Horton

Re: pgbench logging broken by time logic changes

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Thomas Munro <thomas.munro@gmail.com>
Cc: Gregory Smith <gregsmithpgsql@gmail.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>, david.christensen@crunchydata.com
Date: 2021-06-17T00:46:42Z
Lists: pgsql-hackers 
On Thu, Jun 17, 2021 at 12:36:10PM +1200, Thomas Munro wrote:
> For that reason, I'm not super attached to that new pg_time_usec_t
> stuff at all, and wouldn't be sad if we reverted that piece.  I am
> moderately attached to the sync changes, though.  pgbench 13 is
> objectively producing incorrect results in that respect.

There is another item in this area where pgbench uses incorrect maths
when aggregating the stats of transactions mid-run and at the end of a
thread, issue caused by 547f04e as this code path forgot to handle the
s <-> us conversion:
https://www.postgresql.org/message-id/CAF7igB1r6wRfSCEAB-iZBKxkowWY6+dFF2jObSdd9+iVK+vHJg@mail.gmail.com

Wouldn't it be better to put all those fixes into the same bag?  If
you drop the business with pg_time_usec_t, it looks like we don't
really need to do anything there.
--
Michael

Commits

  1. Fix pgbench timestamp bugs.

  2. Fix pattern matching logic for logs in TAP tests of pgbench

  3. pgbench: Improve time logic.

  4. pgbench: Synchronize client threads.

  5. pgbench: refactor handling of stats tracking