RE: [hlds_linux] DUAL CPU (smp) problem or not?
/dons nomex suit That's something I noticed too. In fact, even under 1.4 and 1.5, we had horrible performance from a monster of a machine, a quad Xeon 550 MHz with 2 MB cache. It was an IBM Netfinity box, 4 GB RAM, SCSI RAID array, etc. Despite all that expensive gear, Aztec was unplayable starting at 12 players, with 99% CPU common, with no other game servers running. Of course it was pegging only 1 CPU, meaning it was really only running as single 550 proc. However, at a lan last week we had 14 players on aztec under 1.6 running on a p3 600 dedicated server, and we never saw over 50% usage, even with all the plugins and VAC (we had internet). I'm not sure how to explain a disparity in use like this, but I think SMP is certainly a strong candidate for investigation. Why run single-threaded app under an multithreaded environment, unless you use win32 and processor affinity? In any case, the CPU usage issues we had with the quad Xeon box disappeared when we conslidated all that silicon onto one P4 2.53. Of course, on this list, such talk is heresy. But I'm pretty glad we didn't go SMP...1U blades running single procs are cheap to build, and save $$$ on rackspace :) -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] The CPU usage issue.. what should I go with: Linux or Win32?
James Couzens wrote: >Better yet, why don't you? I'm hoping that if people keep asking that question enough that eventually some bright person will get annoyed and answer it, so that we can all stop watching and get back to killing each other. Troy, comma delimited would be fine :P -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] The CPU usage issue.. what should I go with: Linux or Win32?
Daniel, I truly do apologize for your lack of basic education. It's breaking my heart. The rest of your post was all hot air, so I will let it vanish as such. Let me know when you have something to add to the discussion, besides your own exceptionally high opinion of yourself. Would anybody like to talk about any successes they've had with 1.6? Anything that's worked for them so far? -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] The CPU usage issue.. what should I go with: Linux or Win32?
> You are so painfully wrong, stop pestering the list with such false bullshi > t, > windows DOES RUN HLDS BETTER AT THE MOMENT. Wow, this is like batting practice... Sindre, How about you write us up a little white paper on the differences between the way Task Manager, top in it's various revisions, and the hlds stats command measure, interpret, and display the actual, valid, real time system usage, load, and statistics of processes running under each and every one of the varied OSes, builds, kernels, platforms, omg I could go on all day but I'm farking tired but you get my basic point, then maybe, I'll listen to what you have to say about CPU usage as the definitive god's own truth and be silent. k? So if you have a tool that will give us this information, then speak up. If not, then go back to watching your CPU(s) spike. -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] The CPU usage issue.. what should I go with: Linux or Win32?
In reply to: > I can't do much about the number of results (I've tried to make it as > easy as I could with that version checker thing) but tell me what you > want it to sort by and I'll make it happen. Troy, thank you very much for setting that up. It's slick and very useful, it just needs a view all funtion and a sorter. And a ton of responsible, detailed submissions to it's database ;) It's important to set a standard, as in, are the numbers you're posting an average or a peak, how many iterations of the test etc. In this case the variation of the open source model is working against us. In return, it's forcing us all to learn some things about our systems that we might never have probed otherwise. Hell, I got into computers this way. I played Q2 one day almost 6 years ago, and I never got tired of figuring out how to make it better. Games come and go, but being an admin is a way of life. Anyway...enough nostalgia. What would be really be the heat would be for a 1337 coder to design an HLDS server benchmarker. Something with normalized results would be nice, with win32 and Linux/BSD distros. I bet Valve has a copy lying around the office somewhere :-/ -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] Pingboost
> Hi! > > Is it possible to get an *explanation* what -pingboost 1|2|3 actually DO? > http://www.hlplayers.com/pingboost.html -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] The CPU usage issue.. what should I go with: Linux or Win32?
> > A few? I think we are the majority here. > An observation on human psychology: negative events are more often recollected than positive ones. From my perspective, I see a lot of people getting positive results here, enough to know that I'm far from alone. From the perspective of an admin having CPU issues, they see a lot of people having problems. I guess we could take a poll, but the effort at quantifying data would be better spent on something more productive. That performance database that was started here was a good idea, but it needs to be enhanced so that it correlates the CPU usages with the platforms. If this were taken systematically, it could be fairly easy to determine which platform/settings/software gives the best results. Flaming back and forth and carping about a _free product_ is not going to help anyone. Out of curiosity, if it turns out that your platform of choice was lousy at running hlds, would you switch to one that was better? [Mage]-BadPing- www.ClanMage.net ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] The CPU usage issue.. what should I go with: Linux or Win32?
> Sorry James, but I have to deny it, because it's true. I've spent more Doh. Broke my own rule and wrote an email before my first cup of coffee. I obviously meant, I have to deny it, because it's false BTW, that's with MM 1.16.2, AM 2.50.57-beta, StatsMe 2.71 and VAC running, with pingboost at -3. ...coffee ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] The CPU usage issue.. what should I go with: Linux or Win32?
> Lets not deny that this release is using A LOT more cpu than the last. > Its pretty much FACT - unless you happen to be one of the lucky few with > a SUPER KERNEL that runs 5% cpu with a 32 player serverand we all > know that's bs. Sorry James, but I have to deny it, because it's true. I've spent more time looking at top on full gameservers in my life than I care to admit, and 1.6 runs as well or better than 1.5. 18 players never even spikes past 35%, even on office or aztec, running a single P4 and Slack 8. Terribly sorry for the issues you and a few others seem to be having...but don't say it can't be done. Best of luck to you. Oh, and to the original question in this thread...I wouldn't hang a winblows box out on the Internet unless I was a masochist ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] ./steam -update and +autoupdate and multiple servers
> I must be doing something wrong but I don't see the drawback yet. > > I running about 12 servers with 4 games each, every update is done under one > account and each hlds process is started under a diffrent linux user login. > So one steam account and 48 diffrent users running it. Everything seems to > work and update fine. > > What am I missing? > I'm doing it similarly, with one steam account and one process user per customer. Works fine so far. There's little chance of customers getting our steam account info, since they don't have shell access, and pure-ftp is masking .files. (assuming the ClientRegistry.blob has some encypted info in it, in addition to the plaintest email account name) -BP -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, > >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
