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

> To: Dave Cramer <davecramer at gmail.com>
> Cc: slony <slony1-general at lists.slony.info>
> Sent: Thursday, 16 October 2014, 16:26
> Subject: Re: [Slony1-general] Lag time increasing but there are no events
> 
> 
>>  From: Dave Cramer <davecramer at gmail.com>
>> To: Glyn Astill <glynastill at yahoo.co.uk> 
>> Cc: slony <slony1-general at lists.slony.info> 
>> Sent: Thursday, 16 October 2014, 15:41
>> Subject: Re: [Slony1-general] Lag time increasing but there are no events
>> 
>> 
>> 
>> Actually I think it is because a pg_dump of the db is going on.
>> 
>> 
>> Can you dump a slave ? 
>> 
>> 
>> Do you have to exclude the slony clusters (which would make sense)?
> 
> 
> Yeah from past experience the locks taken by pg_dump cause replication to lag. I 
> normally use the -N switch in pg_dump to exclude the schema.  For pg_restore I 
> usually create a TOC with the -l option and pipe it to grep to filter out the 
> references to the slony schema that get dumped for the triggers, and then -L 
> option to use it with my restore. 

Also I assumed you were using 2.0+ there, with 1.2 there's some slight fiddling done in pg_catalog to disable triggers, from memory pg_trigger.tgrelid is pointed to an index rather than the table or something like that.  I used to dump out with the slony schema, and run uninstallnode() and drop schema cascade against the restored database.