Re: POC: Cache data in GetSnapshotData()

On Thu, Mar 17, 2016 at 9:00 AM, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com>
wrote:
>If you see, for the Base readings, there is a performance increase up till
64 clients and then there is a fall at 88 clients, which to me >indicates
that it hits very high-contention around CLogControlLock at 88 clients
which CLog patch is able to control to a great degree (if >required, I
think the same can be verified by LWLock stats data). One reason for
hitting contention at 88 clients is that this machine >seems to have
64-cores (it has 8 sockets and 8 Core(s) per socket) as per below
information of lscpu command.

I am attaching LWLock stats data for above test setups(unlogged tables).

*BASE* At 64 clients Block-time unit At 88 clients Block-time unit
ProcArrayLock 182946 117827
ClogControlLock 107420 120266
*clog patch*

ProcArrayLock 183663 121215
ClogControlLock 72806 65220
*clog patch + save snapshot*

ProcArrayLock 128260 83356
ClogControlLock 78921 74011
This is for unlogged lables, I mainly see ProcArrayLock have higher
contention at 64 clients and at 88 contention is slightly moved to other
entities.

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