2008/2/20, Geoffrey <lists@serioustechnology.com>:
> We've just started seeing these errors. Research I've done seems to
> indicate that it's related to temp tables. Question is, we didn't start
> seeing these errors until we started using slony to replicate our data.
>
> The errors only showed up shortly after the initial replication of the
> data was complete. That is, once the replication server 'caught up'
> with the production server.
>
I've ever seen the same error. At that time I didn't use slony but
replicated some postgresql servers by an application layer. It seemed
that not only temp table but regular relations are dropped and created
quickly when the error occurred.
There could be a chance to drop information about relations or
themselves between SELECT catching relation OID and begining scan the
relation actually.
I have no solution for this until now. But just for your information.
Hitoshi Harada