On 05/24/2012 11:44 AM, Robert Haas wrote:
> On Thu, May 24, 2012 at 11:25 AM, Gurjeet Singh<singh.gurjeet@gmail.com> wrote:
>> It'd be great if one of standard utilities like pg_restore supported this,
>> by spawning every concurrent index build in separate backends. Just a
>> thought.
> If parallel restore doesn't already take this into account when doing
> job scheduling, that would be a worthwhile improvement to consider.
>
>
Parallel restore allows it in some cases, but not in others. You can't
create an index-backed constraint concurrently with any other index
because ALTER TABLE takes an ACCESS EXCLUSIVE lock.
We could get around that changing pg_dump to create the index first and
then add the constraint using the existing index.
cheers
andrew
