[Date Prev][Date Next]
Re: Backing up LDAP database
- To: Lars Uffmann <email@example.com>
- Subject: Re: Backing up LDAP database
- From: Julio Sánchez Fernández <firstname.lastname@example.org>
- Date: Mon, 04 Oct 1999 11:49:24 +0200
- Cc: openldap-software@OpenLDAP.org
- References: <19990930094101.A14909@larry.cc.swin.edu.au> <126.96.36.199.19990929174323.00951ac0@localhost> <19990930113756.A15275@larry.cc.swin.edu.au> <37F2BF05.182FEE83@OpenLDAP.org> <37F318DB.6E6E46A2@mediaways.net>
Lars Uffmann wrote:
> This is exactly what I tried. But what about the ordering problem with
> ldbmcat? ie if the database has more then 256 Entries, ldbmcat swaps
> it output, the entries are in wrong order! The result is not usable as
> input to ldif2ldbm.
Have you tried it? With a recent version? There were old bugs in
database creation tools that created problems (some indexes would be
hosed and NEXTID would not be set up properly). I killed those bugs
months ago. All current branches are supposed to be fixed (2-0 alpha
uses a different approach now). I don't know for how long that part
was broken, possibly was broken in the Umich code base already. So
you might be used to *know* it did not work.
> I still do backups using a subtree search with
> '(objectclass=*)' filter.
Problem is you will start losing operational attributes if you keep
that approach in 2.0 when it is released.