[SkeetHunter]

Well the only thing about that is Adminmod and metamod 1.17.3 are working fine on my servers...but I like the idea that someone has fixed my problem for me so I'll try that next. For now, I am still running them without the pingboost to see if that has an effect.

[SkeetHunter]

Well Hulla's with or without pingboost is stable on FC2.

CPU   In    Out   Uptime  Users   FPS    Players
11.00 28.05 36.75     944    31  327.23      11

[PM]

So.. does this mean Hullu's metamod version fixes your problems? That would be great

[ToT | V!PER]

Originally Posted by PM So.. does this mean Hullu's metamod version fixes your problems? That would be great

No crashes yet since my first post about hullu´s metamod.

[SkeetHunter]

Yes, hlds service uptime proves no crashes.

CPU   In    Out   Uptime  Users   FPS    Players
19.20 44.31 58.95    1206   344  198.33      17

CPU   In    Out   Uptime  Users   FPS    Players
23.00 49.22 67.94    1207   158  330.69      19

NOTE: I have installed the hullu metamod. However, during testing on my systems I also determined that the pingboost option is not necessary. In fact, removing has improved overall system performance.

Another noteworthy event is that with the hullu metamod is server is MUCH better on memory usage (less leak).

top - 21:23:03 up 20:13,  1 user,  load average: 0.80, 0.75, 0.67
Tasks:  53 total,   2 running,  51 sleeping,   0 stopped,   0 zombie
Cpu(s): 20.3% us,  8.1% sy,  0.0% ni, 67.9% id,  2.7% wa,  0.3% hi,  0.7% si
Mem:   2075100k total,   693236k used,  1381864k free,   140040k buffers
Swap:  2040212k total,        0k used,  2040212k free,   232020k cached

 2883 *****      15   0  122m 115m 5740 S 12.0  5.7 111:13.52 hlds_i686
 2887 *****      15   0  112m 105m 5760 S 18.3  5.2  35:12.62 hlds_i686

Both servers are on the same box. I am running Dual 3.2 Xeons w 2MB cache and HT turned off. I have installed run-2.0-3.i386.rpm fron FC1 which allows setting of CPU affinity for my processes. Note the run command followed by -b (bias) and the CPU number.

screen -A -dmS ava run -b 0 ./hlds_run -tos -pingboost 1 +heapsize 512000 -game dod +ip 65.69.37.140 +maxplayers 26 +map dod_avalanche -autoupdate

screen -A -dmS switch run -b 1 ./hlds_run -tos -pingboost 1 +heapsize 512000 -game dod +ip 65.69.37.141 +maxplayers 32 +map dod_switch -autoupdate

[ToT | V!PER]

You are using +heapsize?
Why do you specify it?

[SkeetHunter]

That's a good question. I saw that setting when looking for the best configuration of hlds once awhile ago...as well as the pingboost option. Simply, I've been doing this for awhile now. But I am always looking to make my server perform better. The heapsize option as I see it simply allocates the given amount of memory to the hlds process.

[SkeetHunter]

Just documenting uptime for servers...

top - 20:04:50 up 2 days, 18:55,  1 user,  load average: 0.45, 0.38, 0.31
Tasks:  53 total,   2 running,  51 sleeping,   0 stopped,   0 zombie
Cpu(s): 14.5% us,  0.7% sy,  0.0% ni, 84.0% id,  0.0% wa,  0.5% hi,  0.3% si
Mem:   2075100k total,   768484k used,  1306616k free,   144244k buffers
Swap:  2040212k total,        0k used,  2040212k free,   239256k cached

CPU   In    Out   Uptime  Users   FPS    Players
30.00 58.82 78.55    4011   928  247.95      21

CPU   In    Out   Uptime  Users   FPS    Players
 0.00  0.00  0.00    4012   374  249.56       0

 PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  CMD
 2883 *****      15   0  172m 162m 5740 S 28.3  8.0 307:19.00 hlds_i686
 2887 *****      15   0  128m 121m 5760 S  0.0  6.0  93:12.27 hlds_i686

[illegal-op]

I've been having the same exact problems. Something interesting is that every time it crashed, my cpu usage would shoot up to almost 100% and then it would crash.

P4 2.8
1 Gig dual channel ram
SuSE 9.0

I will give that hulu metamod a try and see what happens.

[NiLuJe]

Does it still work with metamodp15 ?