Re: Choosing parallel_degree

On 12 April 2016 at 07:58, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:

> On Tue, Apr 12, 2016 at 3:15 AM, Simon Riggs <simon(at)2ndquadrant(dot)com>
> wrote:
>
>> On 8 April 2016 at 17:49, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
>>
>>
>>> With the patch, you can - if you wish - substitute
>>> some other number for the one the planner comes up with.
>>
>>
>> I saw you're using AccessExclusiveLock, the reason being it affects
>> SELECTs.
>>
>> That is supposed to apply when things might change the answer from a
>> SELECT, whereas this affects only the default for a plan.
>>
>>
> By this theory, shouldn't any other parameter like n_distinct_inherited
> which just effects the plan required lower lock level?
>

It should, yes, and I'm as surprised to see it isn't as you are.

Thread: Fabrizio was asked by Robert to provide or document an analysis of
why each setting was OK to change; 9 days later he had not done so or
replied, so I committed a reduced version of the patch that matched
existing tests and code comments.

I guess we could have salvaged some more from it, but we didn't and there's
never enough time.

If RMT allows, that can be changed or it can wait.

--
Simon Riggs http://www.2ndQuadrant.com/
<http://www.2ndquadrant.com/>
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services