slurpd problem in cluster configuration

["Luigi Formisano" <luigi.formisano@keypartners.it>]

Hi all,
sorry if I send this mail again, but I don't know how to resolve this
problem.

I would like to install openldap (2.0.25) on a SUN cluster of two machines
(Solaris 8 with SUN cluster 3.0).

My ldap architecture is made by one master server, one replica hub and two
slaves.

The master and the replica hub are on the cluster. The slaves on some linux
servers.

In normal condition (when all the cluster's machines are up and running) the
master and the replica hub run on the same server. If one machine fail the
two daemons will switch on the good one.
So, in normal case, I have two slapd and two slurpd daemon on the same
server.
I configured the master slapd.conf and the replica slapd.conf to use
different directory for the database, different replicalogfile. The two
slapd run on two different ip addresses of the same machine, so they don't
conflict.

My problem is with the replication. The replication from the master to the
replica hub is ok, but the replica hub can't replicate to the two slaves.

I saw on the slurpd debug that also if I say to the two slurpd to use two
different log file, the two slurpd copy the log files to the same directory
(/usr/local/openldap/var/openldap-slurp/replica). I think that the two
slurpd get confused and don't replicate on the two slaves.
Infact they write on the same file, slurpd.replog.

Do you have any idea how to resolve this problem ?
Is it possile to run two instances of slapd and slurpd on the same machine ?
Is it possible to run openldap on a cluster configuration as the one that I
have ?

Thanks a lot for the help,
Luigi