On 2014-07-22 10:17:15 -0400, Tom Lane wrote:
> Or even more to the point, investigate why it's there in the first
> place; perhaps there's an actual fixable bug somewhere in there.
I think MauMau's scenario of a failover to another database explains
their existance - there's no step that'd remove them after promoting a
standby.
So there indeed is a need to have a sensible mechanism for removing them
at some point. But it should be about removing, not ignoring them.
Greetings,
Andres Freund
-- Andres Freund http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services