Re: Fix pg_stat_reset_single_table_counters function

Masahiro Ikeda <ikedamsh@oss.nttdata.com>

From: Masahiro Ikeda <ikedamsh@oss.nttdata.com>
To: pgsql-hackers@lists.postgresql.org
Cc: mitsuru.hinata.5432@gmail.com
Date: 2023-08-10T05:09:57Z
Lists: pgsql-hackers

Attachments

On 2023-08-01 15:23, Masahiro Ikeda wrote:
> Hi,
> 
> My colleague, Mitsuru Hinata (in CC), found the following issue.
> 
> The documentation of pg_stat_reset_single_table_counters() says
>> pg_stat_reset_single_table_counters ( oid ) → void
>> Resets statistics for a single table or index in the current database 
>> or shared across all databases in the cluster to zero.
>> This function is restricted to superusers by default, but other users 
>> can be granted EXECUTE to run the function.
> https://www.postgresql.org/docs/devel/monitoring-stats.html
> 
> But, the stats will not be reset if the table shared across all
> databases is specified. IIUC, 5891c7a8e seemed to have mistakenly
> removed the feature implemented in e04267844. What do you think?

I fix an issue with the v1 patch reported by cfbot.

The test case didn't assume the number of databases will change in
the test of pg_upgrade.

Regards,
-- 
Masahiro Ikeda
NTT DATA CORPORATION

Commits

  1. Fix pg_stat_reset_single_table_counters() for shared relations