Hi Dirk,

On Mon, Mar 7, 2016 at 6:44 PM, Dirk Kastens <dirk.kastens@uni-osnabrueck.de> wrote:

Hi,

I am running the same package version openldap-2.4.40-6.el6_7.x86_64 as
you do. I've double checked the
/etc/openldap/slapd.d//olcDatabase={2}bdb.ldif, the olcDbCheckpoint:
1024 15 has already been defined in that file, and I did restarted
slapd, but the log files are still glowing. What else I could still be
missing?

How big is your db? I've set the logsize to 100 MB and only the last two files are kept. Install the db4-utils and run the command db_archive. This will show you the logfiles that are no longer used.
"db_archive -d" will delete the unused logfiles. But this should be done by slapd automatically.

The logsize is set to 10 MB, it creates a 10 MB log file every 2 - 3 minutes (it grows so fast, like the movie "Evolution" :-)), I ran the db_archive -d 10 hours ago when the disk was full, and I have run the db_archive -d now as the disk is full again despite the server is not too busy. Why doesn't the slapd do it automatically? Apart that issue, the ldap is running very well. I really want to figure out what is going on, but if I cannot find a solution, I have no choice but run a cron job to call db_archive -d every 8 hours. I guess that should not cause ldap any side effect, right?

Thank you very much.

Dirk