Kolab High Availability and Scalability

[suffocator at gmx.de]

Hi List,

I'm looking for some information on Kolab High Availability and Scalability.
As mentioned in the thread "Kolab numbers" there are a few projects
concerning this point of interest. Do you have any information to share?
Setting up e.g. two cluster nodes based on Redhat AS with somehow mirrored
shared storage with heartbeat and ha service sounds good, but what is your
experience with Postfix and Cyrus IMAPd? What kind of storage is 
recommended? Are you facing any problems regarding message-ids handled by
more than one mta? Do you run only one LDAP database or do you replicate
with slurpd switching to master-role during a possible failover?
What about CPU-Load on the server regarding SSL/TLS encryption in times with
much traffic? As I've seen there has been a HA Kolab installation on CEBIT,
maybe you can share some details on how things fit together? 
How did you design your very own concept? 

Still many more questions and I'm curious about the answers...

TIA,

Magnus

-- 
support our band!!!
visit http://www.humanbastard.de

support the search for extraterrestrial intelligence!
visit http://setiathome.berkeley.edu

"Sie haben neue Mails!" - Die GMX Toolbar informiert Sie beim Surfen!
Jetzt aktivieren unter http://www.gmx.net/info