hi
i'm using slony 1.0.5 (we have big problems with slony 1.1 replicating 
unicode databases, and i dont know if i can trust 1.0.6 at this moment).
slon in this version is having 3 parameters.
i read about them, but frankly i dont understand what they mean.
for example - -s - i lowered the number in hope if will check and replicate 
more often, but dblag skyrocketed.

anyway
our database is under quite heavy load of inserts and updates
and default slon parameters dont work. after 30 hours i have 24 hours of 
replication lag.

could anyone of you help me understanding the parameters to tune the 
replication system to work with our load?

and additionally - what can i do to catchup with replication - as fast as it 
is possible.
i would even consider recreating the replication, but:
1. the only way that i know - (drop node <slave_db>, create node, subscribe 
set) leaves old entries in sl_event and sl_log (according to documentation), 
which makes me think that after such recreation db lag will start to grow 
again.
2. i would like to remove the replication from both master and slave, and 
reinit cluster, but i have no clue on how to "drop node ( id = 1)". any 
help?

depesz
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://gborg.postgresql.org/pipermail/slony1-general/attachments/20050824/e2f91f39/attachment.html