Web Hosting Talk







View Full Version : Why are my proccess so HIGH?????????


NiceRsx2002
09-08-2002, 03:12 PM
Why are my servers proccess so high? I have another server more active than this one with 400 less proccesses that uses more bandwidth and has more sites on it... Here's the output when running top:

3:16pm up 3 days, 19:25, 1 user, load average: 1.46, 1.51, 1.41
561 processes: 560 sleeping, 1 running, 0 zombie, 0 stopped
CPU states: 3.6% user, 23.4% system, 0.3% nice, 72.5% idle
Mem: 524288K av, 301856K used, 222432K free, 54036K shrd, 0K buff
Swap: 1048576K av, 25076K used, 1023500K free 124456K cached

PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND
14333 root 11 0 1512 1512 852 R 23.7 0.2 0:14 top
14418 apache 0 0 20288 19M 19216 S 0.9 3.8 0:00 httpd
14402 apache 0 0 20292 19M 19212 S 0.7 3.8 0:00 httpd
14340 apache 0 0 20328 19M 19160 S 0.5 3.8 0:00 httpd
3433 root 1 0 332 316 240 S 0.1 0.0 92:08 syslogd
5332 root 1 0 18688 18M 18124 S 0.1 3.5 20:50 proftpd
12735 mysql 10 10 67832 47M 2244 S N 0.1 9.1 0:00 mysqld
13971 mysql 10 10 67832 47M 2244 S N 0.1 9.1 0:00 mysqld
14337 apache 0 0 20316 19M 19156 S 0.1 3.8 0:00 httpd
14398 apache 0 0 20304 19M 19168 S 0.1 3.8 0:00 httpd
14408 apache 0 0 19868 19M 19092 S 0.1 3.7 0:00 httpd
1 root 0 0 104 60 36 S 0.0 0.0 0:02 init
4376 root 0 0 68 0 0 SW 0.0 0.0 0:00 klogd
4442 daemon 0 0 128 52 44 S 0.0 0.0 0:00 atd
4473 root 0 0 1656 1656 716 S 0.0 0.3 11:15 snmpd
4492 named 0 0 2168 1936 1432 S 0.0 0.3 0:00 named
4495 named 0 0 2168 1936 1432 S 0.0 0.3 0:01 named
4496 named 0 0 2168 1936 1432 S 0.0 0.3 0:00 named
4497 named 0 0 2168 1936 1432 S 0.0 0.3 0:00 named
4498 named 0 0 2168 1936 1432 S 0.0 0.3 0:00 named
4512 root 0 0 456 340 272 S 0.0 0.0 0:18 sshd
4529 root 0 0 204 0 0 SW 0.0 0.0 0:00 sshd
4558 root 0 0 492 356 268 S 0.0 0.0 1:06 xinetd
4576 root 0 0 188 0 0 SW 0.0 0.0 0:00 safe_mysqld
4597 mysql 10 10 67832 47M 2244 S N 0.0 9.1 0:24 mysqld
4599 mysql 10 10 67832 47M 2244 S N 0.0 9.1 0:03 mysqld

Anyone know how I can find out why there are so many proccesses???

Shyne
09-08-2002, 03:28 PM
Check how many processes of mysql is running('myadmin processlist' is the commnad). Check the apache logs.

NiceRsx2002
09-08-2002, 04:00 PM
It says "command not found" I'm running an Ensim server with 7.2 if that make a difference...

mlovick
09-08-2002, 04:03 PM
Its a typo

mysqladmin processlist

;)

DD-SNC
09-08-2002, 04:18 PM
That's linux for ya.. Get BSD. I have shell servers running that many procs plus some and still under 0.

mlovick
09-08-2002, 04:25 PM
You can also use:

ps aux | grep mysqld
ps aux | grep httpd

This will filter out the processes for the individual services running.

2host.com
09-08-2002, 07:39 PM
pstree, netstat, lsof, top, ps, etc. will all help find any information you probably will need for any task.

Shyne
09-08-2002, 10:38 PM
Originally posted by DD-SNC
That's linux for ya.. Get BSD. I have shell servers running that many procs plus some and still under 0.

Linux has nothing to do with it. How does a shell server compare to a production server? Do you even know how many clients he has on the box? You can't just switch the OS of the server in a snap (software and hardware compatibilities). I see that switching to BSD is your solution for everything.

If my keyboard breaks I'll make sure to switch to BSD.

2host.com
09-08-2002, 10:42 PM
BSD does have better process control and handles things a little better. However, I agree it's not Linux at fault here just because of a high load, and comparing them, they aren't that much different, even if one might be a little better in my opinion, the other isn't so much worse that it really matters.

mind21_98
09-08-2002, 11:31 PM
23.4% system vs 3.6% user CPU usually indicates a lot of context switching or hard drive access (for IDE). You may need to increase the Min/MaxSpareServers values in your httpd.conf to reduce the number of processes Apache has to fork per minute.