Re: Auto-tuning work_mem and maintenance_work_mem

On 18/02/14 03:48, Tom Lane wrote:
> Gavin Flower <GavinFlower@archidevsys.co.nz> writes:
>> On 17/02/14 15:26, Robert Haas wrote:
>>> I don't really know about cpu_tuple_cost.  Kevin's often advocated
>>> raising it, but I haven't heard anyone else advocate for that.  I
>>> think we need data points from more people to know whether or not
>>> that's a good idea in general.
>> Processors have been getting faster, relative to spinning rust, over the
>> years.  So it puzzles me why anybody would want to raise the
>> cpu_tuple_cost!
> The case where this is sensible is where your database mostly fits in
> RAM, so that the cost of touching the underlying spinning rust isn't
> so relevant.  The default cost settings are certainly not very good
> for such scenarios.
>
>             regards, tom lane
Thanks.

That is obvious... once you pointed it out!


Cheers,
Gavin