On Thursday 04 January 2007 19:06, Quanah Gibson-Mount wrote:
--On Thursday, January 04, 2007 10:01 AM +0100 poberli@hsr.ch wrote:
Any ideas what this could be?
This is a default debian installation which ldapd version: slapd -V @(#) $OpenLDAP: slapd 2.2.23 (May 30 2005 08:52:42) $
Yes, you need to upgrade to a later release. Preferably, 2.3, which has auto recovery. It sounds like your database is not being shut down cleanly, and you are not running db_recover when that happens. There were many issues with the 2.2.23 release, IIRC. 2.2.30 was the last release in the 2.2 series, and 2.3 is now up to 2.3.32.
2.3.31 is the latest (by tarballs available, and tags) that I can see ... and here I was just about to build new packages with patches for the most recent 3 post-2.3.31 ITSs ...
2.3 adds many nice new features in addition to significant performance increases.
Of course, there could be other reasons for the problems the user is having.
E.g., this could be ldbm data integrity problems (just switching to bdb should fix that). Upgrading to 2.3.x with ldbm may not have any benefit ...
Patrick, it may be best giving some details of your database backend, and any configuration you have for it (e.g. checkpoint and use or not of a DB_CONFIG file in the case of bdb). If you're not sure, the relevant bits of your slapd.conf would be useful.
Regards, Buchan