Re: Overhead for stats_command_string et al, take 2

Tom Lane wrote:
> Stefan Kaltenbrunner <stefan(at)kaltenbrunner(dot)cc> writes:
>> Tom Lane wrote:
>>> It'd be interesting to compare 8.1 and HEAD for the no-overhead case;
>>> I don't think you need to redo all four cases, but I'd like to see that one.
>
>> 8.1: 50,50,49
>> HEAD: 49,48,49
>
> OK, so that seems comparable to my results on a dual Xeon ... probably,
> both your machine and my newer one have fast-to-read clock hardware.
> We need to get some numbers from one of the people who have complained
> about EXPLAIN ANALYZE overhead.
>

Data from two (identical) dual P-III, one running Linux and one running
Freebsd - both doing the 100000 SELECT 1 test:

Freebsd 6.1:
- 8.1 21.5 (median times)
- HEAD 22.2

Linux 2.6.16
- 8.1 16.1
- HEAD 17.2

The variation in run times seems to be up to 0.5 seconds, so I'm not
sure that I'm seeing a real difference between 8.1 and HEAD (though this
test seems to run noticeably slower on Freebsd - recall from my previous
posting featuring these boxes that EXPLAIN ANALYZE seems to have a
*much* higher overhead on Freebsd).

(8.1 is 8.1.3 on the Freebsd box and 8.1.4 on the linux one. HEAD is
from today).

regards

Mark