Re: LWLocks in DSM memory

On Wed, Aug 17, 2016 at 9:12 PM, Andres Freund <andres(at)anarazel(dot)de> wrote:

> >Yes. I want a long wait list, modified in bulk - which should be the
> >case with the above.
>

I ran some pgbench. And, I do not see much difference in performance, small
variance in perf can be attributed to variance in probability of drawing
the particular built-in script.

Server configuration:
./postgres -c shared_buffers=8GB -N 200 -c
min_wal_size=15GB -c max_wal_size=20GB -c checkpoint_timeout=900 -c
maintenance_work_mem=1GB -c checkpoint_completion_target=0.9 &

pgbench configuration: initialized with scale_factor 300
./pgbench -c $threads -j $threads -T 1800 -M prepared -b simple-update(at)1
-b select-only(at)20 postgres

Simple-update : select-only = 5:95

*cilents* *8* *64* *128*
*Current Code* 38279.663784 258196.067342 283150.495921
*After Patch revert* 37316.09022 268285.506338 280077.913954
*% diff * -2.5171944284 3.9076656356 -1.0851409449

Simple-update : selec-only = 20:80

*cilents* *8* *64* *128*
*Current Code* 23169.021469 134791.996882 154057.101004
*After Patch revert* 23892.91539 135091.645551 150402.80543
%diff 3.1244043775 0.2223044958 -2.3720396854

And this was done on our 8 socket intel machine machine

--
Thanks and Regards
Mithun C Y
EnterpriseDB: http://www.enterprisedb.com