I shouldn't speak for the list, but I have played on some underpowered 8v8 L4D2 servers -- When there is a large horde of zombies that gets pipebombed... It may affect some of your other servers. I don't know much about that processor, but try and load up those configured servers beforehand just in case.
GL and have fun! On Fri, Jun 18, 2010 at 7:16 AM, Josh Bost <dislexs...@dislexsick.com>wrote: > Thanks for the info guys, much appreciated :) > > And now I can stop hunting down the source of my 45% choke on the LAN > server while testing it. Didn't realize bots were so CPU hungry > :yikes: (~45 FPS: 1 server with 1 human/31 bots) > > If anybody has benchmarks for similar setups though, I would still > like to see them if possible, though I'm thinking 3x24 player TF2 + 1 > stock + 1x16 player L4D2 server across 2 cores should run just fine > (There will be other servers on the other cores, Trackmania, UT, etc) > > Cheers, > > Joshua 'Dislexsick' Bost > > > On Fri, Jun 18, 2010 at 6:53 PM, Saul Rennison <saul.renni...@gmail.com> > wrote: > > Bots taken far, far more CPU cycles than real players. Even the best > > servers can only handle about 16 bots before they begin to lag... many > > GSPs ban the use of too many bots they're so resource intensive. > > > > On Friday, June 18, 2010, Josh Bost <dislexs...@dislexsick.com> wrote: > >> Hi All, > >> > >> I'm preparing a server to be used at a local LAN party next month, the > >> server is pretty much enterprise specs (Dell PowerEdge 2950 - Intel > >> Xeon 5365, 16GB RAM, 6x146GB 15K RPM SAS drives in RAID-5) and I plan > >> to have several game servers running at once. I want to figure out > >> how many source servers I could run at once (Until now all my TF2 > >> servers have been through GSPs), and so I made a little benchmark - a > >> 32 man server, but since I can't populate that on my NAT'd (stupid > >> landlord, fml) ADSL2 connection I decided to stick 32 bots in it, > >> since I assume the AI would use just as many CPU cycles as a real > >> player. > >> > >> Anyway, onto the questions: > >> Has anybody tried this before? Does the benchmark given by running > >> bots translate easily into real world performance (Even if it is > >> something like take the CPU usage and multiply it by 1.5 to get an > >> approximation of how much a real player would use, etc)? > >> > >> I am only interested in CPU here, as that is the weak point of the > >> system (Allocating 500MB RAM to each server would give me 30 servers + > >> 1GB RAM to the Host OS, and there is no way I would need that much - > >> Similarly the RAID-5 Setup gives my 180-250MB/sec read speeds on the > >> HDD - Dual Gigabit NICs will do just fine for a LAN party) > >> > >> Alternatively, is anybody using a similar setup and could give me some > >> benchmarks for usage running TF2 (or L4D2, though it will only be > >> running one or two L4D2) servers? > >> > >> Cheers, > >> > >> Joshua 'Dislexsick' Bost > >> > >> _______________________________________________ > >> To unsubscribe, edit your list preferences, or view the list archives, > please visit: > >> http://list.valvesoftware.com/mailman/listinfo/hlds > >> > > > > -- > > > > Thanks, > > - Saul. > > > > _______________________________________________ > > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > > http://list.valvesoftware.com/mailman/listinfo/hlds > > > > _______________________________________________ > To unsubscribe, edit your list preferences, or view the list archives, > please visit: > http://list.valvesoftware.com/mailman/listinfo/hlds > _______________________________________________ To unsubscribe, edit your list preferences, or view the list archives, please visit: http://list.valvesoftware.com/mailman/listinfo/hlds
