Re: Should we cacheline align PGXACT?

Hi,

On Fri, Feb 24, 2017 at 12:22 PM, Ashutosh Sharma <ashu(dot)coek88(at)gmail(dot)com>
wrote:

On Fri, Feb 24, 2017 at 10:41 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
> > On 24 February 2017 at 04:41, Ashutosh Sharma <ashu(dot)coek88(at)gmail(dot)com>
> wrote:
> >>
> >> Okay. As suggested by Alexander, I have changed the order of reading and
> >> doing initdb for each pgbench run. With these changes, I got following
> >> results at 300 scale factor with 8GB of shared buffer.
> >
> >
> > Would you be able to test my patch also please?
> >
>
> Sure, I can do that and share the results by early next week. Thanks.
>

So, Here are the pgbench results I got with '
*reduce_pgxact_access_AtEOXact.v2.patch*' on a read-write workload.

*pgbench settings:*
pgbench -i -s 300 postgres
pgbench -M prepared -c $thread -j $thread -T $time_for_reading postgres

where, time_for_reading = 30mins

*non default GUC param*
shared_buffers=8GB
max_connections=300

pg_wal is located in SSD.

CLIENT COUNT TPS (HEAD) TPS (PATCH) % IMPROVEMENT
4 2588 2601 0.5023183926
8 5094 5098 0.0785237534
16 10294 10307 0.1262871576
32 19779 19815 0.182011224
64 27908 28346 1.569442454
72 27823 28416 2.131330194
128 28455 28618 0.5728342998
180 26739 26879 0.5235797898
196 27820 27963 0.5140186916
256 28763 28969 0.7161978931

Also, Excel sheet (results-readwrite-300-SF.xlsx) containing the results
for all the 3 runs is attached.