Re: Choosing parallel_degree

On 17/03/2016 11:23, Julien Rouhaud wrote:
> On 17/03/2016 02:07, James Sewell wrote:
>>
>> On Thu, Mar 17, 2016 at 5:05 AM, Julien Rouhaud
>> <julien(dot)rouhaud(at)dalibo(dot)com <mailto:julien(dot)rouhaud(at)dalibo(dot)com>> wrote:
>>
>>
>> attached v3 drops the GUC part.
>>
>>
>> This looks good good. I do think that some threshold control would be
>> good in the long term - but you are right Robert it just feels strange.
>>
>> Maybe once the final formula is implemented in 9.7+ and this gets some
>> real world use cases it can be revisited?
>>
>> One thing I really, really like about the way the new patch works is
>> that you can set parallel_degree on an inheritance parent, then that
>> will set the minimum workers for all associated children (when accessing
>> from the parent).
>>
>> Currently this patch will not fire on small tables even
>> when parallel_degree is set, can we fix this by adding a check
>> for ref->parallel_degree to the table size condition?
>>
>
> Actually, a parallel plan will be created, since in this case the
> rel->reloptkind will be RELOPT_OTHER_MEMBER_REL, not RELOPT_BASEREL.
>

And after re-reading your mail I see that it was what you meant, sorry :)

With the current threshold, you need a table bigger than 8 MB to be able
to force parallel workers. I'm not sure there'll be benefits for
multiple workers on a table smaller than 8 MB, since setting up all the
parallel stuff takes time.

--
Julien Rouhaud
http://dalibo.com - http://dalibo.org