Re: Extended statistics is not working on Vars hidden under a RelabelType

On 15 October 2017 at 06:49, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
> On Fri, Oct 13, 2017 at 4:49 PM, David Rowley
> <david(dot)rowley(at)2ndquadrant(dot)com> wrote:
>> tps = 8282.481310 (including connections establishing)
>> tps = 8282.750821 (excluding connections establishing)
>
> vs.
>
>> tps = 8520.822410 (including connections establishing)
>> tps = 8521.132784 (excluding connections establishing)
>>
>> With the patch we are making use of the extended statistics, which we
>> do expect to be more work for the planner. Although, we didn't add
>> extended statistics to speed up the planner.
>
> Sure, I understand. That's actually a pretty substantial regression -
> I guess that means that it's pretty important to avoid creating
> extended statistics that are not needed, at least for short-running
> queries.

To be honest, I ran that on a VM on my laptop. I was getting quite a
bit of noise. I just posted that to show that the 12x slowdown didn't
exist. I don't know what the actual slowdown is. I just know extended
stats are not free and that nobody expected that they ever would be.
The good news is that they're off by default and if the bad ever
outweighs the good then the fix for that starts with "DROP STATISTICS"

I personally think it's great we're starting to see a useful feature
materialise that can help with poor row estimates from the planner.

--
David Rowley http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services