Re: Introduce a new view for checkpointer related stats - Mailing list pgsql-hackers

From Bharath Rupireddy
Subject Re: Introduce a new view for checkpointer related stats
Date
Msg-id CALj2ACV7hGqu95A7smwEM3SnY1OsyzGm+BicqrUH4VeWxyuLiw@mail.gmail.com
Whole thread Raw
In response to Re: Introduce a new view for checkpointer related stats  (sirisha chamarthi <sirichamarthi22@gmail.com>)
List pgsql-hackers
On Fri, Dec 2, 2022 at 12:54 PM sirisha chamarthi
<sirichamarthi22@gmail.com> wrote:
>
> On Thu, Dec 1, 2022 at 9:50 PM Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com> wrote:
>>
>> On Wed, Nov 30, 2022 at 5:15 PM Bharath Rupireddy
>> <bharath.rupireddyforpostgres@gmail.com> wrote:
>> >
>> > I don't have a strong opinion about changing column names. However, if
>> > we were to change it, I prefer to use names that
>> > PgStat_CheckpointerStats has. BTW, that's what
>> > PgStat_BgWriterStats/pg_stat_bgwriter and
>> > PgStat_ArchiverStats/pg_stat_archiver uses.
>>
>> After thinking about this a while, I convinced myself to change the
>> column names to be a bit more meaningful. I still think having
>> checkpoints in the column names is needed because it also has other
>> backend related columns. I'm attaching the v4 patch for further
>> review.
>> CREATE VIEW pg_stat_checkpointer AS
>>     SELECT
>>         pg_stat_get_timed_checkpoints() AS timed_checkpoints,
>>         pg_stat_get_requested_checkpoints() AS requested_checkpoints,
>>         pg_stat_get_checkpoint_write_time() AS checkpoint_write_time,
>>         pg_stat_get_checkpoint_sync_time() AS checkpoint_sync_time,
>>         pg_stat_get_buf_written_checkpoints() AS buffers_written_checkpoints,
>>         pg_stat_get_buf_written_backend() AS buffers_written_backend,
>>         pg_stat_get_buf_fsync_backend() AS buffers_fsync_backend,
>>         pg_stat_get_checkpointer_stat_reset_time() AS stats_reset;
>
>
> IMO, “buffers_written_checkpoints” is confusing. What do you think?

Thanks. We can be "more and more" meaningful by naming
buffers_written_by_checkpoints, buffers_written_by_backend,
buffers_fsync_by_backend. However, I don't think that's a good idea
here as names get too long.

Having said that, I'll leave it to the committer's discretion.

--
Bharath Rupireddy
PostgreSQL Contributors Team
RDS Open Source Databases
Amazon Web Services: https://aws.amazon.com



pgsql-hackers by date:

Previous
From: Richard Guo
Date:
Subject: Re: Missing MaterialPath support in reparameterize_path_by_child
Next
From: Peter Eisentraut
Date:
Subject: refactor ExecGrant_*() functions