[Mageia-sysadm] [sysadmin-reports] Hobbit [80535] krampouezh.mageia.org:disk warning (YELLOW)

[Buchan Milne]

On Thursday, 9 February 2012 12:16:19 Pascal Terjan wrote:
> On Thu, Feb 9, 2012 at 06:55,  <root at mageia.org> wrote:
> > yellow Thu Feb  9 07:55:02 CET 2012 - Filesystems NOT ok
> > &yellow / (90% used) has reached the WARNING level (90%)
> > 
> > Filesystem         1024-blocks      Used Available Capacity Mounted on
> > /dev/xvda1             2466700   2084572    256824      90% /
> > /dev/xvdb              1032088     68144    911516       7% /srv/mx2-dd0
> > 
> > See
> > http://xymon.mageia.org/xymon-cgi/bb-hostsvc.sh?HOST=krampouezh.mageia.o
> > rg&SERVICE=disk
> 
> [root at krampouezh ~]# df
> Sys. de fich.         Tail. Occ. Disp. %Occ. Monté sur
> /dev/xvda1            2.4G  2.0G  247M  90% /
> /dev/xvdb            1008M   67M  891M   7% /srv/mx2-dd0
> [root at krampouezh ~]# du -sh /var/lib/ldap/backup
> 915M	/var/lib/ldap/backup


For now, I have manually changed:

# How many days to keep archived transaction logs for. This should be just
# greater than the backup interval on these files. Default is 7
# KEEP_ARCHIVES_DAYS=7

to:
# How many days to keep archived transaction logs for. This should be just
# greater than the backup interval on these files. Default is 7
# KEEP_ARCHIVES_DAYS=7
KEEP_ARCHIVES_DAYS=1

in /etc/sysconfig/ldap, and ran /etc/cron.daily/ldap-hot-db-backup -v4

[root at krampouezh ~]# du -sh /var/lib/ldap/backup
389M    /var/lib/ldap/backup
[root at krampouezh ~]# df
Sys. de fich.         Tail. Occ. Disp. %Occ. Monté sur
/dev/xvda1            2.4G  1.4G  872M  62% /
/dev/xvdb            1008M   67M  891M   7% /srv/mx2-dd0

But, I am not sure how backups are implemented, and we could also consider 
enabling automatic archive log removal (in DB_CONFIG) for any slaves.


It may also be that the find call in ldap-common should be using -mtime 
instead of -ctime (as we still have >2 days of archive logs).

Regards,
Buchan