Filling my /var on Freebsd 5.3

[NEO^AMiGA]

Heya!

First of all.-.. thnx for a great program.

Ran into som problem tonight. My /var/ got filled. =/ What does cacti keep there and does it grow over time or does it stick to a resonable size? Maybe my problem is not only cacti-related but this happend after running if for a couple of days so it's where i'll start to look for errors.

/var/db >du -hd1
8.0K ./entropy
2.0K ./ipf
4.7M ./pkg
18K ./ports
14M ./sup
26K ./portaudit
33M ./mysql
54M .

/var/db/mysql >ll
total 32872
drwx------ 2 mysql mysql 4096 May 2 23:55 cacti
-rw-rw---- 1 mysql mysql 25088 Apr 29 03:19 ib_arch_log_0000000000
-rw-rw---- 1 mysql mysql 5242880 Apr 29 03:25 ib_logfile0
-rw-rw---- 1 mysql mysql 5242880 Apr 29 03:19 ib_logfile1
-rw-rw---- 1 mysql mysql 10485760 Apr 29 03:25 ibdata1
drwx------ 2 mysql mysql 512 Apr 29 03:19 mysql
drwx------ 2 mysql mysql 512 Apr 29 03:19 test
-rw-rw---- 1 mysql mysql 12567368 May 3 00:10 fakehost-bin.001
-rw-rw---- 1 mysql mysql 23 Apr 29 03:25 fakehost-bin.index
-rw-rw---- 1 mysql mysql 1248 Apr 29 03:25 fakehost.xxx.org.err
-rw-rw---- 1 mysql mysql 6 Apr 29 03:25 fakehost.xxx.org.pid

/var/db/sup/ports-all >ll
total 14416
-rw-r--r-- 1 root wheel 14741549 Apr 29 00:27 checkouts.cvs:.

/var/spool >du -hd1
2.0K ./lock
2.0K ./lpd
2.0K ./mqueue
2.0K ./opielocks
4.0K ./output
32M ./clientmqueue
32M .

Where /clientmqueue is filled with files like this:
-rw-rw---- 1 smmsp smmsp 972 May 2 03:06 qfj42162H8041753
-rw-rw---- 1 smmsp smmsp 972 May 2 03:06 qfj42162H9041753
-rw-rw---- 1 smmsp smmsp 926 May 2 03:06 qfj42162HA041753
-rw-rw---- 1 smmsp smmsp 713 May 2 03:06 qfj42162Ip041707
-rw-rw---- 1 smmsp smmsp 710 May 2 03:06 qfj42162ne041751
-rw-rw---- 1 smmsp smmsp 0 May 2 23:01 qfj42L12HE073334
-rw-rw---- 1 smmsp smmsp 897 May 3 00:00 qfj42M030v075593

Does this make sense to someone here? I'm kinda lost at the moment and if you need more info from me, just ask. Also... When I ran out of diskpace I got a "hole" in all my graphs, is that fixable?

[rony]

Don't know about you client queue. But turn on binary loggin in your mysql server conf. /etc/my.cnf

[NEO^AMiGA]

Don't know about you client queue. But turn on binary loggin in your mysql server conf. /etc/my.cnf

This is already in that conf:

# Replication Master Server (default)
# binary logging is required for replication
log-bin

Okay? OR should it look different... Soon my var is full again! =(((( And I have no idea what of those stuffs I can delete.

[NEO^AMiGA]

The spool/clientmqueue was a minconfigured mail-thingy. And after a fix of that and reboot i have some space left. But still wonder about that /db one...

[Linegod]

If you are not using MySQL replication, you probably want to turn binary logging _off_, not on.

Or you can manually clean out the bin files by logging into mysql and doing:

RESET MASTER;