Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I have a VPS running CentOS 5.7 (Final), 32-bit.

Right after boot, when I run # cat /proc/user_beancounters I see that my dcachesize usage is at about 4100000, while my VPS is set up with a limit of 4600000. So I am already using up about 90% of this cache somehow. Lately I've been having memory issues, which I've figured out I've been able to hotfix by service mysqld restart, since the usage of that builds up over time. But each restart takes it back down to about 410k, and then the cycle repeats.

I don't have much else running at the time -- just basic things like sshd or lighttpd. I suppose I could try stopping other services to see if the dcachesize value drops and write down each one. But it's surprising that it would just boot up with so much used. Is a value of 4 million for an idle state unusual?

Here's an output from ps -ef

UID        PID  PPID  C STIME TTY          TIME CMD
root         1     0  0 Sep26 ?        00:00:00 init [3]
root     11754 19881  0 11:50 ?        00:00:00 sshd: root@ttyp0
root     11784 11754  0 11:51 ttyp0    00:00:00 -bash
root     12246 11784  0 11:51 ttyp0    00:00:00 ps -ef
root     18358     1  0 Sep26 ?        00:00:00 /sbin/udevd -d
root     19499     1  0 Sep26 ?        00:00:03 brcm_iscsiuio
root     19776     1  0 Sep26 ?        00:00:01 syslogd -m 0
root     19785     1  0 Sep26 ?        00:00:00 klogd -x
dbus     19819     1  0 Sep26 ?        00:00:00 dbus-daemon --system
root     19881     1  0 Sep26 ?        00:00:00 /usr/sbin/sshd
root     19896     1  0 Sep26 ?        00:00:00 xinetd -stayalive -pidfile /var/run/xinetd.pid
tinydns  19963     1  0 Sep26 ?        00:00:00 /usr/local/bin/tinydns
root     19964     1  0 Sep26 ?        00:00:00 tcpserver -vDRHl0 -x tcp.cdb -- 0.0.0.0 53 /usr/local/bin/axfrdns
root     19976     1  0 Sep26 ?        00:00:00 sh ../sbin/kloxo-wrapper.sh
qmails   20232     1  0 Sep26 ?        00:00:00 qmail-send
qmaill   20241 20232  0 Sep26 ?        00:00:00 splogger qmail
root     20242 20232  0 Sep26 ?        00:00:00 qmail-lspawn |preline procmail
qmailr   20243 20232  0 Sep26 ?        00:00:00 qmail-rspawn
qmailq   20245 20232  0 Sep26 ?        00:00:00 qmail-clean
root     20338     1  0 Sep26 ?        00:00:00 /usr/libexec/courier-authlib/authdaemond
root     20340     1  0 Sep26 ?        00:00:00 /usr/bin/tcpserver -v -R -H -l myvpsname 0 110 /usr/sbin/pop3login /usr/bin/pop3d Maildir
root     20342     1  0 Sep26 ?        00:00:00 /usr/bin/tcpserver -v -R -H -l myvpsname 0 995 /usr/bin/couriertls -server -tcpd /usr/sbi
root     20344     1  0 Sep26 ?        00:00:00 /usr/bin/tcpserver -v -R -H -l myvpsname 0 143 /usr/sbin/imaplogin /usr/bin/imapd Maildir
root     20346     1  0 Sep26 ?        00:00:00 /usr/bin/tcpserver -v -R -H -l myvpsname 0 993 /usr/bin/couriertls -server -tcpd /usr/sbi
root     20380     1  0 Sep26 ?        00:00:00 crond
root     20417 20338  0 Sep26 ?        00:00:00 /usr/libexec/courier-authlib/authdaemond
avahi    20440     1  0 Sep26 ?        00:00:00 avahi-daemon: running [myvpsname.local]
avahi    20441 20440  0 Sep26 ?        00:00:00 avahi-daemon: chroot helper
root     26232     1  0 05:12 ?        00:00:00 /bin/sh /usr/bin/mysqld_safe --datadir=/var/lib/mysql --socket=/var/lib/mysql/mysql.soc
mysql    26274 26232  0 05:12 ?        00:00:40 /usr/libexec/mysqld --basedir=/usr --datadir=/var/lib/mysql --user=mysql --pid-file=/va
apache   32232     1  0 04:53 ?        00:01:46 /usr/sbin/lighttpd -f /etc/lighttpd/lighttpd.conf
lxlabs   32312     1  0 04:53 ?        00:00:00 /usr/local/lxlabs/ext/lxlighttpd/sbin/kloxo.httpd -f /usr/local/lxlabs/kloxo/file/light
root     32511 19976  0 04:54 ?        00:00:00 ../sbin/kloxo.exe master

In short, what are good techniques for reducing your usage of dcachesize? Because I am on a VPS, I don't have access to OpenVZ to raise this limit.

share|improve this question
    
did you ever figure this out? –  doitlikejustin Aug 14 '13 at 23:21
1  
It turned out there was nothing I could do on my own. I would have had to contact my host, so they could adjust the limit for my VPS. However, I think a few weeks after this post I ended up buying a new VPS with CentOS 6 and just copied all of my files and settings over, and the new VPS had much better memory and configuration. –  Crisu Aug 16 '13 at 13:56
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.