On Mon, Aug 27, 2007 at 04:30:29PM -0500, Dan Falconer wrote:

> have no idea what causes it.  Going with the suggestions of David Rees, I 
> stopped slons & ran some vacuums: on each server, I vacuumed every time I 
> found within the "_pl_replication" schema.  At most, the vacuums took 30 
> seconds to run, and I saw nothing terrible in regards to the number of rows 
> removed, rows that couldn't be removed, etc.

Hrm.  Did it recover any dead tuples at all?  And what about the main
database?  Remember, slony has to find the rows to update and delete. 
Those are rows in the _replicated_ tables.  Are transactions running
long on your targets?

A

-- 
Andrew Sullivan  | ajs at crankycanuck.ca
When my information changes, I alter my conclusions.  What do you do sir?
		--attr. John Maynard Keynes