Re: [hlds_linux] How hard would it be to make hlds
- -Original Message- - From: [EMAIL PROTECTED] - - True, but look at the intake of funds. It would more then - cover the cost. - - Just my $0.02 - - Original Message - - From: localhost [EMAIL PROTECTED] - To: [EMAIL PROTECTED] - Sent: Sunday, April 25, 2004 7:57 PM - Subject: Re: [hlds_linux] How hard would it be to make hlds - 64 players? - - - it would be hell for admins like us. you wouldnt be able - to fit more - than 4 servers on a good size box - - m0gely wrote: - Would this be a huge undertaking? I realize it's pretty - late in the - game to even suggest such a revision if it is indeed - that big. But - it could make some great new mapping possibilities and - game play. - And I'm not talking about making Dust2 64 player. :P - LOLOLOL! I didn't know hlds was ported to sparc!! That's the only way you're gonna get 4 64-player servers on one box, let alone ONE!! Hell, you can barely fit a 32 on one x86 proc under 3.1.1.0 with a few addons. Wishful thinking at best. /BA /Michigan Burbs ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
[hlds_linux] Re: Online FPS will soon be dead (Was: Transparent walls with D3D?)
-Original Message- - From: [EMAIL PROTECTED] - From: muldy - - We all know VALVe and Microsoft sleep on the same bed, - Microsoft wants OpenGL out of the way to impose (one more!) - the DirectX Api. - - OpenGL is portable, DirectX is not! Don't even talk - about WineX, Why emulate clients if you can make a client - for other platforms? - - VALVe has got an excellent game on it's hands, and does - not need to team up with Microsoft or any other major company(ATI?). - Microsoft VALVe will chose the requirements for - playing Half-Life from now on... Which is why HL2, and all it's mods will become single-player FPS games, or small LAN games run in the confines of home. I agree, I'm waiting for the day when the end will come, and it will given the choices made by Valve. I really wish I am wrong, but it doesn't look that way. Money is made from the sale of client copies, and that will be the focus. MS and Valve will bank on the fact that people will run Windows clients and connect to Windows servers. One of the driving factors for the popularity of HL so far has been the proliferation of servers available to play on, and all of the assistance available to would-be server operators. It's no fun trying to play an online FPS when you can't find a decent server to play on. For example, I have a ton of people pleading with me to open up a Nascar Thunder 2004 server on one of my OC3 links. Unfortunately, this will not happen unless EA releases a unix dedicated server. Because I, like many of you, provide game servers as a service and not a commercial entity, I have no recoup of my costs in operation. Those of you who do run commercial services have to worry about your solvency. Running Windows servers is not economically feasable. It's simple math: Windows 2003 Server, 5 license = $995.00 X 5 servers = $4975.00 FreeBSD 5.1 (my personal choice), unlimited = $0.00 X 5 servers = $0.00 Now apply those figures to a commercial entity with 100 servers, then factor in the additional management to do monthly hotfixes or service packs, and the downtime involved. Since those of you who do run commercially know the slim margins you're on now, can you afford to amortize in the costs of the OS as well, yet realize a profit during the lifespan of the game? So the OpenGL/D3D argument is really a moot issue here. It doesn't matter what method of rendering is used, if there aren't servers for the client to connect to, the networked game is dead. /BA Michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
[hlds_linux] Re: Online FPS will soon be dead (Was: Transparent walls with D3D?)
From: Kevin J. Anderson [EMAIL PROTECTED] To: [EMAIL PROTECTED] Michael Ressen wrote: -Original Message- - From: [EMAIL PROTECTED] - From: muldy - - We all know VALVe and Microsoft sleep on the same bed, - Microsoft wants OpenGL out of the way to impose (one more!) - the DirectX Api. - - OpenGL is portable, DirectX is not! Don't even talk - about WineX, Why emulate clients if you can make a client - for other platforms? - - VALVe has got an excellent game on it's hands, and does - not need to team up with Microsoft or any other major company(ATI?). - Microsoft VALVe will chose the requirements for - playing Half-Life from now on... Which is why HL2, and all it's mods will become single-player FPS games, or small LAN games run in the confines of home. I agree, I'm waiting for the day when the end will come, and it will given the choices made by Valve. I really wish I am wrong, but it doesn't look that way. Money is made from the sale of client copies, and that will be the focus. MS and Valve will bank on the fact that people will run Windows clients and connect to Windows servers. One of the driving factors for the popularity of HL so far has been the proliferation of servers available to play on, and all of the assistance available to would-be server operators. It's no fun trying to play an online FPS when you can't find a decent server to play on. For example, I have a ton of people pleading with me to open up a Nascar Thunder 2004 server on one of my OC3 links. Unfortunately, this will not happen unless EA releases a unix dedicated server. Because I, like many of you, provide game servers as a service and not a commercial entity, I have no recoup of my costs in operation. Those of you who do run commercial services have to worry about your solvency. Running Windows servers is not economically feasable. It's simple math: Windows 2003 Server, 5 license = $995.00 X 5 servers = $4975.00 FreeBSD 5.1 (my personal choice), unlimited = $0.00 X 5 servers = $0.00 Now apply those figures to a commercial entity with 100 servers, then factor in the additional management to do monthly hotfixes or service packs, and the downtime involved. Since those of you who do run commercially know the slim margins you're on now, can you afford to amortize in the costs of the OS as well, yet realize a profit during the lifespan of the game? So the OpenGL/D3D argument is really a moot issue here. It doesn't matter what method of rendering is used, if there aren't servers for the client to connect to, the networked game is dead. /BA Michiganburbs.com I dont quite understand your point? what makes you think that games will stop making linux servers? Valve certainly will not stop linux server development, and pretty much any *popular* game that has even remotely intelligent developers will have one as well. for example EA/Dice finally got around to making a battlefield linux server. pretty much everything you just said is complaining about an irrelevant, non issue in a vacuum? or am I missing something? kev Awww, Kevin... Of all ppl on this list I figured you would totally understand what I was talking about! Well, maybe I'm not that clear, so I apologize for that. My reply to Muldy was geared around his last statement: Microsoft VALVe will chose the requirements for playing Half-Life from now on... What I meant by my reply is the fact that any really popular gaming platform does have a dedicated server platform to run on multiple OS bases. With MS and Valve, it seems the focus has been working it's way toward a Windows-optimized, or eventual Windows-only system. Although I have pretty much given up on Valve delivering a viable product (I am still on 3.1.1.0c), there are a few signs of a directional shift toward MS products: 1. We've all agreed the binaries we have been given in the last 6-9 months are performing better on Windows machines, and the linux binaries have become CPU hogs, leaving many on this list frustrated and in wonder of why more hasn't been done. Granted, Alfred has told us the problem is in the compiler, as they have not seen any improvements in using different ones, but I read every post on this list, and I have seen people offer to help, make suggestions, etc, that could lead to a better linux product. I question why these offers have been dismissed, and if the code is compiled properly, why are the Win binaries performing so much better on the same hardware? 2. Microsoft's oversight of Valve, due to financial involvement is going to mean an excertion of influence to make the Windows product more appealing to consumers. Linux is competition to MS; an ever-growing cut into their bottom line. MS is not going to be party to a product that promotes Linux as superior, or even on-par with their OS platforms. The shareholders and dollars rule in this realm. 3. The topic that sparked this original thread, about the push toward D3D rendering for the new platforms
Re: [hlds_linux] Man, cheating has never been so bad for my server. VAC, are you still there?
- -Original Message- - From: dual_bereta_r0x [EMAIL PROTECTED] - Citando Ryan Schulze [EMAIL PROTECTED]: - - switch to a different game, probably bf1942. - - We are having a great success with Wolfenstein: Enemy - Territory and Call of Duty Game servers. Both are full 24x7, - didn't have more cheaters than players, are fast even in - Windows platform as a server and people are having much more - fun than our CS players. - - -- - dual_bereta_r0x -- Alexandre Hautequest - ArenaNetwork Lan House Cyber -- www.arenanetwork.com.br - ICQ 126063524 - - We are having success with BF1942 and the desertcombat mod as well, although our CS 1.5 servers are still full most of the time. It's an even split right now as far as usage. Best A/C idea up yet is BF1942 1.6 will have an integrated 3rd party system. Those of you here on the bf mailing list know what I mean ;), plus there's now support from UA on a system as well. What's better is at least those servers are UP, and not stuck in limbo trying to update themselves. =) /BA michiganburbs ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] CS 1.5 without valve/won ?
- -Original Message- - From: Jonathan M [EMAIL PROTECTED] - - I have though of a way to keep the 1.5 community active - - But developing a fake won server - - Which unfortunatly woudnt be able to block bad keys - - I belive Valve shoudnt stop serving WON keys - - as most people will continue to play the old games for a long time - - At least until steam is a stable and well developed product That would be best for everyone, however I think it's clear Valve's intent to move to Steam, as in theory (!?) it offers them an advantage of easier distribution and lower costs of operation, as well as a unique ability to manage upgrades and patches and valid installs. I can't argue with the benes of Steam - properly done it will have it's place. However closing the WON system is more of a move to shed costs associated with operation and maintenance, and forcing clients over to Steam. As a shot-in-the-dark, I sent Eric an email offering to aquire the entire WON network, including the cd-key database and crypto key used in the auth mechanism.I truly don't believe they'll even consider it, but what the hell, ya never know - they might be strapped for cash when they get their bandwidth bill for this month! /Michael ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
[hlds_linux] RE: Not Upgrading
I can completely understand those of you with more patience than I have. I appreciate Brian's feeling that things will eventually work themselves out, too. It just that this isn't a 2 or 3 day issue. The members of this list have been voicing their concerns to Valve about many different issues during the course of the last 9 months since the 1.6 beta came out. We've made issues about the updating, the cpu loads, asked for the courtesy to be able to test this stuff in advance of the client rollouts. We've offered our insight as to features, compared the benefits of the updates, and more. We aren't here to bash Valve, we're here to help it succeed by giving input they can't get from anywhere else. It's after these months of bringing up issues and telling them when we (and the userbase) want and watching our pleas be either ignored or deferred. At this point I feel that we've been cheated, and this whole upgrade was handled badly, despite warnings from members of this list. Some of us are in business to make money hosting, some do it for fun. Those who profit from hosting or renting are the ones that will be hurt the most. What about the LAN party companies and coordinators? This whole thing is akin to telling your kids not to touch the stove when they're young - then watching them do it over and over again. I personally feel we've been ignored, put off, disrespected, and since I have no ties to Valve I can make a choice. Everyone can. For purposes of comparison I know a few of you are also BF1942 admins, and a couple also subscribe to their linux list as well. Tell the rest here: Has Andreas (or DICE) ever treated us in the same respect that Valve is treating us here, not to mention their whole userbase? I just didn't want some of you to think this is an impulsive decision, just look at what's gone on in the past 6-9 months and it's pretty easy to tell that with some communication, listening, and testing all of this could have been avoided. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] CDeath Problem
-Original Message- From: Jonathan M [EMAIL PROTECTED] Why are you running cheating death as a metamod plugin Read the information in the readme on how to install it Regards Jonathan Jonathan, CD can be installed two ways: As a Metamod plugin, or as a standalone plugin. Refer to the install instructions at docs.unitedadmins.com and you will see both well described. Regards, Michael Ressen, Michigan Burbs Network Administrator, UA Staff www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
[hlds_linux] High in game pings
[EMAIL PROTECTED] previously said: Hello, I am running 3.1.1.0c with boffix on a dual amd mp 2400 with a gig of ram running multiple servers. Running FreeBSD 5.1. I setup one server for 32 slots. My out of game ping to this server and all others on this machine is from 55-75. In game it is about the same till server reaches about 18 people playing. And from there on out the in game ping increases with each player. At about 30 players my in game ping is about 120-130 while at the same time the out of game ping is still normal at 55-75 and all other servers on the machine are still running normal pings. Top shows the full 32 player server using about 55% of one cpu so the machine is not over loaded. I have tried different pingboost settings and even no pingboost but all have the same problem. I even disabled metamod and all its plugins and still no go. Any idea on what to do to correct this? I used to run a 24 player server a month or two back and I don't recall ever having this problem when its full as this starts happening at 18 players on the new one. Thanks Jeremy How did you compile the kernel? Did you use OPTIONS HZ=1000, or some other value? Also, which scheduler have you used - the 4.x one or the new ULS one? What about some of your sysctl values like machdep.cpu_idle_hlt' and 'machdep.hlt_logical_cpus'? These 2 can be really important for the HT that goes on in conjunction with the SMP kernel. I've done a bit of experimentation and checking, and I also run 5.1 on a dual 2.8 xeon box with these values: OPTIONS HZ=1000 Machdep.cpu_idle_hlt=0 Machdep.hlt_logical_cpus=0 Compiled with ULS scheduler, and stripped clean of everything else. Then I don't use -pingboost 3 anymore, I just use -1 because the increase to the kernel HZ really makes a massive difference. I'm able to run a 32-player CS server with 3.1.1.0c + boffixv2, MM 1.16, HLG with the aimbot checks on, and amx with a bunch of plugins and pings in-game only increase about 10-15ms total when full. The hlds process chews up about 80% of one proc, and I still have a 40 player BF1942 server and another 20 player CS server on the same box. Mine is really smooth in-game, and will support players with high rate settings all the way up. One thing I notice does increase pings is cs_office and de_inferno. Those two maps really eat up resources, but it seems to be on the client end more than the server end. Yes, pings climb much more on those maps, but they don't come close to doubling. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Possible addition to help catch cheaters.
[EMAIL PROTECTED] previously said: Now that is a WHOLE lot of work to catch cheaters but an interesting idea. Mind sharing the meta plugins? Jeremy Under Adminmod 2.50.52 it was a plugin called admin_ss.sma, and under AMX 0.9.3 there's actually a couple of them. The stock ones tell the client about the screenshot, so you have to remove that code. Email me OT and I can probably send my edited .sma's to you. And as Stefan says, theres ways to intercept the call, and detect it, but for a while it was a pretty good way - we caught a lot of cheaters amongst our regulars who were tearing it up. If no one was suspicious or complained about them, we didn't hassle them at all, but a few complaints and we'd take a few SS's. Still, when we got one - the evidence could not be disputed, and we would post the screenshot on our web forums for everyone to see. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
[hlds_linux] Trim your posts
Guys... Trim your posts please. Use -snip- if needed, just make them shorter and cut out the useless crap! Tks! /Michael ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE:[hlds_linux] Patch Boffix for 3.1.1.0 hlds
gault [EMAIL PROTECTED] previously said: Hi all, Could the creator ( i want to thanx him for his work first) answer me for a= bug in patch. When you try to changemap in rcon changelevel de_dust2, in g= ame the server crash ... auto up but sometimes need 3 or 4 times to changel= evel its pretty annoying :p Have you got some feedback from that ? Specs: Dual 2.8 Ghz Xeon DELL 2650 OS: FreeBSD 5.0 hlds : 3.1.1.0c with boffix loaded Waiting for you :p Cheers, S.Gault I have almost identical hardware, FBSD 5.1, 3.1.1.0c + boffix v2 - rcon changelevel does not crash mine. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Updated - 3.1.1.1d Performance
-Original Message- From: Daniel Stroven [EMAIL PROTECTED] I have never manually set the sys_ticrate. I have pingboost 1 enabled and Im pretty sure the kernels are compiled at 1000hz. This is on RH. Should I try putting +sys_ticrate 1000 in the command line? I would as long as your kernel is compiled with it. I'm not sure bout redhat's defaults, but if ya wanna give it a go - throw it in there and see what happens! Can't hurt. /Michael ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
[hlds_linux] RE: hlds_linux digest, Vol 1 #2508 - 9 msgs
James Couzens [EMAIL PROTECTED] previously said: Wait I have more to say on this subject. rantI think its shitty that there are all these sites that post so publicly open notices of exploits, and often the code that goes along with this. But then again, it seems that when software companies are informed that there is a problem, they tend to ignore it when it comes to some insignifcant security bug. So perhaps the hackers are right to release it into the wild, so that we all get raped over this and bitch and complain and then voila said expoit is fixed! But does it really need to come to that? Can't software companies be privately informed, and immedtiately put forth a fix? Then problems like this would be few and far between. Perhaps its too late in the game to change the way things work here, but meh, just my two cents./rant James - Original Message - From: Scott Pettit [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Wednesday, July 30, 2003 5:02 AM Subject: RE: [hlds_linux] HLDS Expolits 3 Servers have been killed of mine now and as I type this another 2 have just been killed. Hurry :/ Now I'm officially pissed. Every 20 mins I have 4 servers getting hit. There is no solution, and I also share the fear of someone exploiting a shell so until this is fixed all my hlds servers are offline now. Valve - Someone must address this issue NOW. This is a top priority, as it potentially compromises security of our servers. Please let us know that you are aware of the situation and it is being worked on. I could care less about a timetable right now, just tell us something IS being done!! Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] _working_ workaround for hlds exploit on
-Original Message- From: john [EMAIL PROTECTED] On Wed, 30 Jul 2003, Michael Ressen wrote: Haspers [EMAIL PROTECTED] previously said: Also available at http://dl.gameservers.net/patch/boffix.tar.gz Mirrored at http://www.michiganburbs.com/boffix.tar.gz on OC192 as well. Go 4 it. Whats the bet that the server would die before you can get that much bandwidth out of it? john 100%. lol that box is just a little ultra5, but I figured a little load testing wouldn't hurt! ;) ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
[hlds_linux] Update on this fix that's commin today???
It's getting late in the day. What's goin on? Eric? Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Update on this fix that's commin today???
-Original Message- On Wed, Jul 30, 2003 at 05:49:26PM -0500, Michael Ressen wrote: It's getting late in the day. What's goin on? Eric? Michael Ressen, Michigan Burbs Network Administrator Fuck sake, shut up. -- James. Really, James. Glad to see you're not in any way concerned about the issue here. I also don't think it's unreasonable for us to get some sort of update. 50% of the public servers are either vulnerable or offline, and a few of mine are offline as well. This is a major issue and with a security risk this big, I for one would like to know what status is, even if it's still not ready. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] Half-Life Primary Server x.1.1.1d
-Original Message- From: Eric Smith [EMAIL PROTECTED] What are the main issues with updating to v3.1.1.1d? CPU usage? -Eric Increased CPU loads are the biggie, Eric. Figure it this way: You size your servers and plan your capacity based on 3.1.1.0x versions, which have been the standard for quite a while. Now, migrating to 3.1.1.1x reduces your overall capacity by not just the ~20-25% that you see in increased load, but can cause even bigger probs for those who run multiple hlds on a physical box. For example, my biggest server is a 2x2.8 Xeon machine, using FreeBSD 5.1, with hyperthreading enabled, and the HZ compiled in at 1000. All extra stuff is stripped out. Since hlds is only single-threaded, the scheduler is running tasks across one cpu at a time, although the scheduler is alternating between phyical cpus to balance the load. Now, here's my hlds command line: ./hlds_i686 -noipx -game cstrike +ip 64.108.152.130 +port 27018 +maxplayers 32 +servercfgfile b1server.cfg +map de_dust2 -pingboost 3 +sys_ticrate 1000 -tos. I have VAC, metamod 1.16, hlguard 1.50 using aimbot, cvar, and file checks, and amx 0.9.3 with 21 plugins. Watching my 'top' output right now here's what I get at 10 sec intervals on cs_assault: CPU % (1 proc) # players --- 74.22 26 71.09 25 77.34 25 80.47 27 82.03 27 82.81 26 83.59 26 If this was cs_office, or de_torn, she'd be pegged at 99%, and there'd be some much higher pings in there. My only recourse is to turn the pingboost down, which is going to result in adding +10ms or so to all players and it won't be as smooth. Now imagine running an identical server on the same box. With 3.1.1.0x I can, but 3.1.1.1x I probably cannot. So, that puts me in a predicament; I can change my setup and lower the load a bit, but my patrons won't like it. I'll lose people doing that. On the other hand, to keep 2 servers running, at the same level of performance, I now have to go out and spend another $2200 on new hardware, plus space fees, additional router port, etc. The costs to me for just one server are more than I am going to undertake, so you can hopefully realize some of these guys here with 100 servers, who might find they no longer have the capacity to handle the load. They're out revenue. We do like all the things that you folks have put into the game, and that's what's helped make it so successful, but we get frustrated when we look at some of the other game engines out there and compare. For example, that same box that hosts my 32 player CS server also hosts a 20-player rental. It also hosts a 20 player BF1942 server. Right now, that server is full, running coral_sea, and the load is 30.47% (1 proc). The efficiency of that binary is way ahead of the hlds binary, and it uses 2.4 times the bandwidth player for player. I'm no super-duper programmer, I'm just some dumbass network engineer, so I'm sure I'm gonna get flamed for this, but I'm also sure that there will be more than a few people who will echo my statements here. Anything that can be done to aid in the load issue we're facing is going to be the most welcome relief to us. Thanks! Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Cleanup CDKEYs
-Original Message- From: Simon Garner [EMAIL PROTECTED] You cannot ban a player just for using the same key as another player. Sure I can. Key-sharing? No, I don't believe it. Multiple keys are used for one thing: cheating. Anyone I see flip flop between keys, I'll ban them on sight if it comes from the same ip. ___ 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 CS map load
Steven Hartland [EMAIL PROTECTED] previously said: mapplayersavgcpu% de_dust21625.464137 de_aztec1632.970308 cs_italy 16 36.829508 de_cbble16 37.502273 de_train 16 38.70 de_prodigy 16 39.395652 de_nuke 16 39.771429 cs_747 16 40.585 de_inferno 16 45.45 cs_office1645.908333 Do these results show what every 1 else is seeing. This is CPU% on a 2.4Ghz Xeon 2Gb RAM sample spacing 1 min running metamod + wwcl I've tried without metamod and the was no deserble difference. Now were as de_dust2 load is excellent the rest leave something to be desired they are now significantly worse than bf1942 I could run a 40 player bf server for the CPU used on cs_office. Steve / K I haven't even tried running steam versions yet, but those ratios are pretty close to what to get. Here's 20/32 players on starwars: CPU states: 14.7% user, 0.0% nice, 2.2% system, 0.5% interrupt, 82.5% idle PID USERNAME PRI NICE SIZERES STATE C CPU COMMAND 8537 cstrike 1130 81700K 77496K CPU2 2 53.12% hlds And with 30/32 players: CPU states: 17.3% user, 0.0% nice, 2.0% system, 1.5% interrupt, 79.3% idle PID USERNAME PRI NICE SIZERES STATE CCPU COMMAND 8537 cstrike 1100 88140K 83972K *Giant 0 67.97% hlds I get about the same as this on office and inferno, less on others, much less on dust and dust2. Here's 23/32 on italy: CPU states: 12.6% user, 0.0% nice, 2.3% system, 1.5% interrupt, 83.6% idle PID USERNAME PRI NICE SIZERES STATE CCPU COMMAND 8537 cstrike 1140 97396K 93256K RUN2 50.78% hlds Pretty much the same spread as you have, just some different values. I believe it has to do with the number of, and complexity of the sprites used in the map. I may be wrong though. Dual 2.8Xeon/512/400/FreeBSD5.1-RELEASE(CustomSMP+HT)/Hlds 3.1.1.0c(ticrate=1000 + updaterate 100)/amx0.9.3 + 21 plugins including PTB + RBBoT's mapvoting + stats logging/hlg 1.41(aimbot + cvar)/VAC/Metamod 1.16/remote logs. /Michael ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
[hlds_linux] FreeBSD 5.1 SMP + Hyperthreading Results
I figured I should share some of this with you, since it's downright amazing: I recently upgraded the OS on one of our CS servers from FreeBSD 5.0-RELEASE to 5.1-RELEASE to take advantage of the new SMP features, and the new scheduler. The server is a Dell Poweredge 1600 series with dual 2.8 Xeons, 400 mhz bus, 1.12G PC2100 ECC 2.5cl ram, and a U320 PERC controller with a Seagate Cheetah 15k U320 drive. The Before setup was a single Xeon on FreeBSD 5.0 (HZ=100), running sshd, and hlds 3.1.1.0c(sys_ticrate=1, no minrate or maxrate, sv_maxupdaterate 100) + amx 0.9.3 (21 plugins), HLG 1.41(11000), as a 32-player single CS server. Since BSD 5.0 only had the 4.x-version kernel scheduler, that's what was used, on a pretty generic kernel. With 32 players, running CS on a high-load map such as cs_office, typical loads were reported by 'top' to be around 85%-90% CPU, with generally stable pings, and few spikes, even during heated conflict. Total system load, which includes other essential services shows to be the same, at 85%-90%. Maps such as aztec and dust would load in the mid-70's range with 32 players. The After configuration includes a 2nd Xeon, an upgrade from source to FreeBSD 5.1 (HZ=1000), enabling the Hyperthreading features in the Xeons, and using the newer (and still experimental) ULE scheduler. FreeBSD's has support for HT, but it sees the HT as physical CPU's, so there is a switch that actually halts logical CPU's by default as so not to cause poor performance because they might compete for resources. I have enabled the logical CPU's on my system. With 32 players and no changes in hlds configuration, 'top' is reporting hlds load on cs_office to be the same 85%-90% on one CPU. Overall system load shows to be 18% total. On dust and aztec the per-cpu load is the same, but overall load is 11%-12%, with no detectable spikes. (short pause) Yes. You read that correctly. 11-12%. Another thing that is noted is that while hlds is not a multithreaded or SMP aware process, it doesn't continually run, so it runs-pauses-runs-pauses. This is allowing the scheduler to shift the load to an unused cpu. While this isn't taking advantage of the chip's cache to improve speed, it isn't forcing a perticular CPU to become the processes host. Perhaps after BSD gets some better support for SMP we will have the ability to control processor affinity, but for right now, I could care less - the performance increase is f***ing amazing! Another point of note, using the linux compatibility mode in BSD does not affect hlds in the manner that some linux kernels do, where you see the cpu loads shoot up to 70% with no players playing. I have found one of our machines running SuSE does exhibit that problem. Stability so far looks really good, and I would recommend 5.1 for anyone who is going to use it strictly as a gaming server, despite the fact it is not considered a stable release yet. Also, end of September is the target date for the next major snapshot of the 5.x branch, and we expect further improvements in the kernel scheduler and SMP subsystems. Sorry if I sound like a commercial, but I'm just thrilled with the results so far!! Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] RE: hlds cpu spikes under no load
Steven Hartland [EMAIL PROTECTED] wrote: You don't need to be in single user mode I do ours live a great starting point is: http://home.nyc.rr.com/computertaijutsu/cvsup.html Scary.. More scary than I wanna know! But.. I am gonna test that on one of my local boxes here. But as far as my production box, I bought a ticket and another CPU for it, so I'm gonna stay a week and make a vacation out of it while I'm there! =) Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] 3111c
[EMAIL PROTECTED] previously said: Is the majority of people running this 3111c? Or are you sticking with the 3110beta ? I've been watching and see different opinions and info posted regarding the differences and it seems that CPU usage is still a concern regarding these new releases. Is everyone just accepting it and suffering performance loss or staying with 3110beta? 3.1.1.0c on all servers for the following observed reasons: - Stability is equal or better - No additional features warrant upgrading - Performance is equal - Resource consumption is less - Better playerslot/MHZ ratio - Works with all my existing plugins and mods - 7 month proven track record Point of note, if 3.1.1.1x was required in it's current form, hlds would be gone from my servers in favor of something else. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
[hlds_linux] RE: hlds cpu spikes under no load
Sorry I can't quote anyone here - I've dumped my trash, but recall the discussion about hlds sitting idle, then punching up to these unreal cpu loads, and back down again?? Well, I did some testing, by accident. =P I had done some monitoring to see if the issue affected my servers. So, using my test rig (PPro200/FreeBSD5.0) I started up hlds and both logged the cpu load and watched it run on top. Nada. Didn't do jack. Just sat there at 0.76% like nothin bothered it. Ok. Fair enough. Might be some differences between that and the live server (IX2.8/FreeBSD5.1). Nope. That machine did much the same thing, although I had to wait to 5am to find it with no one playing. So, figuring it wasn't affecting my servers, I dropped the issue. Until yesterday. I installed amx onto one of the other units. This one is SuSe with the 2.4.18-64GB-SMP kernel, compiled under gcc 2.95.3. After pluggin in amx and removing adminmod and statsme, I sat n watched. 85% with no one in there? WTF?Back to 2%, then 45%. It was weird. Thought it was the amxmod.. So I pulled that out and just ran it clean - no plugins except metamod. Same thing. Back to adminmod/statsme/hlguard. Same deal. Hlds 3.1.1.0c in all cases I could try had the prob with the 2.4.18 linux kernel, but not with the BSD5 kernels. Is anyone in position to verify what I have noted? With 3.1.1.1x? With different kernels? With same config as me? Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] RE: [hlds] 32 players maps
-Original Message- Message: 8 Date: Fri, 6 Jun 2003 14:21:23 -0700 (PDT) From: Tony Di Schino [EMAIL PROTECTED] Subject: Re: [hlds_linux] RE: [hlds] 32 players maps To: [EMAIL PROTECTED] Reply-To: [EMAIL PROTECTED] --- Michael Ressen [EMAIL PROTECTED] wrote: About 3 weeks ago, we went from a couple 20-man servers to a 32 player one. I ended up modifying quite a few maps with additional spawn points. Here's the list of what I have that's playable now: are you hosting them for DL? If so, I would love to refer to your URL for other admins that come looking. Else, I I would love to add to my files section (even as a mirror), email me off list if you are interested. T --- I've put up 31 maps in 32-player versions in the downloads section of our website. If you're a CS server admin, and you'd like any of them, feel free to grab what you want! If you have requests, email me - there are more on the way, so check back in a few weeks to see what else I've added. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] Monitoring HLDS Servers with MRTG
Byron Carceller [EMAIL PROTECTED] previously said: From: Byron Carceller [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] Monitoring HLDS Servers with MRTG Date: Mon, 9 Jun 2003 16:44:14 +1000 Reply-To: [EMAIL PROTECTED] Oh, this script of yours only supports 3111? Damn, We're all running 3110c. Is there any other way I can go about this, or other script that's been made to work with the 3110c binary somehow? Yeah, I found that out too whilst I was up late trying it out! No worries though, Brian's script does work well on my test 3.1.1.1 server. I'll just hang onto it, because we all know sooner or later we're going to have to make that switch. As far as MRTG or RRDtool based logging goes, I was actually contemplating such a thing with MRTG in conjunction with RRDtool anyway. In my idea, I was going to turn on SNMP on my servers, and just query info that way. I can also keep it secure in that manner by filtering the snmp requests thru my router and firewall ACLs. If I come up with something for 3.1.1.0, I'll be happy to share. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] HZ=1000 and CPU-usage on FreeBSD 5.0
-Original Message- Date: Sun, 08 Jun 2003 11:09:42 -0300 To: [EMAIL PROTECTED] From: Capriotti [EMAIL PROTECTED] Subject: Re: [hlds_linux] HZ=1000 and CPU-usage on FreeBSD 5.0 Reply-To: [EMAIL PROTECTED] It increases CPU usage for me too. I was using 5.0 but the high CPU use for NAT made me give it up, some time ago. Maybe now it's better. Looks like latency is lower under FreeBSD 4.8-stable as well, according to my customers. Yeah, at the moment 5.0 is still catching up in some areas, but it should quickly improve in areas it needs too. Tomorrow there's a scheduled press release for FreeBSD 5.1. It should be on the ftp's already. There are some advancements, bug fixes, but most importantly - better SMP support. I would suggest checking it out, I plan on it. Here's the release schedule: http://www.freebsd.org/releases/5.1R/schedule.html. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
RE: [hlds_linux] nat issues
Doing what you say below and it still doesn't work? Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com From: Darren Mansell [EMAIL PROTECTED] To: '[EMAIL PROTECTED]' [EMAIL PROTECTED] Subject: RE: [hlds_linux] nat issues Date: Mon, 2 Jun 2003 16:27:35 +0100 Reply-To: [EMAIL PROTECTED] Don't know if this makes a difference but im running squid proxy on port 3128 and DHCPD which I had the idea that iptables may be getting confused with. I am running hlds exactly as you say there. ./hlds_run -game valve +maxplayers 12 +map boot_camp +ip 192.168.0.2 +port 27015 the router is setup to forward the ports to that ip on the same port number. Thanks for the help so far everyone ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] HLDS 3111c
LOL!!! admin_unllama Britt There. Now maybe you'll make some sense when you talk. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com From: Britt [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] HLDS 3111c Date: Mon, 2 Jun 2003 13:09:57 -0500 Reply-To: [EMAIL PROTECTED] Apperently more so than you do. - Original Message - From: Michael Ressen [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Sunday, June 01, 2003 9:49 PM Subject: Re: [hlds_linux] HLDS 3111c WHAT? Are you on glue? Explain to me how micro$oft, let alone, windoze has ANYTHING to do with this??? Did they just buy Valve? Or did they covertly purchase (or steal) RFC-768 recently without my knowledge (or bid)? So how (I cant wait for this answer) would changing the way they handle UDP have ANYTHING to do with what we're discussing? We're not talking about pings here, nor are we discussing routing. This is a simple matter of CPU usage. DjoDJo was talking about the fact that with the huge increase in cpu load, he can't run well because the insane resource requirements cause his servers to melt - thus increasing pings to an unplayable level. The problem is not the ping, it is the need for clockcycles. ISP's and latest routing technology? I'm sure whoever is in charge of the ISP's decisions on routing policies has a pretty decent handle on what's best for their organization. If you were familiar with this concept, you'd know there isn't one cure-all for routing, protocol or otherwise. And valve? Slap them 100M and get a low-ping server and network? Um, Valve isn't a charity organization, they're a company, in business to make money, and throwing money at something doesn't guarantee that it will either work, or be fixed (windows anyone?). Ok, Im sorry for the flame here, but calling the discussion BS and telling people to chill out because it makes you sick really isn't too proper for a listesrv, especially when you don't know what you're talking about. Ok, back to my little cave. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] HLDS 3111c
Actually, I agree with you, but in a different way. Game companies generally hate anything that isn't windows. For example, when was the last time you saw a port of a game for MacOS?Same thing with all these server ports for Linux. Where's native *nix support? What if I wanna put up 200 CS servers on my Sun Fire V880, and pipe em thru my OC-12?? I have access to several Sun boxes that are just sitting around in racks doing nothing - and they're in the best places to be: Regional POPs of backbone providers. It's a shame I can't use them for this. I understand it's all in the bucks. Make for what is popular, and what everyone has, not for what's best. I for one would love to see this work under sparc architecture. Imagine the companies who do server rentals. The idea of running a high I/O game server engine, in multiple threads, on a 64-bit OS, using a 64-bit processor(s)!! Instead of purchasing 50 x86 boxes to run Slack, we could be using the scalability of Solaris, HP-UX, AIX, or Tru-64 on hardware designed for running servers. As an administrator, I'd much rather manage 2-4 servers running a combined 50 halflife mod servers than 50 separate boxes. Reduced management, reduced admin interface, better user management, less cost per virtual server, better scalability, and better performance. I mean, how many of you have gone off your own servers to play some games on other people's servers to see how they feel? I do, quite often. I check out all sorts of stuff, then compare it to my own. Now BSD is a sorta different animal, because it runs on many different platforms, so logically that would be the best choice to start with. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com From: =?iso-8859-1?Q?Jonas_Andr=E9n?= [EMAIL PROTECTED] To: [EMAIL PROTECTED] Subject: Re: [hlds_linux] HLDS 3111c Date: Mon, 2 Jun 2003 22:48:00 +0200 Reply-To: [EMAIL PROTECTED] just a q while you at valve ar compiling for different 486/686/AMD couldnt you do a native FreeBSD port too? :) im not a programmer so it might be harder than i think.. but i would be really happy if that could be done. And if you would do it would us freebsd users see any big gains from using native freebsd instead of linux emu? //root ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
Re: [hlds_linux] HLDS 3111c
I am well familiar with 1855. You are right, I did go off on him, and poorly formatted it, and it was rude to the rest on the listserv, to whom I apologize. About the UDP handlers, what I was referring to was Britt's assertation that if microsoft changed their handling of UDP, we would have an improvement. My reply was strictly to do with the fact that we weren't discussing how the kernel is affected by UDP, but by the massive cpu needs with 3.1.1.1. I was simply taken aback at the statement about it being microsoft's fault, or partial fault. I am well aware of the implications of UDP on a kernel, and wasn't in any way trying to imply it has no effect. That effect however, isn't the cause of our problems. Michael Ressen, Michigan Burbs Network Administrator www.michiganburbs.com Subject: Re: [hlds_linux] HLDS 3111c From: Jeremy Brooking [EMAIL PROTECTED] To: [EMAIL PROTECTED] Organization: Date: 03 Jun 2003 11:14:34 +1200 Reply-To: [EMAIL PROTECTED] On Mon, 2003-06-02 at 14:49, Michael Ressen wrote: WHAT? Are you on glue? Explain to me how micro$oft, let alone, windoze has ANYTHING to do with this??? Did they just buy Valve? Or did they covertly purchase (or steal) RFC-768 recently without my knowledge (or bid)? So how (I cant wait for this answer) would changing the way they handle UDP have ANYTHING to do with what we're discussing? We're not talking about pings here, nor are we discussing routing. This is a simple matter of CPU usage. If youre saying (and it appears as such) the way your kernel handles UDP, and the way HLDS handles UDP, has no effect on CPU usage, then you are the one whos retarded, and need to shit down and shut the fuck up. Perhaps start by reading RFC1855. ___ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds_linux
