From: | Noah Misch <noah(at)leadboat(dot)com> |
---|---|
To: | Amit Kapila <amit(dot)kapila(at)huawei(dot)com> |
Cc: | pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: Parallel Sort |
Date: | 2013-05-15 18:12:22 |
Message-ID: | 20130515181222.GC234183@tornado.leadboat.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Wed, May 15, 2013 at 12:26:52PM +0530, Amit Kapila wrote:
> On Monday, May 13, 2013 7:59 PM Noah Misch wrote:
> > We can allocate a small amount of permanent shared memory for
> > coordination
> > among a group of processes, but sorting will benefit from a region as
> > large as
> > maintenance_work_mem. Expect on-demand memory sharing.
>
> Will the shared memory used for coordinating tuples between master and
> worker be fixed or varying depending on size of tuples to be sorted or
> number of workers associated.
> If it is varying, then it can sometimes encounter situation where required
> memory is not available and in that case it has to revert to serial sorting
> How will the parallel sorting tasks be divided and assigned to each worker?
I haven't selected answers for those details, yet.
--
Noah Misch
EnterpriseDB http://www.enterprisedb.com
From | Date | Subject | |
---|---|---|---|
Next Message | Kevin Grittner | 2013-05-15 18:18:35 | Re: counting algorithm for incremental matview maintenance |
Previous Message | Noah Misch | 2013-05-15 18:11:37 | Re: Parallel Sort |