Horton

Re: Proposals for EXPLAIN: rename ANALYZE to EXECUTE and extend VERBOSE

Fujii Masao <masao.fujii@oss.nttdata.com>

View thread View on postgr.es
From: Fujii Masao <masao.fujii@oss.nttdata.com>
To: Jelte Fennema-Nio <postgres@jeltef.nl>, David Rowley <dgrowleyml@gmail.com>
Cc: Michael Christofides <michael@pgmustard.com>, Guillaume Lelarge <guillaume@lelarge.info>, Greg Sabino Mullane <htamfids@gmail.com>, Robert Haas <robertmhaas@gmail.com>, "David G. Johnston" <david.g.johnston@gmail.com>, Nikolay Samokhvalov <samokhvalov@gmail.com>, Tom Lane <tgl@sss.pgh.pa.us>, pgsql-hackers@postgresql.org
Date: 2025-05-30T14:36:40Z
Lists: pgsql-hackers 

On 2025/05/30 21:43, Jelte Fennema-Nio wrote:
> On Fri May 30, 2025 at 1:55 PM CEST, David Rowley wrote:
>> By my count, those who think we should change log_buffers to true are:
>> Robert [1], Michael Christofides [2], Fujii [3].
> 
> +1 for me on changing the default. It just seems confusing to have
> different defaults for auto_explain and EXPLAIN. I'd even vote for
> making that change still for PG18, since this feels like an oversight
> that the defaults were not kept aligned.
> 
> 
>> log_min_duration = 0, log_buffers = off: 11231 tps
>> log_min_duration = 0, log_buffers = on: 11106 tps
>> log_min_duration = -1 (off): 13297 tps
>>
>> So it is possible to measure it, at least.
> 
> Seems small enough not to worry about it.

Users can simply disable log_buffers if they enable log_analyze
but don't need buffer usage details and want to avoid the overhead.
If there were no way to avoid the overhead, I'd agree not to make
this change. But since we can disable log_buffers when needed,
I think it's acceptable.

Regards,

-- 
Fujii Masao
NTT DATA Japan Corporation

Commits

  1. Fix further fallout from EXPLAIN ANALYZE BUFFERS change

  2. Add missing BUFFERS OFF in regression tests, take 2

  3. Add missing BUFFERS OFF in select_into regression tests

  4. Enable BUFFERS with EXPLAIN ANALYZE by default