We just rolled out slony 1.05 for our production cluster today, and it seems
to be working like a charm.  Now I need to determine a path for upgrades of
our application and its databases.  For now, I am advocating the method of
breaking the mirror, upgrading the master, and then recreating a mirror,
since at the moment this takes only about 20 minutes.  We would like to
explore the option of using the 'execute' statement to syncronize changes to
both machines.
 
If we alter tables to add/remove columns, will slony's triggers be properly
updated?
If we alter the database to add/remove tables, will we need to manually
update the mirror set, or does this happen automagically as well?
 
Thanks,
 
-- 
Adam Haberlach
Database Administrator
adam at polimetrix.com
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gborg.postgresql.org/pipermail/slony1-general/attachments/20050610/f8ee008c/attachment.html