> From: Glyn Astill <glynastill at yahoo.co.uk>

> To: Slony List <slony1-general at lists.slony.info>
> Sent: Friday, 20 November 2015, 16:00
> Subject: [Slony1-general] Slony replication lag with moderate quantity of    deletes
> 
> The cause seems to be chunks of deletes on 2 specific tables (within a 
> transaction), although I can't see a reason why these specific tables cause 
> issues other than the tables having pretty poor primary keys and 
> indexes (Table definitions at the bottom).  > 

So after digging into this properly, the fault really has nothing to do with Slony at all.

After copying a dump of the data from the last slow copy into an isolated test table with the logApply trigger on it, and autoexplain enabled it appears the issue is purely the per row inserts into one of the tables getting a sub-optimal execution plan.