We know that if there are old idle-in-transaction transactions lying
around, we won't get any useful results out of a VACUUM.

A thought, therefore, is that the cleanup thread might take a peek to
see if there is such a transaction that has been running since the
_last_ cleanup cycle.  If that is the case, there's no point to
VACUUMing any of the Slony-I tables as NONE of the tuples killed since
the last cycle will get cleaned out.  It would be a good idea to
report this in the log, perhaps even reporting an "offending" PID.

Thoughts?  If it works, this ought to get rid of useless VACUUM
cycles...  We might ANALYZE instead, I suppose...
-- 
"cbbrowne","@","ca.afilias.info"
<http://dev6.int.libertyrms.com/>
Christopher Browne
(416) 673-4124 (land)