Re: [HACKERS] Proposal: Local indexes for partitioned table

Peter Eisentraut wrote:
> On 12/29/17 15:38, Alvaro Herrera wrote:
> > I just realized there's a further problem in the area: when a partition
> > is detached from its parent, its indexes are not made independent of the
> > indexes on parent.  So they can't be dropped on their own (booh!); and
> > dropping the index on the former parent partitioned table drops the
> > index on the former partition also (hiss!).  I think the fix for this is
> > to delete all dependencies, and re-register normal ones.  Should be
> > straightforward, but I'm not doing it this year.
> 
> We're waiting for an updated patch in this thread, aren't we?

Dunno.  The problems reported so far are rather very minor details that
don't prevent reviewing the rest of it.  I will of course send an
updated version at some point, but there's only one me so I can either
spend the whole commitfest focused on fixing tiny issues my own patches,
or help with other stuff.

-- 
Álvaro Herrera                https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services