Re: Gather performance analysis

On Fri, Sep 24, 2021 at 2:01 AM Robert Haas <robertmhaas@gmail.com> wrote:
>
> On Thu, Sep 23, 2021 at 4:00 PM Tomas Vondra
> <tomas.vondra@enterprisedb.com> wrote:
> > I did find some suspicious behavior on the bigger box I have available
> > (with 2x xeon e5-2620v3), see the attached spreadsheet. But it seems
> > pretty weird because the worst affected case is with no parallel workers
> > (so the queue changes should affect it). Not sure how to explain it, but
> > the behavior seems consistent.
>
> That is pretty odd. I'm inclined to mostly discount the runs with
> 10000 tuples because sending such a tiny number of tuples doesn't
> really take any significant amount of time, and it seems possible that
> variations in the runtime of other code due to code movement effects
> could end up mattering more than the changes to the performance of
> shm_mq. However, the results with a million tuples seem like they're
> probably delivering statistically significant results ... and I guess
> maybe what's happening is that the patch hurts when the tuples are too
> big relative to the queue size.

I am looking at the "query-results.ods" file shared by Tomas, with a
million tuple I do not really see where the patch hurts? because I am
seeing in most of the cases the time taken by the patch is 60-80%
compared to the head.  And the worst case with a million tuple is
100.32% are are we pointing to that 0.32% or there is something else
that I am missing here.

>
> I guess your columns are an md5 value each, which is 32 bytes +
> overhead, so a 20-columns tuple is ~1kB. Since Dilip's patch flushes
> the value to shared memory when more than a quarter of the queue has
> been filled, that probably means we flush every 4-5 tuples. I wonder
> if that means we need a smaller threshold, like 1/8 of the queue size?
> Or maybe the behavior should be adaptive somehow, depending on whether
> the receiver ends up waiting for data? Or ... perhaps only small
> tuples are worth batching, so that the threshold for posting to shared
> memory should be a constant rather than a fraction of the queue size?
> I guess we need to know why we see the time spike up in those cases,
> if we want to improve them.

I will test with the larger tuple sizes and will see the behavior with
different thresholds.  With 250 bytes tuple size, I have tested with
different thresholds and it appeared that 1/4 of the queue size works
best.  But I will do more detailed testing and share the results.

-- 
Regards,
Dilip Kumar
EnterpriseDB: http://www.enterprisedb.com