Instead of having the same actual IP for each server, why not use virtual IPs with something like heartbeat? This way if replica-A@siteA goes down, replica-A@siteB will take over. This entirely depends on your network topology, but may be a step in the right direction.
Having servers with the same IPs doens't seem to be the most elegant solution; same with rsync'ing your database instead of using replication. Jim On Thu, Apr 26, 2012 at 7:03 AM, Paul Robert Marino <[email protected]>wrote: > Well that's not a great way to set that up but its workable. You will need > to sync without ssl and do a destination nat betwean them > On Apr 26, 2012 4:05 AM, "Maurizio Marini" <[email protected]> wrote: > >> I have a disaster recovery scenario: >> on a remote location I have the same servers with the same hostnames and >> the >> same ip's, exactly all the same. >> Nightly I use rsync to keep all the servers in sync. >> One of this server is a CentOS5 with centos-ds and samba as pdc. >> I cannot use replica between current and dr, as the 2 server have the >> same ip >> and hostname. >> I am using ldap2db to import the nightly ldif backup. >> /usr/lib/dirsrv/slapd-centos-ds/ldif2db -n userRoot -i >> /tmp/backup-yyddmm.ldif >> It seems work, it's dirty but does work. >> Do u see any side-effects? Have u some suggestion? >> >> -m >> -- >> 389 users mailing list >> [email protected] >> https://admin.fedoraproject.org/mailman/listinfo/389-users > > > -- > 389 users mailing list > [email protected] > https://admin.fedoraproject.org/mailman/listinfo/389-users >
-- 389 users mailing list [email protected] https://admin.fedoraproject.org/mailman/listinfo/389-users
