Adrian Bridgett wrote:
> On Tue, Apr 24, 2007 at 13:08:46 +0100 (+0100), Brian Chivers wrote:
>> What I normally do is simply copy the slapd.conf & ldap.conf files from
>> one of the other slaves and then simply slapadd -l master.ldif which is a
>> dump of our main ldap server. This is where the problem starts the
>> database backend that all our other servers use is set to ldbm not the
>> default bdb. I've found this via google but it doesn't seem to help much
>
> hdb is recommended these days (over bdb at any rate). I use the
> syncrepl overlay and just point it at the master, then get updates
> pushed out from the master (I have the push version rather than the
> pull version - as soon as you change something on the master, it's on
> the slaves). It grabs the whole DB over on the first install.
>
> One problem I've had is the naff error messages - basically you don't
> get anything nice like "oops, someone's added a schema on the master
> but not the slave". Need to check contextCSN attribute to see if all
> the updates are applied.
>
>> Does anyone know how to "add" ldbm or can I simply use the default bdb.
>> If this is the case does anyone know if the replication will work via
>> slurpd and this will take care of the conversions ??
>
> Slurpd has been deprecated in favour of the syncrepl overlay.
>
> Adrian
>
And there was me thinking I'd got it all sorted with slapd & slurpd working together :-(
Do you know if you have a ldbm db one the master but hdb on the slave the replication will work ??
Brian
------------------------------------------------------------------------------------------------
The views expressed here are my own and not necessarily
the views of Portsmouth College
