Re: Wait free LW_SHARED acquisition - v0.9

On Sat, Oct 11, 2014 at 7:00 AM, Andres Freund <andres(at)2ndquadrant(dot)com>
wrote:
> On 2014-10-11 06:49:54 +0530, Amit Kapila wrote:
> > On Sat, Oct 11, 2014 at 6:29 AM, Andres Freund <andres(at)2ndquadrant(dot)com>
> > wrote:
> > >
> > > On 2014-10-11 06:18:11 +0530, Amit Kapila wrote:
> > > I've run some short tests on hydra:
> > >
>
> > Could you please post numbers you are getting for 3000 scale
> > factor for client count 128 and 175?
>
> Yes, although not tonight....

No issues, whenever you get it.

> Also from hydra?

Yes. One more thing I would like to share with you is that while doing
tests, there are some other settings change in postgresql.conf

maintenance_work_mem = 1GB
synchronous_commit = off
wal_writer_delay = 20ms
checkpoint_segments=256
checkpoint_timeout =35min

I don't think these parameters matter for the tests we are doing, but
still I thought it is good to share, because I forgot to send some of
these non-default settings in previous mail.

> > > > Nothing specific, for performance tests where I have to take
profiles
> > > > I use below:
> > > > ./configure --prefix=<installation_path>
> > CFLAGS="-fno-omit-frame-pointer"
> > > > make
> > >
> > > Hah. Doing so overwrites the CFLAGS configure normally sets. Check
> > > # CFLAGS are selected so:
> > > # If the user specifies something in the environment, that is used.
> > > # else: If the template file set something, that is used.
> > > # else: If coverage was enabled, don't set anything.
> > > # else: If the compiler is GCC, then we use -O2.
> > > # else: If the compiler is something else, then we use -O, unless
> > debugging.
> > >
> > > so, if you do like above, you're compiling without optimizations...
So,
> > > include at least -O2 as well.
> >
> > Hmm. okay, but is this required when we do actual performance
> > tests, because for that currently I don't use CFLAGS.
>
> I'm not sure what you mean? You need to include -O2 in CFLAGS whenever
> you override it.

okay, thats what I wanted to ask you, so that we should not see different
numbers due to the way code is built.

When I do performance tests where I don't want to see profile,
I use below statement:
./configure --prefix=<installation_path>

> And since
> your general performance numbers are a fair bit lower than what I see
> with, hopefully, the same code on the same machine...

You have reported numbers at 1000 scale factor and mine were
at 3000 scale factor, so I think the difference is expected.

With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com