How long can a LDAP frontend be down (master and slave) ?

[Alex Waldus <alex.waldus@cmg.com>]

I have a question about the replication between MASTER and SLAVE ldap
servers.

How long can a LDAP Front-end (SLAVE) be down before the slurpd running on
the backends can not update(sync) it anymore.

Is it the total free diskspace?
Does it depend on the number of changes made in the down time?
Can I configure this some where?

I'm asking this because if I want to update (eg patches on the OS) an
Front-end (Slave) i want to know how much time i have to do it.

I allready asked some people and they say it is possible but nobody until
now is able to tell me exactly.
Does anybody here know?
Please provide me a detailed answer.

Regards,

Alex