Re: OpenLDAP + RHEL4 - time to time database crash (BDB backend)

[Michal Dobroczynski <michal.dobroczynski@gmail.com>]

On 22/09/05, Dusty Doris <openldap@mail.doris.cc> wrote:
> > I found out a few things that might cause the problem but I do not
> > have enough experience to judge which one is the most likely:
> > - bdb database tuning, meaning I should configure checkpointing,
> > cachesize, idlcachesize etc. - I'll refer to faq-o-matic for some
> > clues about the possible values;
> > - running db_checkpoint (uid same as slapd's uid) (but how often?)
> >
>
> I am using 2.3.7 but here is my configs for bdb tuning.
>
> checkpoint, cachesize, idlcachesize goes slapd.conf.  Don't need to run
> db_checkpoint manually.
>
> cachesize       50000
> idlcachesize    150000
> checkpoint      1024 5
> (that says to flush to disk every 1M or 5 minutes)
>
>

I added these three lines, but with slightly different values
(according to the size of my database).

> Berkely DB environment goes in DB_CONFIG in your bdb directory
>    ie: /var/db/openldap-data
>
> set_cachesize 0 536870912 1
> set_lg_regionmax 1048576
> set_lg_bsize 2097152
> set_lg_max 10485760

Almost same here. I do not use the DB_LOG_AUTOREMOVE. I also upgraded
over the night to 2.2.26.

> set_flags DB_LOG_AUTOREMOVE
>
> (Note DB_LOG_AUTOREMOVE only works if using bdb 4.2 w/ the supplied patch
> by openldap or bdb 4.3)
>
>
>
>

Now I just have to wait and see if everything is going to be okay.
Anyway, in two weeks I plan a total change of the distribution and
then I'll pay much attention to BDB version and so on.

Btw. - are you using 2.3.7 in productive environment?

Regards,
Michal