telling me cpu use. > You may well have lower CPU usage with 2.4.9 - however 2% cpu usage for > a 16 player server is simply INCORRECT. There is no point debating this > fact! > 2% is laughable. You state that it is incorrect, but then you do not back this information up. I am attempting to back up every statement that I am making here, and you simply refute them with, a big fat nothing. Instead of continuing to make useless replies to these messages, do everyone a favour, and prove me wrong. I will be more than happy to conceed on this subject if someone can just supply some answers. And there is point debating this fact. If I could run a 16 player cstrike server (allbeit 1.3) on an AMD K62-250 at ~40-50% cpu, what is to say, that I can not run the same server at negligable mounts of cpu on a processor thats 8 times faster, with a memory bus thats 4 times faster, and has had substantial improvements in architecture? Why is that just so ludacris to you? If you find my previous post regarding dod useage even remotely believeable, they were achieved using the same method to attain those counter-strike results. James ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
I've tried it, and it works, you can easily adjust fps up to a stable 350 (with sys_ticrate and hz 700). That beeing said, after testing the 2.4.9 kernel with hz-patch and different hz', I do NOT believe it's any better than the new kernels. Yes, it show less cpu-use with few players, but as soon as you max your cpu out, it show the same high. - Sindre >= Original Message From Matt Heler <[EMAIL PROTECTED]> = >I doubt that patch applies cleanily to 2.4.9, I tried it and got rejects.. > >On Friday 19 September 2003 06:41 am, Sindre wrote: >> That's why you use: >> http://www.kernel.org/pub/linux/kernel/people/rml/variable-HZ/v2.4/variable >>-hz-rml-2.4.20-pre10-ac2-1.patch >> >> :) >> >> - Sindre >> >> >= Original Message From "James Couzens" <[EMAIL PROTECTED]> = >> >Daniel, >> > >> >Thats not a good idea, as previously tested by Andrew from Divo networks, >> >you will get bad results, even kernel panics. And rightly so, mucking >> > with the timer in unpatched 2.4.x kernels won't do (there is a backport >> > of the 2.5/2.6 work by Robert Love). Roberts' actual words "Without this >> > patch, you cannot just change HZ, since the system exports various values >> > via system call or /proc that are measured in ticks. In other words, >> > things that are in the units of "ticks per seconds" >> > >> >Not only that, but also by changing the timer in an unpatched kernel the >> >date would wrap in 49.7 days instead of the 497 you would get with 100. >> > >> >James >> > >> >--- >> >James Couzens >> >My Half-Life Admin >> >http://myHLAdmin.com >> >- Original Message - >> >> From: "Daniel Stroven" <[EMAIL PROTECTED]> >> >> >To: <[EMAIL PROTECTED]> >> >Sent: Thursday, September 18, 2003 10:00 PM >> >Subject: Re: [hlds_linux] Crazy Usage >> > >> >> James, >> >> >> >> Do me a favor and put one of your kernels at 1000hz and post me some >> >> data? Just so I can see what you get from it. Im betting top will be >> >> off as far as cpu usage..but I want to see what you show for cpustat and >> >> all that. >> >> >> >> thanks >> >> - Original Message - >> >> From: "James Couzens" <[EMAIL PROTECTED]> >> >> To: <[EMAIL PROTECTED]> >> >> Sent: Friday, September 19, 2003 12:45 AM >> >> Subject: Re: [hlds_linux] Crazy Usage >> >> >> >> > > Do we need to? I mean really, look at his numbers, >> >> > > they are beyond wrong, and beyond feasability. >> >> > > >> >> > > In the ChangeLog for 2.4.10 there is this entry: >> >> > > >> >> > > pre11: >> >> > > - Neil Brown: md cleanups/fixes >> >> > > - Andrew Morton: console locking merge >> >> > > - Andrea Arkangeli: major VM merge >> >> > > >> >> > > There were major VM changes throughout the 2.4 kernel >> >> > > series due to problems, but that is not the only >> >> > > problem here. His problem is procps and the kernel >> >> > > most likely...ie. his version of procps does not like >> >> > > his kernel version. Procps (This includes top, ps, >> >> > > vmstat, etc) can be very dependent on kernel versions, >> >> > > I suspect procps and maybe his kernel are both at >> >> > > fault. Quickly scaning some procps changelogs, there >> >> > > are numerous entries about fixes for incorrect CPU >> >> > > usage reporting. There was even a CPU usage reporting >> >> > > bug fixed as recently as december. I don't feel we >> >> > > should have to backup the fact James is wrong. His >> >> > > numbers speak for themselves. Using 0% CPU has nothing >> >> > > to do with this, using 0% of your brain and ignoring >> >> > > the fact his numbers are impossible is more the issue >> >> > > here. >> >> > > >> >> > > -sb >> >> > >> >> > For the last time, i'm not using top, ps, or any other such tools. >> >> > But >> >> >> >> that >> >> >> >> > point is moot, since they agree with the inf
RE: [hlds_linux] Crazy Usage
>NO ITS NOT. GO RUN WINDOWS AND SEE FOR YOUR SELF IDENTICAL CPU USE. It's an established fact that windows can be very inaccurate when reporting CPU usage. I myself have also seen this happen. >I state again, that I can run THREE 16 player servers ALL with > 40fps most >near 50, on the same server running 2.4.9, that with a 2.4.2x kernel, I can >only run 1 SINGLE 16 player server (adding a second takes pings to 300 at >about 8 players, so to be fair 1 1/2 servers). This is not imaginary data. >This if FACT. And I've STILL got oodles of cpu to spare with 2.4.9. I'm >shocked you can so easily toss this very valid information out the window. You may well have lower CPU usage with 2.4.9 - however 2% cpu usage for a 16 player server is simply INCORRECT. There is no point debating this fact! 2% is laughable. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of James Couzens Sent: 19 September 2003 18:16 To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Crazy Usage James, > YOUR CPU USAGE IS INCORRECT. > It is NOT POSSIBLE to have a FULL 16 PLAYER SERVER RUNNING AT 2% CPU > USAGE. > (Assuming you don't have a new Intel prototype server running at 40ghz) NO ITS NOT. GO RUN WINDOWS AND SEE FOR YOUR SELF IDENTICAL CPU USE. > > I don't disagree that you may see lower usage in 2.4.9 - however I > believe the figure of 2% is incorrect. > > In what way does each server "completely change the load 'dynamic'"? > If you are seeing exponential load increases - it just shows how dodgy > those usage figures really are. > > When I run a server - and see 30% usage- it will still be 30% usage > regardless if there is another server running at 30% or not. > I would only expect to see a change in CPU usage if I was operating at > the upper limits of the CPUs power. (eg 90%+ or so) > > > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of James > Couzens > Sent: 19 September 2003 13:52 > To: [EMAIL PROTECTED] > Subject: Re: [hlds_linux] Crazy Usage > > ----- Original Message - > From: "James Sykes" <[EMAIL PROTECTED]> > To: <[EMAIL PROTECTED]> > Sent: Friday, September 19, 2003 5:29 AM > Subject: RE: [hlds_linux] Crazy Usage > > > > If your server has 16 players - and is showing 1-2% cpu - there is > > something very wrong there. > > > > I remember having a similar problems with HLDS running on windows. > > All the servers would show less than 3% cpu usage - which of course > was > > complete rubbish. Just by looking at CPU temperature we could see it > was > > running very fast. (probably 80-90%) > > > > Anyone out there that thinks 1-2% cpu usage is CORRECT - then you > should > > be able to run about 20 full 16 player cs servers - and not even hit > 50% > > cpu! > > No, thats not the case at all. Each server you add will completely > change > the load "dynamic" if you will, exponentially adding to the load. Lets > look > at it this way. With kernels greater then 2.4.9, I can run a SINGLE 16 > player server (if I run two, the players can enjoy 300ms pings \o/). If > I > use 2.4.9, judging by what I saw last night, 3 of the 4 servers were > full, > and everyone who was local had sub 60 ping, in many cases sub 40. > > Everyone looks at my results and thinks I'm trying to tell them you can > run > an exceptional number of servers per box. Thats not only not the case, > but > not my point. My point is to show more reasonable cpu use by hlds. YES > fps > is lower, but at the expensive of lower cpu use? I think thats a damn > fine > trade off if I can run a more respectable number of game servers per > machine. > > > Try it - I bet you start lagging before the 5th fills up. > > > > --- > James Couzens > My Half-Life Admin > http://myHLAdmin.com > > > > ___ > 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 > ___ 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
RE: [hlds_linux] Crazy Usage [OT]
Please, I've had my inbox spammed for something like 3-4 days now to the point where I'm getting messages daily to the tune of a few hundred posts. Take the argument off list - this list is to assist with problems regarding HLDS on Linux, not to scream and shout at each other about who is right or wrong. Feel free to post results of testing these kernel changes but myself and I'm sure a lot of others don't need to watch a dog fight resulting in inbox spam. Cheers, Scott ___ Scott Pettit [EMAIL PROTECTED] -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Daniel Stroven Sent: Saturday, 20 September 2003 8:36 a.m. To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Crazy Usage [OT] This is all I was trying to argue last night, but James and Daniel aren't even open to the possibility that the numbers are even remotely innacurate you mean james and brian. I was the first to dispute the numbers and mention security issues with going back to such an old kernel. I just disagree in an agreeable manner. With testing and my opinions. Im not trying to insult anyone or get into an argument. My boxes all normally run 2.6.0-test kernels at 1000hz. Cpu usage is high. I think this is valve issue mostly, but it can't hurt to test and share results between us to find something that works as good as possible with the current usage issues. - Original Message - From: "Stan Bubrouski" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Friday, September 19, 2003 3:05 PM Subject: Re: [hlds_linux] Crazy Usage [OT] > David Sharpe wrote: > > > That's one of the best worded (albeit not grammatically) > > Replies to a flame I have ever seen, I applaud you ;) > > > > > > Explain to me why. First of all he compares HLDS 1.1.1.0 > performance to HLDS 1.1.1.1 which has nothing to do with > the argument. I've never once mentioned any version of > HLDS. > > I was argueing his false claims that Linux Kernels before > 2.4.10 perform better than later kernels. I even posted > a response from the former maintainer of the 2.4 kernel > Alan Cox to explain why the CPU stats on kernels previous > to 2.4.10 are highly innacurate. The damn thing has 100HZ > sampling which is tells nothing of CPU load as he claims > it does because it doesnt sample enough to give accurate > results. People have posted on here time and time again > that an HZ of 512-1024 is appropriate to achieve accurate > results. In the face of facts, he simply replied to > my older post a total of 6 times each time with different > flames and more misinformation. > > The fact remains that you cannot run 3 half-full or full > HLDS at only 30% total. And that a 10/15 player server > is only using 2% of CPU. This is what I was arguing > not whether or not he is kid. His constant denial and > flames of others who offered factual reasons why his results > are wrong only go to prove that calling him a kid is > appropriate because he is acting like one. I didn't > misquote Alan has he suggests either, I posed to Alan > a simple question last night: > > Why would the CPU usage reporting on a Linux kernel > prior to 2.4.10 show such wildly different results > on later subsequent kernels (I did mention the app > was hlds to him, not that it probably meant anything > more to him than it is a CPU bound process). > > His response was simple: > "It may even just be a timing effect. We sample at 100Hz which > means that bursty cpu reporting is wildly inaccurate. There are debug > tools that can give accurate answers (oprofile) but they hit overall > performance in doing so." > > Alan is referring to the fact that in 2.4 kernels > prior to 2.4.10 the default HZ was set to 100, it > currently resides at 1000HZ I believe. The problem > with the low HZ value is that it doesn't sample the > CPU enough to guarentee its not catching idle time > and thus seeing far too low or too high values. > > In later kernels, the developers decided higher > HZ values were fairer and represented a truer > representation of the CPU usage on all processes. > > The bottom line here is that if you set HZ to 100 > on later 2.4 kernels you get similar if not almost > identical numbers as you would on 2.4.9, but that > doesn't mean they are correct. This is all I was > trying to argue last night, but James and Daniel > aren't even open to the possibility that the numbers > are even remotely innacurate. Read the thread he > just shoots everyone down. BTW isn't it cute > how he flamed the same message from me multiple > times? I wonder how old he really is. > > -sb > > > > > > _
Re: [hlds_linux] Crazy Usage
Brian are you rebuilding world or just kernel in these tests? If its just kernel it would be good to see if a full build world changes the results at all as it could well be an incompatibility there somewhere. Steve / K - Original Message - From: "Brian A. Stumm" <[EMAIL PROTECTED]> > I run an even OLDER kernel than he and have similar results... Thats why I > argue this side > > CPU InOut Uptime Users FPSPlayers > 1.67 17.14 19.18 583 169 66.64 10 > > CPU InOut Uptime Users FPSPlayers > 1.40 16.02 16.75 587 169 94.18 9 > > AMD Barton 2500+ with 1 gig DDR case you missed earlier posts. default > redhat 7.2 install (the EVERYTHING install for redhat 7.2 to be specific). > > Those 'stats' are from last night. stats commands are logged btw, case you > want to go back and retrieve your numbers. > > Someone aluded that I may never see higher than 10% cpu because top is > reporting wrong, yet I can reach 100% while compiling under gcc with load > avg's over 1.0. I've seen at least 1.4. > > > > ___ > 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
Re: [hlds_linux] Crazy Usage
Daniel Stroven wrote: r5-cs5- [CS5] Statistics: CPU: 2% FPS: 50 Players: 0/17 Map: de_dust r5-cs4- [CS4] Statistics: CPU: 0% FPS: 51 Players: 15/17 Map: de_dust 2% with no players and 0% with 15 players? Definitely not right. :) doh! Even 2% with 16 players is just... So nice ! Come on guys, it means you can run up to _*50*_ (fifty) 16 players servers on this computer (before reaching 100% CPU) OK, let's say you don't want to stress your hardware, you'll just run 40 servers of 17 players on it. From a GSP point of view, this is better than ever. What are you guys waiting for to become rich ?! Or there is something else about these numbers ? David. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
Brian A. Stumm wrote: I don't believe that redhat 7.2 fresh install numbers are wrong. Thats what I run and get similar CPU usage to what he reports. Granted I run TFC but I still see 2% cpu with 10 players connected. Jesus can't one of the Valve guys just come out and say you can't run an hlds which only uses 2% CPU with 10 players? -sb ___ 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
Re: [hlds_linux] Crazy Usage
I think if people would leave their kernels alone - and use some tools that have been provided on this list - you'll see. I was frustrated myself with the 3111d binaries on Intel CPU's and requested valve to look into and showed them the difference between AMD and Intel - which is amazing -and keep in mind - everything is default - no tweaking modifying HZ's - nothing - just a plain simple install of Red Hat 8.0 - This weekend will be a great test for Steam on its CPU use. Britt - Original Message - From: "Matt Heler" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Friday, September 19, 2003 12:16 AM Subject: Re: [hlds_linux] Crazy Usage > On Thursday 18 September 2003 09:40 pm, Stan Bubrouski wrote: > > Matt Heler wrote: > > > Thoose are io schedulers that you were refering to they control the disk > > > reading / writing.. I believe your refering to the O(1) scheduler which > > > schedules how much proccesor usage a process gets, and how long it shall > > > get it for. > > > > I/O is huge. There are multiple I/O schedulers and multiple > > process schedulers, again I wasn't clear. Currently in -mmX > > series Nick and Con's (O20.1int) patches are for process scheduling. > Correct , there is also Nick's Policy Scheduler that someone should try out > too.. Eventually if I get around i'll give it a try and see how that > performs. Fundmentally it's a rewrite of the cpu scheduler code from the > ground up and has shown some fairly good numbers in various benchmarks since > its infancy. > > > > > -sb > > > > > Matt > > > > > > On Thursday 18 September 2003 08:32 pm, Stan Bubrouski wrote: > > >>Daniel Stroven wrote: > > >>>I just saw 50% cpu spike. So the max spike on 2.4.9 has been 20% lower > > >>>then average usage on the previous 2.6.0 kernel. Nothing concrete > > >>> yet. I haven't joined the server yet either to see. > > >>> > > >>>James, I will be stopping in to see your servers later tonight I think. > > >>>Just to get a feel. > > >>> > > >>>dan > > >> > > >>This hasn't been mentioned enough, and I apologize > > >>for not bringing it up in a previous post, but > > >>the scheduler in 2.6 is vastly different from 2.4. > > >> > > >>Currently in 2.6 there are competing schedulers and > > >>schedulers which can be used for different tasks. > > >> > > >>For instance there are the AS (the one in the current > > >>test build enabled by default), CFQ, Deadline, and > > >>I think others as well. Before dismissing the 2.6 > > >>kernel, I think it is important to understand that > > >>the scheduler plays a huge roll in performance as well. > > >>The VM and all that are meaningless without good > > >>scheduling and if anyone wants to run a server on > > >>2.6 test kernels I suggest you read up on the > > >>schedulers and decide which one to use. > > >> > > >>-sb > > >> > > >> > > >> > > >>___ > > >>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 > > > > ___ > > 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 > ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
lol - i agree! I've left my kernel alone and loaded up a few server and I'm not seeing the CPU problems like I was before on Intel CPU's on 3111d - with Steam ... I was in AWE when I seen how much better the AMD's did with the AMD Binaries - and so far Steam is matching that on the Intel CPUs. Also note - we're testing like Valve does - just what they release - no plugins/addons, etc... We'll get to that stage once we're certain these binaries are tight. Then get the authors of the plugins to optimize their applications. Britt - Original Message - From: "Brian A. Stumm" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Friday, September 19, 2003 12:02 AM Subject: Re: [hlds_linux] Crazy Usage > On Thu, 18 Sep 2003, James Couzens wrote: > > > Stan, > > > > I am not some "child" with no clue to which I speak. Oh, did I miss > > something? Was Alan Cox in here stating to everyone that 2.4.9 has > > inaccurate cpu use? I'm putting out information here, and none of you can > > prove me wrong, all you can do is call me a "child" or refute my results, > > with what proof? > > > > I challenge someone to correct me here. I am simply looking for the answers > > like the rest of you. Rather than trying to figure out why it is I am > > seeing significantly improved results, you result to dismissal and denial. > > GG. > > Right beside you... I've asked for the proof as well but noone wants to > offer it or even look. They'd rather dismiss it as "impossible" even > though I've ran this kernel for a few years now. It ran great under a > celeron 533. I was told that cpu couldn't run a hlds server several times > yet I did so for years. Sure it ate 40-60% of that cpu but it worked for a > single server. It wasn't til I wanted to run 2 servers that it became a > prob, so when I upgraded to an AMD 2500+ with a mirror image of the > install I see 2% avg cpu per 14 player server and now I'm told all along > my cpu reporting is wrong. If thats true the 533mhz celeron should never > have worked. yet it did and did quite well. Nooone wants to comment on > that. Even though it had a 2 year track record. It even ran 3111d recently > with 16 live players at a lan party and did quite well. Noone comments on > that. They'd rather claim that our cpu reporting is inaccurate supported > by no proof and then state things like (and this is paraphrasing) "i dont > need to look it up because its wrong and there is its not my job to > educate you on why its wrong". > > That mentality is what got these people to there 60% usage of a 2+ gig cpu > with a single 20 player server. This is what got these people to set HZ to > 1000 without understanding what HZ means or does. Sure some understand but > others are just following trying to have a 1337 server. They don't > understand but it must be good if Bob is doing it. > > *rolls eyes* > > > ___ > 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
Re: [hlds_linux] Crazy Usage
At 00:09 9/19/2003 -0500, you wrote: James Couzens wrote: Eric, ? It fluctuated between 5-8% during round start and peaks of 32-38% when people were running around. To be fair avg cpu for that map would have to be in the 20% range. James James Couzens wrote: Its very real. Your ignorance is shocking. Please explain to me how 32% cpu on de_airstrip with 15 players is "not real"? There's no way that you could ever run a 15-player server on de_airstrip and usage be at exactly 32%. It's a mathmatical impossibility. Silly. ...I was being sarcastic :P It's hard to tell on this list sometimes. -- Eric (the Deacon remix) ___ 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
RE: [hlds_linux] Crazy Usage
>What do ya think? The windows solution becoming a "think thru"? If it wasn't for the damn price tag I would have switched long ago. Jeremy ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
No not trolling, and yes I run multiple servers/machines. I just didn't word myself correctly being in a hurry. I was refering to him not being able to tell what kind of difference he was getting on his machine since his tools were inaccurate. Jeremy -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Eric (Deacon) Sent: Thursday, September 18, 2003 7:22 PM To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Crazy Usage [EMAIL PROTECTED] wrote: >>Incorrect or not, you do realize that there is a SIGNIFICANT difference, >>like, an *exponential* difference, between two 14-player servers and one >>28-player server, right? > > How do you know there is a difference if there is no way to measure the > usage? > > Jeremy ...? Are you new to running HL servers and unfamiliar with basic Linux system/CPU monitoring tools, or are you just trolling? -- Eric (the Deacon remix) ___ 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
RE: [hlds_linux] Crazy Usage
>Yes, let's. How about running SETI and seeing what happens? What does Setti have to do with anything? NM don't answer that it doesn't matter. Jeremy ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
well, is there an HLDS_WINDOW list? cauz I'm really thinking of using Windows now. Who cares about 10% cpu given to windows if I can save out 20% and give out better performance to my clients?? What do ya think? The windows solution becoming a "think thru"? - Original Message - From: "Sindre" <[EMAIL PROTECTED]> To: "Alfred Reynolds" <[EMAIL PROTECTED]>; "Half-Life Dedicated Linux Server Mailing List" <[EMAIL PROTECTED]> Sent: Thursday, September 18, 2003 8:09 PM Subject: RE: [hlds_linux] Crazy Usage > Have you actually tried 32 players? > I mean, my dual mp2800+ cannot run 32 players without fps dropping below 30 on > some maps, and I'm the kind of guy that really wants 100fps, which shouldn't > be that hard with such "state of the art" hardware. > Your windows binaries are far better though, it ran the same 32 player server > with better performance on a Thunderbird 1400. (Yes, it was 4.1.1.1d) > > - Sindre > > >= Original Message From Alfred Reynolds <[EMAIL PROTECTED]> = > >These are the 3 machines we actively test with. None have revealed anywhere > >near the CPU load other people report. > > > >Build Machine: > >>gcc -v > >Reading specs from /usr/local/lib/gcc-lib/i686-pc-linux-gnu/2.95.3/specs > >gcc version 2.95.3 20010315 (release) > >>uname -a > >Linux linuxbuild2 2.2.5 #1 Fri Apr 2 16:37:56 MEST 1999 i686 unknown > >SuSE Linux 6.1 (i386) > >> cat /proc/cpuinfo > >processor : 0 > >vendor_id : AuthenticAMD > >cpu family : 6 > >model : 8 > >model name : AMD Athlon(TM) XP 2200+ > >stepping: 0 > >cpu MHz : 1795.387510 > >fdiv_bug: no > >hlt_bug : no > >sep_bug : no > >f00f_bug: no > >fpu : yes > >fpu_exception : yes > >cpuid level : 1 > >wp : yes > >flags : fpu vme de pse tsc msr 6 mce cx8 9 sep 12 pge 14 cmov > >fcmov 17 22 mmx 24 30 3dnow > >bogomips: 1789.13 > > > >GNU C Library production release version 2.0.7 > > > > > >64 bit build machine: > >>gcc -v > >Thread model: posix > >gcc version 3.2.2 (SuSE Linux) > >> uname -a > >Linux 64bitcompiler 2.4.19 #1 Wed Apr 30 15:17:44 UTC 2003 x86_64 unknown > >UnitedLinux 1.0 (AMD64) > >> cat /proc/cpuinfo > >processor : 0 > >vendor_id : AuthenticAMD > >cpu family : 15 > >model : 4 > >model name : Athlon HX > >stepping: 0 > >cpu MHz : 1595.496 > >cache size : 1024 KB > >fpu : yes > >fpu_exception : yes > >cpuid level : 1 > >wp : yes > >flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca > >cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow > >bogomips: 3185.04 > >TLB size: 1088 4K pages > >clflush size: 64 > >address sizes : 40 bits physical, 48 bits virtual > >power management: ts ttp > > > >GNU C Library stable release version 2.2.5 > > > > > >Test1: > >>uname -a > >Linux alfred_linux 2.4.19-16mdkcustom #3 Fri Feb 21 14:26:04 PST 2003 i686 > >unknown unknown GNU/Linux > >Mandrake Linux release 9.0 (dolphin) for i586 > >>cat /proc/cpuinfo > >processor : 0 > >vendor_id : GenuineIntel > >cpu family : 6 > >model : 5 > >model name : Pentium II (Deschutes) > >stepping: 1 > >cpu MHz : 400.914 > >cache size : 512 KB > >fdiv_bug: no > >hlt_bug : no > >f00f_bug: no > >coma_bug: no > >fpu : yes > >fpu_exception : yes > >cpuid level : 2 > >wp : yes > >flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca > >cmov pat pse36 mmx fxsr > >bogomips: 799.53 > > > > > >GNU C Library stable release version 2.2.5 > > > > > > > >dual_bereta_r0x wrote: > >> Alfred Reynolds wrote: > >>> The "stats" command uses the value from /proc//stat , which is > >>> the same value that top uses. Perhaps the difference you are > >>> encountering is due to the sampling intervals (hlds smoothes the > >>> usage over a 5 second window but top simply shows the instantaneous > >>> value). > >>> > >>> - Alfred > >> > >> Would you mind to send us what are the base distro used by Valve to > >> test/deploy hlds? I mean, kernel version, glibc, utils, and stuff. > >> Even if we hack all the possible distros and kernels, we could have > >> YOURS as a base value. > > > >___ > >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 ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
-r5-cs3- [CS3] Statistics: CPU: 2% FPS: 50 Players: 15/17 Map: cs_assault Uptime: 2067 m -r5-cs4- [CS4] Statistics: CPU: 3% FPS: 50 Players: 7/17 Map: cs_italy Uptime: 968 m -r5-cs5- [CS5] Statistics: CPU: 1% FPS: 50 Players: 6/17 Map: de_train Uptime: 433 m -r5-cs2- [CS2] Statistics: CPU: 6% FPS: 50 Players: 3/17 Map: de_dust2 Uptime: 2239 m -r5-cs3- [CS3] Statistics: CPU: 6% FPS: 47 Players: 15/17 Map: cs_assault Uptime: 2071 m -r5-cs4- [CS4] Statistics: CPU: 5% FPS: 50 Players: 9/17 Map: cs_italy Uptime: 972 m -r5-cs5- [CS5] Statistics: CPU: 4% FPS: 50 Players: 7/17 Map: de_train Uptime: 436 m -r5-cs2- [CS2] Statistics: CPU: 8% FPS: 50 Players: 9/17 Map: de_aztec Uptime: 2245 m -r5-cs3- [CS3] Statistics: CPU: 12% FPS: 43 Players: 11/17 Map: cs_office Uptime: 2076 m -r5-cs4- [CS4] Statistics: CPU: 9% FPS: 50 Players: 8/17 Map: cs_italy Uptime: 978 m -r5-cs5- [CS5] Statistics: CPU: 8% FPS: 50 Players: 8/17 Map: de_train Uptime: 442 m -r5-cs2- [CS2] Statistics: CPU: 24% FPS: 50 Players: 9/17 Map: de_aztec Uptime: 2249 m -r5-cs3- [CS3] Statistics: CPU: 26% FPS: 37 Players: 15/17 Map: cs_office Uptime: 2081 m -r5-cs4- [CS4] Statistics: CPU: 25% FPS: 50 Players: 8/17 Map: cs_italy Uptime: 982 m -r5-cs5- [CS5] Statistics: CPU: 24% FPS: 50 Players: 8/17 Map: de_train Uptime: 446 m -r5-cs2- [CS2] Statistics: CPU: 8% FPS: 50 Players: 11/17 Map: de_aztec Uptime: 2252 m -r5-cs3- [CS3] Statistics: CPU: 15% FPS: 42 Players: 15/17 Map: cs_office Uptime: 2083 m -r5-cs4- [CS4] Statistics: CPU: 11% FPS: 50 Players: 8/17 Map: cs_italy Uptime: 984 m -r5-cs5- [CS5] Statistics: CPU: 11% FPS: 50 Players: 8/17 Map: de_train Uptime: 449 m -r5-cs2- [CS2] Statistics: CPU: 12% FPS: 50 Players: 12/17 Map: de_aztec Uptime: 2254 m -r5-cs3- [CS3] Statistics: CPU: 13% FPS: 43 Players: 15/17 Map: cs_office Uptime: 2085 m -r5-cs4- [CS4] Statistics: CPU: 16% FPS: 50 Players: 7/17 Map: cs_italy Uptime: 986 m -r5-cs5- [CS5] Statistics: CPU: 15% FPS: 50 Players: 0/17 Map: de_dust2 Uptime: 450 m I will keep looking and try to get results of when all servers are actually full. You can see 0% because of the way I check for cpu. If it is less than 1% it is rounded down to 0%. And I take 1 sample every 1 second, as you all know hlds servers will spike useage, more at the start and early stages of a round, and much less when players are dead, so think before you leap and say its not possible. If you still think i've been out back on the harley huffing some gas, go play on the servers for your self. 24.207.0.203:27015 24.207.0.203:27016 24.207.0.203:27017 24.207.0.203:27018 James. - Original Message - From: <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, September 18, 2003 8:28 AM Subject: RE: [hlds_linux] Crazy Usage > >http://prodigy.redphive.org/images/de_airstrip.jpg > > You are nutts, that is not real. Please use some common sense if you thing > you can run all those players on one server and get those results. > > Sorry > > Jeremy > > > > ___ > 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
Re: [hlds_linux] Crazy Usage
[EMAIL PROTECTED] wrote: Incorrect or not, you do realize that there is a SIGNIFICANT difference, like, an *exponential* difference, between two 14-player servers and one 28-player server, right? How do you know there is a difference if there is no way to measure the usage? Jeremy ...? Are you new to running HL servers and unfamiliar with basic Linux system/CPU monitoring tools, or are you just trolling? -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
Have you actually tried 32 players? I mean, my dual mp2800+ cannot run 32 players without fps dropping below 30 on some maps, and I'm the kind of guy that really wants 100fps, which shouldn't be that hard with such "state of the art" hardware. Your windows binaries are far better though, it ran the same 32 player server with better performance on a Thunderbird 1400. (Yes, it was 4.1.1.1d) - Sindre >= Original Message From Alfred Reynolds <[EMAIL PROTECTED]> = >These are the 3 machines we actively test with. None have revealed anywhere >near the CPU load other people report. > >Build Machine: >>gcc -v >Reading specs from /usr/local/lib/gcc-lib/i686-pc-linux-gnu/2.95.3/specs >gcc version 2.95.3 20010315 (release) >>uname -a >Linux linuxbuild2 2.2.5 #1 Fri Apr 2 16:37:56 MEST 1999 i686 unknown >SuSE Linux 6.1 (i386) >> cat /proc/cpuinfo >processor : 0 >vendor_id : AuthenticAMD >cpu family : 6 >model : 8 >model name : AMD Athlon(TM) XP 2200+ >stepping: 0 >cpu MHz : 1795.387510 >fdiv_bug: no >hlt_bug : no >sep_bug : no >f00f_bug: no >fpu : yes >fpu_exception : yes >cpuid level : 1 >wp : yes >flags : fpu vme de pse tsc msr 6 mce cx8 9 sep 12 pge 14 cmov >fcmov 17 22 mmx 24 30 3dnow >bogomips: 1789.13 > >GNU C Library production release version 2.0.7 > > >64 bit build machine: >>gcc -v >Thread model: posix >gcc version 3.2.2 (SuSE Linux) >> uname -a >Linux 64bitcompiler 2.4.19 #1 Wed Apr 30 15:17:44 UTC 2003 x86_64 unknown >UnitedLinux 1.0 (AMD64) >> cat /proc/cpuinfo >processor : 0 >vendor_id : AuthenticAMD >cpu family : 15 >model : 4 >model name : Athlon HX >stepping: 0 >cpu MHz : 1595.496 >cache size : 1024 KB >fpu : yes >fpu_exception : yes >cpuid level : 1 >wp : yes >flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca >cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext lm 3dnowext 3dnow >bogomips: 3185.04 >TLB size: 1088 4K pages >clflush size: 64 >address sizes : 40 bits physical, 48 bits virtual >power management: ts ttp > >GNU C Library stable release version 2.2.5 > > >Test1: >>uname -a >Linux alfred_linux 2.4.19-16mdkcustom #3 Fri Feb 21 14:26:04 PST 2003 i686 >unknown unknown GNU/Linux >Mandrake Linux release 9.0 (dolphin) for i586 >>cat /proc/cpuinfo >processor : 0 >vendor_id : GenuineIntel >cpu family : 6 >model : 5 >model name : Pentium II (Deschutes) >stepping: 1 >cpu MHz : 400.914 >cache size : 512 KB >fdiv_bug: no >hlt_bug : no >f00f_bug: no >coma_bug: no >fpu : yes >fpu_exception : yes >cpuid level : 2 >wp : yes >flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca >cmov pat pse36 mmx fxsr >bogomips: 799.53 > > >GNU C Library stable release version 2.2.5 > > > >dual_bereta_r0x wrote: >> Alfred Reynolds wrote: >>> The "stats" command uses the value from /proc//stat , which is >>> the same value that top uses. Perhaps the difference you are >>> encountering is due to the sampling intervals (hlds smoothes the >>> usage over a 5 second window but top simply shows the instantaneous >>> value). >>> >>> - Alfred >> >> Would you mind to send us what are the base distro used by Valve to >> test/deploy hlds? I mean, kernel version, glibc, utils, and stuff. >> Even if we hack all the possible distros and kernels, we could have >> YOURS as a base value. > >___ >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
Re: [hlds_linux] Crazy Usage
Frank Stollar wrote: The new VM started with 2.4.10 was (after some bugs) superior as the old VM. Proved by application benchmarks. And the new scheduler in 2.6. adds even more power to the kernel. Why should all other applications gain from the new VM besides HL? Forgive my ignorance, but what are you referring to with "VM"? "Virtual Machine" or "Virtual Memory" management, or what? -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
[EMAIL PROTECTED] wrote: You are proving nothing. You are offering opinions and just assuming your programs are correct. In his defence, he's showing you hard numbers, not opinions, and yes, he's assuming that proc isn't lying. Has innacurate proc info been a major issue reported by anyone else in any other role before, or is this the first time anyone has ever discounted someone's results and "ASSUMING that his programs are INcorrect"? When we all know it is just impossible for the 1% you quote to be true.if you came on the list and gave a low but believable number you might not get all the flack, but common, 1%? Something is wrong in that machine, just face it. Lets do a test. Yes, let's. How about running SETI and seeing what happens? -- Eric (the Deacon remix) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
load average: 1.88, 1.75, 1.69 2579 wicked15 0 148m 132m 9344 R 56.2 6.5 267:46.26 hlds_i686 CPU InOut Uptime Users FPSPlayers 58.00 69.71 105.751345 1164 90.91 19 2410 betatest 15 0 113m 97m 9344 R 52.5 4.8 231:45.91 hlds_i686 CPU InOut Uptime Users FPSPlayers 53.00 41.98 54.011370 1064 76.53 18 2770 sylentas 15 0 121m 106m 9344 S 45.6 5.3 482:32.21 hlds_i686 CPU InOut Uptime Users FPSPlayers 34.20 42.24 48.391227 2095 83.05 18 Pretty close, with the difference being what Alfred pointed out. RH-7.2 2.6.0-test4-mm3-1 #1 SMP at 1000hz dual 2.2ghz xeon 2 gigcs ddr corsair sys_ticrate 100 pingboost 1 Want to see huge jump in server FPS same servers?..same maps with sys_ticrate 1000 instead of 100. 2579 wicked15 0 145m 130m 9344 R 63.5 6.5 270:56.59 hlds_i686 CPU InOut Uptime Users FPSPlayers 78.00 63.29 83.671350 1179 140.39 18 2410 betatest 15 0 113m 97m 9344 S 69.5 4.8 234:36.78 hlds_i686 CPU InOut Uptime Users FPSPlayers 75.67 50.79 61.631376 1083 163.59 18 2770 sylentas 15 0 121m 106m 9344 R 66.2 5.3 485:18.99 hlds_i686 CPU InOut Uptime Users FPSPlayers 67.00 44.07 52.601234 2124 323.42 18 FPS can go higher with sys_ticrate at 1000, but they seem more stable if kept at 100. Between 70-100 on average. - Original Message - From: "Steven Hartland" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, September 18, 2003 7:16 PM Subject: Re: [hlds_linux] Crazy Usage > I must say those stats are laughable what do u have to do on that box to > get something to show above 10% calculate pi to 1000 decimal places :P > > Seriously though I just checked out stats on our boxes and guess what > they don't agree with top or ps: > CPU InOut Uptime Users FPSPlayers > 0.00 0.00 0.00 4 0 50.94 0 > stats > CPU InOut Uptime Users FPSPlayers > 0.00 0.00 0.00 5 0 54.85 0 > stats > CPU InOut Uptime Users FPSPlayers > 0.00 0.00 0.00 5 0 99.19 0 > > top: > 21285 game 80 60044K 54284K nanslp 0 0:09 1.07% 1.07% hlds_i686 > > And on populated server: > CPU InOut Uptime Users FPSPlayers > 0.00 22.90 23.61 2 586 49.71 12 > > top: > 11324 game 80 81788K 76488K nanslp 2 150:24 15.09% 15.09% hlds_i686 > > So Im not to convinced with stats accurately showing the same as top or ps. > > Steve / K > - Original Message - > From: "Brian A. Stumm" <[EMAIL PROTECTED]> > > > > Well add my top to your list of not trusted then Even though this is a > > total default redhat 7.2 install Here is a one hour's worth of stats > > reporting at 2 minute intervals from last night during a more peak playing > > time. Keep in mind there are 4 bots playing during this time which do not > > count under "Players" (valves stats deal dont count fake players). This > > info basically tells me that CPU is not always directly tied to players as > > I see .5% with 8 players and over 1% with as few as 2 players. > > > > 10:34pm > > CPU InOut Uptime Users FPSPlayers > > 1.67 17.14 19.18 583 169 66.64 10 > > > > 10:36pm > > CPU InOut Uptime Users FPSPlayers > > 1.00 20.05 23.39 585 169 51.77 10 > > > > 10:38pm > > CPU InOut Uptime Users FPSPlayers > > 1.40 16.02 16.75 587 169 94.18 9 > > > > 10:40pm > > CPU InOut Uptime Users FPSPlayers > > 0.50 16.73 18.80 589 172 52.23 9 > > > > 10:42pm > > CPU InOut Uptime Users FPSPlayers > > 0.40 16.73 19.60 591 174 96.32 9 > > > > 10:44pm > > CPU InOut Uptime Users FPSPlayers > > 1.00 15.83 17.80 593 174 49.63 9 > > > > 10:46pm > > CPU InOut Uptime Users FPSPlayers > > 2.00 17.06 20.45 595 174 51.03 9 > > > > 10:48 pm > > CPU InOut Uptime Users FPSPlayers > > 1.33 14.15 15.09 597 174 99.31 9 > > > > 10:50pm > > CPU InOut Uptime Users FPSPlayers > > 1.20 15.21 16.72 599 174 51.12 9 > > > > 10:52pm > > CPU InOut Uptime Users FPSPlayers > > 1.75 16.71 20.85 601 174 99.25 8 > > > > 10:54pm > > CPU InOut Uptime Users FPSPlayers > > 0.50 15.25 18.42 603 174 88.28 8 >
RE: [hlds_linux] Crazy Usage
yeah it was my mistake, I'd forgotten ps aux averages the cpu% over the uptime or the process. The hlds stat cpu report and that reported by top are pretty much the same (give or take the odd spike). It's been a while since I've been running on Linux :) -- Nathan Woodcock System Administrator (BarrysWorld/GAME) GAME DIGITAL LTD [EMAIL PROTECTED] http://www.barrysworld.com/ http://play.game.net/ > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of > Alfred Reynolds > Sent: 18 September 2003 18:45 > To: Half-Life Dedicated Linux Server Mailing List > Subject: RE: [hlds_linux] Crazy Usage > > > The "stats" command uses the value from /proc//stat , > which is the same > value that top uses. Perhaps the difference you are > encountering is due to > the sampling intervals (hlds smoothes the usage over a 5 > second window but > top simply shows the instantaneous value). > > - Alfred > > Nathan Woodcock - BarrysWorld wrote: > > how accurate is rcon stats? It doesn't bare much > > resemblance at all to > > the process use as shown by ps. > > > > > >> -Original Message- > >> From: [EMAIL PROTECTED] > >> [mailto:[EMAIL PROTECTED] On Behalf > Of Daniel > >> Stroven Sent: 18 September 2003 15:19 > >> To: [EMAIL PROTECTED] > >> Subject: Re: [hlds_linux] Crazy Usage > >> > >> > >> James, > >> > >> Can you give us some "rcon stats" results on these games? When you > >> the servers are close to full? > >> > >> I have cpustat installed when I tested out the 2.4.9 kernel. Pretty > >> neat, but I would like to see some info from rcon stats if possible > >> from you. > >> > >> Also if you can show some load stats from top for these servers. > >> thanks > >> > >> dan > >> - Original Message - > >> From: "James Couzens" <[EMAIL PROTECTED]> > >> To: <[EMAIL PROTECTED]> > >> Sent: Thursday, September 18, 2003 3:59 AM > >> Subject: Re: [hlds_linux] Crazy Usage > >> > >> > >>> Daniel, > >>> > >>> Believe me, it can, and it does. Taking the server to larger maps > >>> does yield cpu above 1%, please note that I was testing the > >>> smallest map, de_dust. Previously posted on this list were some > >>> screenshots: > >>> > >>> http://prodigy.redphive.org/images/de_airstrip.jpg > >>> http://prodigy.redphive.org/images/de_aztec.jpg > >>> http://prodigy.redphive.org/images/de_aztec2.jpg > >>> http://prodigy.redphive.org/images/de_dust.jpg > >>> http://prodigy.redphive.org/images/de_dust2.jpg > >>> > >>> The cpu use is very REAL, and even for YES 30 slots. I have seen > >>> identical results in windows. As Eric kindly pointed out > >>> previously, 30 slots is a relative term. A 30 slot server, with > >>> all slots bound to the same server would use an exponentially > >>> larger amount of cpu than two > >> servers running > >> 15. > >>> > >>> Cheers, > >>> > >>> James > >>> > >>> - Original Message - > >>> From: "Daniel Stroven" <[EMAIL PROTECTED]> > >>> To: <[EMAIL PROTECTED]> > >>> Sent: Thursday, September 18, 2003 12:34 AM > >>> Subject: Re: [hlds_linux] Crazy Usage > >>> > >>> > >>>> Server FPS I have seen over 500. 100 is what i would > >> ideally want. 50 > >>> fps > >>>> or lower is not what I want. > >>>> > >>>> To me its simple. Regardless of kernel or regardless of > >> OS, a cpu can > >>> only > >>>> handle so much. A cpu is not going to just use 1 percent > >> for near 30 > >>> slots, > >>>> unless its some super duper chip. There is way to much > >> going on for it > >> to > >>>> use 1%, that is not even realistic to think so. > >>>> > >>>> Put that same chip on windows servers which seem to > >> utilize cpu better > >>> than > >>>> linux currently, and you will not see 1% usage. In fact I would > >>>> say its down right physically impossible with the current > >>>> chips..even 3ghz chip. > >>>> > >>>> If it works for you great, for me the
RE: [hlds_linux] Crazy Usage
The "stats" command uses the value from /proc//stat , which is the same value that top uses. Perhaps the difference you are encountering is due to the sampling intervals (hlds smoothes the usage over a 5 second window but top simply shows the instantaneous value). - Alfred Nathan Woodcock - BarrysWorld wrote: > how accurate is rcon stats? It doesn't bare much > resemblance at all to > the process use as shown by ps. > > >> -Original Message- >> From: [EMAIL PROTECTED] >> [mailto:[EMAIL PROTECTED] On Behalf Of Daniel >> Stroven Sent: 18 September 2003 15:19 >> To: [EMAIL PROTECTED] >> Subject: Re: [hlds_linux] Crazy Usage >> >> >> James, >> >> Can you give us some "rcon stats" results on these games? When you >> the servers are close to full? >> >> I have cpustat installed when I tested out the 2.4.9 kernel. Pretty >> neat, but I would like to see some info from rcon stats if possible >> from you. >> >> Also if you can show some load stats from top for these servers. >> thanks >> >> dan >> ----- Original Message - >> From: "James Couzens" <[EMAIL PROTECTED]> >> To: <[EMAIL PROTECTED]> >> Sent: Thursday, September 18, 2003 3:59 AM >> Subject: Re: [hlds_linux] Crazy Usage >> >> >>> Daniel, >>> >>> Believe me, it can, and it does. Taking the server to larger maps >>> does yield cpu above 1%, please note that I was testing the >>> smallest map, de_dust. Previously posted on this list were some >>> screenshots: >>> >>> http://prodigy.redphive.org/images/de_airstrip.jpg >>> http://prodigy.redphive.org/images/de_aztec.jpg >>> http://prodigy.redphive.org/images/de_aztec2.jpg >>> http://prodigy.redphive.org/images/de_dust.jpg >>> http://prodigy.redphive.org/images/de_dust2.jpg >>> >>> The cpu use is very REAL, and even for YES 30 slots. I have seen >>> identical results in windows. As Eric kindly pointed out >>> previously, 30 slots is a relative term. A 30 slot server, with >>> all slots bound to the same server would use an exponentially >>> larger amount of cpu than two >> servers running >> 15. >>> >>> Cheers, >>> >>> James >>> >>> - Original Message - >>> From: "Daniel Stroven" <[EMAIL PROTECTED]> >>> To: <[EMAIL PROTECTED]> >>> Sent: Thursday, September 18, 2003 12:34 AM >>> Subject: Re: [hlds_linux] Crazy Usage >>> >>> >>>> Server FPS I have seen over 500. 100 is what i would >> ideally want. 50 >>> fps >>>> or lower is not what I want. >>>> >>>> To me its simple. Regardless of kernel or regardless of >> OS, a cpu can >>> only >>>> handle so much. A cpu is not going to just use 1 percent >> for near 30 >>> slots, >>>> unless its some super duper chip. There is way to much >> going on for it >> to >>>> use 1%, that is not even realistic to think so. >>>> >>>> Put that same chip on windows servers which seem to >> utilize cpu better >>> than >>>> linux currently, and you will not see 1% usage. In fact I would >>>> say its down right physically impossible with the current >>>> chips..even 3ghz chip. >>>> >>>> If it works for you great, for me the performance simply wasn't >>>> there. >>>> - Original Message - >>>> From: "James Couzens" <[EMAIL PROTECTED]> >>>> To: <[EMAIL PROTECTED]> >>>> Sent: Thursday, September 18, 2003 2:15 AM >>>> Subject: [hlds_linux] Crazy Usage >>>> >>>> >>>>> Jeremy, >>>>> >>>>> I assure you, it is not incorrect. Its very correct. >> Kernel timings >> in >>>>> kernel-2.4.21 which I have also tested against, use identical >>>>> kernel sleeping habbits: >>>>> >>>>> LINUX 2.4.21 : >>>>> >>>>> [EMAIL PROTECTED] asm-i386 # cat >>> /usr/src/linux-2.4.21/include/asm-i386/param.h >>>>> #ifndef _ASMi386_PARAM_H >>>>> #define _ASMi386_PARAM_H >>>>> >>>>> #ifndef HZ >>>>> #define HZ 100 >>>>> #endif >>>>> >>>>> #define EXEC_PAGESIZE 4096 >>>>> >>>>> #ifndef NGROUPS >
RE: [hlds_linux] Crazy Usage
I'm running cs 1.6 on a dual p3-1ghz box as well and get much better performance than you (using kernel 2.4.22). We're seeing it use around 25% of the total cpu available (ie. approx 50% of a gig cpu) with 14-16 players on aztec. If you're seeing 95% of a gig cpu used with 14 players I'd say theres something else in the equation. Our setup is totally defauly bog stadnard hlds 1.1.2.0 with cs 1.6 (no metamod, no plugins etc). -- Nathan Woodcock System Administrator (BarrysWorld/GAME) GAME DIGITAL LTD [EMAIL PROTECTED] http://www.barrysworld.com/ http://play.game.net/ > -Original Message- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On Behalf Of > Daniel Stroven > Sent: 18 September 2003 18:25 > To: [EMAIL PROTECTED] > Subject: Re: [hlds_linux] Crazy Usage > > > James, > > I have no doubt, you and many other on this list have far more linux > knowledge than me. Myself, I learned the little I know by > trial and error, > and using "history" and man pages. If its a comparision of > knowledge of > linux os and its kernels, I am not competent enough to > dispute findings with > technical backup. > > However, I have shown in the 5yrs of running my business, > that I find or > notice problems long before most techs do. I am even really good at > figuring out what is wrong, though sometimes I am personally > unable to fix > it due to limited knowledge. > > As far as this kernel topic goes, I am confident that something is > definitely wrong with hlds binaries. I ran the beta linux > since the day it > was out, and I am religious about watching cpu usage, fps, > ping and in game > performance. Almost fanatical about it. > > I do not believe a kernel that is a couple years out of date, > with a VM that > no longer exists in any kernels since 2.4.9 is the solution > to the usage > issues with steam hlds. > > As I posted, we tried 2.4.9 at 4 different hz variables. 100 > default, and > then 500,700,1000. I do not feel the cpu usage stats from > 100hz was any > more reliable than at 500/700 or 1000hz. > > I know that in game performance on 2.4.9 was no where near > the performance > of my 2.6.0-test-mm3 kernel. Not only by the pings, fps, but > the smoothness > of play. I also know from testing a gcc3 compiled Intel > binary that there > can be improvements to the hlds engine itself. Lower load, > higher fps and > lower usage. Nothing "great", but a marked difference. If I believed > believed the information from cpustat/top/stats, I would have > no fear to > load up another three 20 player servers on the test box. > > But, the fact remains that regardless of kernel or operating > system, a cpu > can only handle so much usage. Prior to 1.6, I could run a > 20 player server > smoothly on a single 1ghz processor under my 2.6.0 kernel. > On 1.6 hlds > binary, I can't run a 12 player server on a 1ghz server. > > Now, the only difference between the now and then, is the > binaries, and the > changes to the game/maps. I would be more likely to believe > an improvment > in stats using a new kernel versus switching to an outdated > one with a VM > that no longer exists. > > 1% cpu I could never believe, even if the pope himself swore > by the results. > > I will undertake one more test. Since testing on dual xeon's > gave great > results from stats point of view, but decreased performance > from an in game > point of viewI will take one of my dual p3 1ghz boxes > with hardware that > is 3yrs old. The same box that ran my two 14 player 24/7 awp > map servers. > In 1.5 and previous I ran both with low ping and smooth > play..with max cpu > usage on each processor at about 70%..while 1.6 usings over > 95% for the same > single 12 player awp server on 1ghz cpu. > > If I can run 2 14 player servers with the same low ping, > smooth play and > much reduced cpu...then I will consider your results. But if > load on the > server is the same, and in game play is as bad with the 2 > servers on, and > cpustat/top/stats show low cpu, then for me it is proven > beyond reasonable > doubt that cpu reporting for the 2.4.9 kernel by ANY program > is faulty. I > personally think that is why it was probably ripped out of the kernel > suddenly. They had all these beautiful stats on cpu, which in the end > turned out to be incorrect. The stats could quite easily be > reading the > proc information correctly, but that doesn't mean the kernel > VM is showing > or reporting that information correctly. > > I will let you know. > > dan > - Original Message - > From: "James Couzens" <[EMAIL PRO
Re: [hlds_linux] Crazy Usage
Brian wrote: Heres something that may prove interesting. If cpu reporting for hlds is innaccurte in older kernels then so should it be innaccurate for other apps, like TOP itself. How much CPU does top use on your system? Well as I stated earlier, you can see various "top" usage at different hz. At 700-1000hz, top says it uses 20-40% cpu. At 100-500hz it says it uses nothing. Top shouldn't use much cpu, I just think the whole reporting of the VM proc in 2.4.9 is faulty. Software measuring it is right im sure, but I think the VM itself does not show proper information. - Original Message - From: "Frank Stollar" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, September 18, 2003 12:24 PM Subject: Re: [hlds_linux] Crazy Usage > [EMAIL PROTECTED] wrote: > >>I am telling you, and PROVING to you, that using this kernel, with DEFAULT > >>kernel timings results in performance vastly superior to any kernel > >>released after. > > > When we all know it is just impossible for the 1% you quote to be true. > > I fully ACK Jeremy. This is not possible, no way. The new VM started > with 2.4.10 was (after some bugs) superior as the old VM. Proved by > application benchmarks. And the new scheduler in 2.6. adds even more > power to the kernel. Why should all other applications gain from the new > VM besides HL? Indeed why should a kernel which only 'manages' the > ressources be guilty for a userspace programm needing 1/10 of the CPU > Power. The calculations done by hlds are done the same way, with 2.4.9 > or 2.4.22. And HLDS is no high-stress program like a database with 100 > nodes and a few 10xgig of data. Why should the hlds do its calculations > 10 times faster with a differnet kernel, just saying 'ho you can now > have some CPU power' and 'here is the memory you wanted'. Normaly the > kernel does almost not need much CPU power but the user programms can > have let's say ~95% of the CPU power. There may be a fault in the HLDS > code, no other dedicated server around like Tribes2, Q3 and so on are > needing that much CPU power from a server. Sometimes I whish I could > have a look for optimizing the HLDS engine or VALVe would make it under > the GPL or any other Licence as HL2 is approching the horizon. Also > Steam is a nice idea and I would like it is a GPL project, many people > can offer there knowledge and talents to it. If Steam would be > open-sorce, nevertheless VALVe could use it as destribution for there > very own products without a hassle, like other open-source online-shops > projects. But I'm dreaming and going to far... > > >With your > > numbers that should mean that you could add 50 16 player 24/7 iceworld > > server. Name them all 24/7 iceworld in the hostname and wait for them to > > fill up. Lets see your server lock up WAY before you ever get to those 800 > > concurrent players. Because with your match you could get all 800 on there > > and still have powere to spare. > > Well nothing to say about this. The numbers speak for themselfes. If > that would be true, HLDS has a bigger problem as anybody ever expected. :) > > cheers > Frank > > > ___ > 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
Re: [hlds_linux] Crazy Usage
James, I have no doubt, you and many other on this list have far more linux knowledge than me. Myself, I learned the little I know by trial and error, and using "history" and man pages. If its a comparision of knowledge of linux os and its kernels, I am not competent enough to dispute findings with technical backup. However, I have shown in the 5yrs of running my business, that I find or notice problems long before most techs do. I am even really good at figuring out what is wrong, though sometimes I am personally unable to fix it due to limited knowledge. As far as this kernel topic goes, I am confident that something is definitely wrong with hlds binaries. I ran the beta linux since the day it was out, and I am religious about watching cpu usage, fps, ping and in game performance. Almost fanatical about it. I do not believe a kernel that is a couple years out of date, with a VM that no longer exists in any kernels since 2.4.9 is the solution to the usage issues with steam hlds. As I posted, we tried 2.4.9 at 4 different hz variables. 100 default, and then 500,700,1000. I do not feel the cpu usage stats from 100hz was any more reliable than at 500/700 or 1000hz. I know that in game performance on 2.4.9 was no where near the performance of my 2.6.0-test-mm3 kernel. Not only by the pings, fps, but the smoothness of play. I also know from testing a gcc3 compiled Intel binary that there can be improvements to the hlds engine itself. Lower load, higher fps and lower usage. Nothing "great", but a marked difference. If I believed believed the information from cpustat/top/stats, I would have no fear to load up another three 20 player servers on the test box. But, the fact remains that regardless of kernel or operating system, a cpu can only handle so much usage. Prior to 1.6, I could run a 20 player server smoothly on a single 1ghz processor under my 2.6.0 kernel. On 1.6 hlds binary, I can't run a 12 player server on a 1ghz server. Now, the only difference between the now and then, is the binaries, and the changes to the game/maps. I would be more likely to believe an improvment in stats using a new kernel versus switching to an outdated one with a VM that no longer exists. 1% cpu I could never believe, even if the pope himself swore by the results. I will undertake one more test. Since testing on dual xeon's gave great results from stats point of view, but decreased performance from an in game point of viewI will take one of my dual p3 1ghz boxes with hardware that is 3yrs old. The same box that ran my two 14 player 24/7 awp map servers. In 1.5 and previous I ran both with low ping and smooth play..with max cpu usage on each processor at about 70%..while 1.6 usings over 95% for the same single 12 player awp server on 1ghz cpu. If I can run 2 14 player servers with the same low ping, smooth play and much reduced cpu...then I will consider your results. But if load on the server is the same, and in game play is as bad with the 2 servers on, and cpustat/top/stats show low cpu, then for me it is proven beyond reasonable doubt that cpu reporting for the 2.4.9 kernel by ANY program is faulty. I personally think that is why it was probably ripped out of the kernel suddenly. They had all these beautiful stats on cpu, which in the end turned out to be incorrect. The stats could quite easily be reading the proc information correctly, but that doesn't mean the kernel VM is showing or reporting that information correctly. I will let you know. dan - Original Message - From: "James Couzens" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, September 18, 2003 10:38 AM Subject: Re: [hlds_linux] Crazy Usage > Daniel, > > Statistics information, IE "cpu jitters" are constantly updated in a single > place in linux, and that is proc. Thats where the HLDS "stats" function > gets it from, and thats where ps, top, and any other utilitiy in linux that > reports CPU % will also obtain it from this directory. This is where > CpuStat gets it from also. > > Provided you do not bugger with the kernel, the information found in this > directory its EXTREMELY accurate, and I can only think that the sheer > disbelief I am getting from most of you on this list is due to a lack of > knowledge regarding how the linux operating system functions. > > Numbers, in this case, do not lie. If you start to bugger with HZ then yes > you are going to get some bizzare and unbelievable results. Provided there > is no interference from some modification such as this the information in > /proc will be true. > > To prove my point to you all, I now have 4 16 player Steam CS 1.6 servers > running on this machine. I shall post results later today. > > James > > - Original Message - > From: "Daniel Stroven" <[EMAIL PROTECTED]>
Re: [hlds_linux] Crazy Usage
[EMAIL PROTECTED] wrote: I am telling you, and PROVING to you, that using this kernel, with DEFAULT kernel timings results in performance vastly superior to any kernel released after. When we all know it is just impossible for the 1% you quote to be true. I fully ACK Jeremy. This is not possible, no way. The new VM started with 2.4.10 was (after some bugs) superior as the old VM. Proved by application benchmarks. And the new scheduler in 2.6. adds even more power to the kernel. Why should all other applications gain from the new VM besides HL? Indeed why should a kernel which only 'manages' the ressources be guilty for a userspace programm needing 1/10 of the CPU Power. The calculations done by hlds are done the same way, with 2.4.9 or 2.4.22. And HLDS is no high-stress program like a database with 100 nodes and a few 10xgig of data. Why should the hlds do its calculations 10 times faster with a differnet kernel, just saying 'ho you can now have some CPU power' and 'here is the memory you wanted'. Normaly the kernel does almost not need much CPU power but the user programms can have let's say ~95% of the CPU power. There may be a fault in the HLDS code, no other dedicated server around like Tribes2, Q3 and so on are needing that much CPU power from a server. Sometimes I whish I could have a look for optimizing the HLDS engine or VALVe would make it under the GPL or any other Licence as HL2 is approching the horizon. Also Steam is a nice idea and I would like it is a GPL project, many people can offer there knowledge and talents to it. If Steam would be open-sorce, nevertheless VALVe could use it as destribution for there very own products without a hassle, like other open-source online-shops projects. But I'm dreaming and going to far... >With your numbers that should mean that you could add 50 16 player 24/7 iceworld server. Name them all 24/7 iceworld in the hostname and wait for them to fill up. Lets see your server lock up WAY before you ever get to those 800 concurrent players. Because with your match you could get all 800 on there and still have powere to spare. Well nothing to say about this. The numbers speak for themselfes. If that would be true, HLDS has a bigger problem as anybody ever expected. :) cheers Frank ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
>I am telling you, and PROVING to you, that using this kernel, with DEFAULT >kernel timings results in performance vastly superior to any kernel released >after. You are proving nothing. You are offering opinions and just assuming your programs are correct. When we all know it is just impossible for the 1% you quote to be true. if you came on the list and gave a low but believable number you might not get all the flack, but common, 1%? Something is wrong in that machine, just face it. Lets do a test. If it is really running at 1%. keep adding 16 player 24/7 iceworld servers until it's at 50%. With your numbers that should mean that you could add 50 16 player 24/7 iceworld server. Name them all 24/7 iceworld in the hostname and wait for them to fill up. Lets see your server lock up WAY before you ever get to those 800 concurrent players. Because with your match you could get all 800 on there and still have powere to spare. Complete bs. Even the original CS that everyone loves couldn't do that, and I thing at one point and tile the 2.4.9 kerel was the current one during the early CS day. Jeremy ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
You can post the results all day long, it is just physically impossible to run that many slots with 1% total cpu usage. Sorry. Jeremy -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of James Couzens Sent: Thursday, September 18, 2003 9:38 AM To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Crazy Usage Daniel, Statistics information, IE "cpu jitters" are constantly updated in a single place in linux, and that is proc. Thats where the HLDS "stats" function gets it from, and thats where ps, top, and any other utilitiy in linux that reports CPU % will also obtain it from this directory. This is where CpuStat gets it from also. Provided you do not bugger with the kernel, the information found in this directory its EXTREMELY accurate, and I can only think that the sheer disbelief I am getting from most of you on this list is due to a lack of knowledge regarding how the linux operating system functions. Numbers, in this case, do not lie. If you start to bugger with HZ then yes you are going to get some bizzare and unbelievable results. Provided there is no interference from some modification such as this the information in /proc will be true. To prove my point to you all, I now have 4 16 player Steam CS 1.6 servers running on this machine. I shall post results later today. James - Original Message - From: "Daniel Stroven" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, September 18, 2003 7:19 AM Subject: Re: [hlds_linux] Crazy Usage > James, > > Can you give us some "rcon stats" results on these games? When you the > servers are close to full? > > I have cpustat installed when I tested out the 2.4.9 kernel. Pretty neat, > but I would like to see some info from rcon stats if possible from you. > > Also if you can show some load stats from top for these servers. thanks > > dan > - Original Message - > From: "James Couzens" <[EMAIL PROTECTED]> > To: <[EMAIL PROTECTED]> > Sent: Thursday, September 18, 2003 3:59 AM > Subject: Re: [hlds_linux] Crazy Usage > > > > Daniel, > > > > Believe me, it can, and it does. Taking the server to larger maps does > > yield cpu above 1%, please note that I was testing the smallest map, > > de_dust. Previously posted on this list were some screenshots: > > > > http://prodigy.redphive.org/images/de_airstrip.jpg > > http://prodigy.redphive.org/images/de_aztec.jpg > > http://prodigy.redphive.org/images/de_aztec2.jpg > > http://prodigy.redphive.org/images/de_dust.jpg > > http://prodigy.redphive.org/images/de_dust2.jpg > > > > The cpu use is very REAL, and even for YES 30 slots. I have seen > identical > > results in windows. As Eric kindly pointed out previously, 30 slots is a > > relative term. A 30 slot server, with all slots bound to the same server > > would use an exponentially larger amount of cpu than two servers running > 15. > > > > Cheers, > > > > James > > > > - Original Message - > > From: "Daniel Stroven" <[EMAIL PROTECTED]> > > To: <[EMAIL PROTECTED]> > > Sent: Thursday, September 18, 2003 12:34 AM > > Subject: Re: [hlds_linux] Crazy Usage > > > > > > > Server FPS I have seen over 500. 100 is what i would ideally want. 50 > > fps > > > or lower is not what I want. > > > > > > To me its simple. Regardless of kernel or regardless of OS, a cpu can > > only > > > handle so much. A cpu is not going to just use 1 percent for near 30 > > slots, > > > unless its some super duper chip. There is way to much going on for it > to > > > use 1%, that is not even realistic to think so. > > > > > > Put that same chip on windows servers which seem to utilize cpu better > > than > > > linux currently, and you will not see 1% usage. In fact I would say its > > > down right physically impossible with the current chips..even 3ghz chip. > > > > > > If it works for you great, for me the performance simply wasn't there. > > > - Original Message - > > > From: "James Couzens" <[EMAIL PROTECTED]> > > > To: <[EMAIL PROTECTED]> > > > Sent: Thursday, September 18, 2003 2:15 AM > > > Subject: [hlds_linux] Crazy Usage > > > > > > > > > > Jeremy, > > > > > > > > I assure you, it is not incorrect. Its very correct. Kernel timings > in > > > > kernel-2.4.21 which I have also tested against, use identical kernel > > > > sleeping habbits: > > > > > > > >
RE: [hlds_linux] Crazy Usage
>http://prodigy.redphive.org/images/de_airstrip.jpg You are nutts, that is not real. Please use some common sense if you thing you can run all those players on one server and get those results. Sorry Jeremy ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Crazy Usage
Are you using the 2.4.9 kernel or the 2.4.9e? As posted by others earlier, one does not want to use the 2.4.9 kernel as it is not secure enough. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of James Couzens Sent: donderdag 18 september 2003 4:33 To: [EMAIL PROTECTED] Subject: [hlds_linux] Crazy Usage Guys, Refer to my previous post, which states, and proves that by using kernel 2.4.9 you can achieve incredibly low useage, using default kernel timings. Here is an example: AMD XP2600 1GB PC2700 DDR333: [7:19:pm] -r5-cs3- [CS3] Statistics: CPU: 1% FPS: 50 Players: 15/17 Map: cs_assault Uptime: 757 m [7:19:pm] -r5-cs2- [CS2] Statistics: CPU: 1% FPS: 50 Players: 14/17 Map: de_chateau Uptime: 922 m Those two servers are on the same machine, and the cpu reported, is the TOTAL CPU use for the ENTIRE server, so its 1% combined. So there are 29 players effectively doing only 1% cpu. And as you can see, each server is still clearly getting 50FPS which is excellent also. VERY IMPRESSIVE. Until Valve can figure out a way to obtain similar using with more modern kernels, instead of complaining further, I would switch to kernel 2.4.9. If you need EXT3 apply Alan Cox's #9 patch for 2.4.9 and then the ext3 patch. All easily found using google. James - Original Message - From: "wad asd" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Wednesday, September 17, 2003 6:22 PM Subject: Re: [hlds_linux] Crazy Usage > -- > [ Picked text/plain from multipart/alternative ] > 32players on my P4 2Ghz takes up 95% > 32players on my dual Xeon takes up 50% on on Cpu > yea.. its crap > > > Daniel Stroven <[EMAIL PROTECTED]> wrote:This is a multi-part message in MIME format. > -- > [ Picked text/plain from multipart/alternative ] > 20858 cstrike 25 0 86864 69m 9400 R 94.6 7.9 427:27.49 hlds_i686 > > wtf. > > I use to run 2 14 player awp map servers on this box since beta days..thru 1.5 as well. Now I can't run a single one smoothly? > > man thats bullshit. > -- > > > > ___ > To unsubscribe, edit your list preferences, or view the list archives, please visit: > http://list.valvesoftware.com/mailman/listinfo/hlds_linux > > > - > Do you Yahoo!? > Yahoo! SiteBuilder - Free, easy-to-use web site design software > -- > > ___ > 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 ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Crazy Usage
Server FPS I have seen over 500. 100 is what i would ideally want. 50 fps or lower is not what I want. To me its simple. Regardless of kernel or regardless of OS, a cpu can only handle so much. A cpu is not going to just use 1 percent for near 30 slots, unless its some super duper chip. There is way to much going on for it to use 1%, that is not even realistic to think so. Put that same chip on windows servers which seem to utilize cpu better than linux currently, and you will not see 1% usage. In fact I would say its down right physically impossible with the current chips..even 3ghz chip. If it works for you great, for me the performance simply wasn't there. - Original Message - From: "James Couzens" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, September 18, 2003 2:15 AM Subject: [hlds_linux] Crazy Usage > Jeremy, > > I assure you, it is not incorrect. Its very correct. Kernel timings in > kernel-2.4.21 which I have also tested against, use identical kernel > sleeping habbits: > > LINUX 2.4.21 : > > [EMAIL PROTECTED] asm-i386 # cat /usr/src/linux-2.4.21/include/asm-i386/param.h > #ifndef _ASMi386_PARAM_H > #define _ASMi386_PARAM_H > > #ifndef HZ > #define HZ 100 > #endif > > #define EXEC_PAGESIZE 4096 > > #ifndef NGROUPS > #define NGROUPS 32 > #endif > > #ifndef NOGROUP > #define NOGROUP (-1) > #endif > > #define MAXHOSTNAMELEN 64 /* max length of hostname */ > > #ifdef __KERNEL__ > # define CLOCKS_PER_SEC 100 /* frequency at which times() counts */ > #endif > > #endif > > LINUX 2.4.9: > > [EMAIL PROTECTED] asm-i386 # cat /usr/src/linux-2.4.9/include/asm-i386/param.h > #ifndef _ASMi386_PARAM_H > #define _ASMi386_PARAM_H > > #ifndef HZ > #define HZ 100 > #ifdef __KERNEL__ > #if HZ == 100 > /* X86 is defined to provide userspace with a world where HZ=100 >We have to do this, (x*const)/const2 isnt optimised out because its not >a null operation as it might overflow.. */ > #define hz_to_std(a) (a) > #else > #define hz_to_std(a) (((a)*HZ)/100) > #endif > #endif > #endif > > #define EXEC_PAGESIZE 4096 > > #ifndef NGROUPS > #define NGROUPS 32 > #endif > > #ifndef NOGROUP > #define NOGROUP (-1) > #endif > > #define MAXHOSTNAMELEN 64 /* max length of hostname */ > > #ifdef __KERNEL__ > # define CLOCKS_PER_SEC 100 /* frequency at which times() counts */ > #endif > > #endif > > As you can clearly see, HZ is still 100. > > So, please explain to me how I am being lied to? I use 2.4.21 I see obscene > usage breaking to almost full capacity of my chip. In addition to not only > SAYING its using that many cycles, it TRULY is, FPS drops, working on the > machine becomes sluggish etc... Yet when I use a 2.4.9 kernel, I see a > remarkable difference, in addition to INCREDIBLY low cpu use, its accurate, > since the FPS is high, and the server is extremely responsive. > > I am not using top to discover cpu use either, and thats even a moot point > since top agrees fully with the cpu use I arrive at, which I manually > calculate by using the proc filesystem. > > Please explain to me the logic you are using to deduce that my results are > incorrect? FPS does not lie. If the server can do 50 FPS, which is its > MAX, then I think i'm on the right track here. > > I am telling you, and PROVING to you, that using this kernel, with DEFAULT > kernel timings results in performance vastly superior to any kernel released > after. > > James > > - Original Message - > From: <[EMAIL PROTECTED]> > To: <[EMAIL PROTECTED]> > Sent: Wednesday, September 17, 2003 7:43 PM > Subject: RE: [hlds_linux] Crazy Usage > > > > >Here is an example: > > > > > >AMD XP2600 1GB PC2700 DDR333: > > >[7:19:pm] -r5-cs3- [CS3] Statistics: CPU: 1% FPS: 50 Players: 15/17 Map: > > >cs_assault Uptime: 757 m > > >[7:19:pm] -r5-cs2- [CS2] Statistics: CPU: 1% FPS: 50 Players: 14/17 Map: > > >de_chateau Uptime: 922 m > > > > > >Those two servers are on the same machine, and the cpu reported, is the > > >TOTAL CPU use for the ENTIRE server, so its 1% combined. So there are 29 > > >players effectively doing only 1% cpu. And as you can see, each server > is> > > >still clearly getting 50FPS which is excellent also. > > > > > >VERY IMPRESSIVE. > > > > That's not impresssive, it's just incorrect. There is no possible way that > > your server has 1% usage with 30 active players. I'm sorry but top is VERY > > incorrect. > > > > Jeremy >
Re: [hlds_linux] Crazy Usage
I've got a Dual P800 1GB of RAM running 2 CS servers (not mps so why not eh). It copes fine when playing dust etc, but when the heavy maps begin the CPU usage shoots up. I am going to have to replace the box, which isn't something I can warrent yet. Daniel Stroven wrote: This is a multi-part message in MIME format. -- [ Picked text/plain from multipart/alternative ] 20858 cstrike 25 0 86864 69m 9400 R 94.6 7.9 427:27.49 hlds_i686 wtf. I use to run 2 14 player awp map servers on this box since beta days..thru 1.5 as well. Now I can't run a single one smoothly? man thats bullshit. -- ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux -- Wireplay Official http://www.wireplay.co.uk/ ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux