Hi,

If a node is a long way behind, and the slon servicing it is 
aggressively trying to catch up, sometimes the application of group 
SYNCS can take a very long time (in my testing, sometimes more than an 
hour).  During a this the cleanup thread apparently does the same thing, 
over and over.   Here's an example:

2007-12-04 21:02:10.900 CUT [1393000] cluster1 slony 10.1.2.133 LOG:  
duration: 6300.830 ms  statement: delete from "_cluster1".sl_log_1 where 
log_origin = '8143' and log_xid < '56583147'; delete from 
"_cluster1".sl_log_2 where log_origin = '8143' and log_xid < '56583147'; 
delete from "_cluster1".sl_seqlog where seql_origin = '8143' and 
seql_ev_seqno < '301904'; select "_cluster1".logswitch_finish();

2007-12-04 21:14:22.366 CUT [1393000] cluster slony 10.1.2.133 LOG:  
duration: 2324.701 ms  statement: delete from "_cluster".sl_log_1 where 
log_origin = '8143' and log_xid < '56583147'; delete from 
"_cluster".sl_log_2 where log_origin = '8143' and log_xid < '56583147'; 
delete from "_cluster".sl_seqlog
where seql_origin = '8143' and seql_ev_seqno < '301904'; select 
"_cluster".logswitch_finish();

I don't think this is particularly problematic, but was this the 
intended design?

JP

-- 
JP Fletcher
Database Administrator
Afilias Canada
voice: 416.646.3304 ext. 4123
fax: 416.646.3305
mobile: 416.561.4763
jpfletch at ca.afilias.info