Hi,

The method suggested by Quanah worked perfectly. I also find very useful the information given by Dieter since I'll take of creating a DB_CONFIG file in new deployments. My last question has to be with db_checkpoint. Should I call it periodically or will Berkeley delete logs by its own?

Thanks a lot

2010/2/12 Quanah Gibson-Mount quanah@zimbra.com

--On Thursday, February 11, 2010 7:21 PM +0100 Dieter Kluenter < dieter@dkluenter.de> wrote:

Quanah Gibson-Mount quanah@zimbra.com writes:

...

----- Echedey Lorenzo echedey@gmail.com wrote:

...

...

I think the best option, as you suggest, is to recreate everything. Is it enough to remove all /var/lib/ldap contents, restart the ldap service, and populate all again? My intention is to have 8M entries as max.

Thanks for your help

There is zero need to recreate everything. Dieter is wrong. Simply stop slapd, create the DB_CONFIG file, run db_recover to regenerate the bdb database profile, then start slapd. This has been the standard way to do this since OpenLDAP 2.1. Dieter should know this.

I have experienced some problems in the past, that's why I prefer a clean recreation.

Then you should have collected data and filed an ITS. I've never had a problem in nearly 10 years of configuring and modifying DB_CONFIG files back to 2.1.4 when it's done correctly.

--Quanah

--

Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc

---

Zimbra :: the leader in open source messaging and collaboration