Re: pgbench regression test failure

From: Steve Singer <steve(at)ssinger(dot)info>
To: pgsql-hackers(at)postgresql(dot)org
Cc: Fabien Coelho <postgresql(dot)org(at)coelho(dot)net>
Subject: Re: pgbench regression test failure
Date: 2017-11-12 02:47:39
Message-ID: 20171112024739.23750.54311.pgcf@coridan.postgresql.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

The following review has been posted through the commitfest application:
make installcheck-world: tested, failed
Implements feature: not tested
Spec compliant: not tested
Documentation: not tested

This causes the pgbench tests to fail (consistently) with

not ok 194 - pgbench late throttling stdout /(?^:processed: [01]/10)/

When I run pgbench manually I get (-t 10 --rate=100000 --latency-limit=1 -n -r)

number of transactions actually processed: 10/10
number of transactions skipped: 10 (100.000 %)

Prior to the patch I was getting.

number of transactions actually processed: 0/10
number of transactions skipped: 10 (100.000 %)

@@ -3539,7 +3542,7 @@ printResults(TState *threads, StatsData *total, instr_time total_time,^M
{^M
printf("number of transactions per client: %d\n", nxacts);^M
printf("number of transactions actually processed: " INT64_FORMAT "/%d\n",^M
- total->cnt - total->skipped, nxacts * nclients);^M
+ total->cnt, nxacts * nclients);^M

I think you want ntx instead of total->cnt here.

The new status of this patch is: Waiting on Author

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Thomas Munro 2017-11-12 03:48:49 Re: A GUC to prevent leader processes from running subplans?
Previous Message Mark Dilger 2017-11-12 00:21:09 Re: Simplify ACL handling for large objects and removal of superuser() checks