Hi,
I hope this isn't answered somewhere else yet - tried to do a quick search but to no avail.
I have a VMware ESX VM with 768 MB ram. Running ispCP 1.0.5 OMEGA build: 20100331 Codename: Priamos. The server is operating 16 extremely low-traffic domains but I need to scale the system as my client base grows.
Code:
leskinensolutions:~# cat /etc/debian_version
5.0.6
The problem is that php5-cgi processes are consuming close 200mb each, providing apparent problems running this server.
If I use php as apache2 module there are no ram-related problems aside the apparent non-usermode issues.
I've tried prefork and worker mpm modules to no avail.
I'm currently running apache2-mpm-worker.
Code:
leskinensolutions:~# ps aux | grep php5-cgi | wc -l
7
Code:
leskinensolutions:~# ps aux | grep apache2 | wc -l
6
Output of top:
Code:
Tasks: 93 total, 2 running, 91 sleeping, 0 stopped, 0 zombie
Cpu(s): 2.7%us, 0.0%sy, 0.0%ni, 97.3%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 769332k total, 613852k used, 155480k free, 10944k buffers
Swap: 522072k total, 217120k used, 304952k free, 259288k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
11706 vu2011 20 0 198m 66m 37m S 0.0 8.9 0:02.76 php5-cgi
11528 vu2002 20 0 194m 63m 37m S 0.0 8.4 0:06.46 php5-cgi
7138 amavis 20 0 191m 55m 2772 S 0.0 7.4 0:02.08 amavisd-new
7139 amavis 20 0 191m 54m 2912 S 0.0 7.3 0:02.36 amavisd-new
11512 vu2021 20 0 192m 53m 27m S 0.0 7.1 0:00.96 php5-cgi
7137 amavis 20 0 188m 49m 2744 S 0.0 6.5 0:00.72 amavisd-new
11648 vu2019 20 0 183m 36m 20m S 0.0 4.8 0:00.32 php5-cgi
10072 vu2000 20 0 179m 30m 19m S 0.0 4.0 0:00.66 php5-cgi
10003 root 20 0 62004 14m 2524 S 0.0 1.9 0:00.18 ispcp-apache-lo
10002 root 20 0 61872 14m 2524 S 0.0 1.9 0:00.18 ispcp-apache-lo
7111 clamav 20 0 193m 14m 5348 S 0.0 1.9 0:04.24 clamd
10246 mysql 20 0 47512 11m 5148 S 0.0 1.6 0:00.42 mysqld
10067 vu2020 20 0 170m 8752 5540 S 0.0 1.1 0:00.02 php5-cgi
10009 www-data 20 0 357m 8412 2012 S 0.0 1.1 0:00.02 apache2
8081 root 20 0 138m 7700 3300 S 0.0 1.0 0:00.64 apache2
11764 www-data 20 0 357m 6900 1844 S 0.0 0.9 0:00.00 apache2
18004 clamav 20 0 57108 5796 5640 S 0.0 0.8 0:01.92 freshclam
2691 postgrey 20 0 57100 5360 2620 S 0.0 0.7 0:01.08 postgrey
10005 www-data 20 0 138m 5032 620 S 0.0 0.7 0:00.02 apache2
10004 www-data 20 0 137m 5028 572 S 0.0 0.7 0:00.00 apache2
8744 polw 20 0 46628 5004 1540 S 0.0 0.7 0:02.36 policyd-weight
5140 root 20 0 103m 3404 1808 S 0.0 0.4 5:19.57 fail2ban-server
2687 polw 20 0 46492 2916 1292 S 0.0 0.4 0:00.90 policyd-weight
2688 polw 20 0 46492 2700 968 S 0.0 0.4 0:01.82 policyd-weight
17429 root 20 0 65932 2344 2172 S 0.0 0.3 0:00.08 sshd
30866 bind 20 0 111m 2140 1352 S 0.0 0.3 0:00.59 named
3315 postfix 20 0 38948 2064 1896 S 0.0 0.3 0:00.04 qmgr
2835 root 20 0 36844 1968 1840 S 0.0 0.3 0:00.14 master
8036 postfix 20 0 38900 1968 1820 S 0.0 0.3 0:00.00 pickup
17446 root 20 0 18848 1728 1392 S 0.0 0.2 0:00.22 bash
2507 root 20 0 119m 1604 884 S 0.0 0.2 0:03.78 rsyslogd
Code:
leskinensolutions:~# free -m
total used free shared buffers cached
Mem: 751 599 151 0 10 253
-/+ buffers/cache: 335 415
Swap: 509 212 297
Since there seems to be a lot of ram in cache, is there something to be alarmed of? To me the php5-cgi ~198mb ram usage seems bit odd.
Anything I could do/check to ensure the ram usage is OK?
Thanks!
BR