Horton

Re: Possible regression in PG18 beta1

Sadeq Dousti <msdousti@gmail.com>

View thread
From: Sadeq Dousti <msdousti@gmail.com>
To: Sami Imseih <samimseih@gmail.com>
Cc: Christophe Courtois <christophe.courtois@dalibo.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2025-05-18T15:56:24Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. EXPLAIN: Always use two fractional digits for row counts.

> In my tests I build from source from both the 17 stable and head branch
> and use the same build options.
>

OK, I built 17 from REL_17_STABLE, and got the same behavior as the
Homebrew PG17:

* First run: 135 ms
* Subsequent runs: 169 ms, 181 ms, 187 ms, 177 ms, 170 ms

For comparison, with PG18:
* First run: 150 ms
* Subsequent runs: 183 ms, 182 ms, 185 ms, 187 ms, 180 ms

On the first and second runs, PG17 is consistently faster.
It's also strange to me that subsequent runs are consistently slower than
the first run.

Best Regards,
Sadeq Dousti