Re: [PATCH] Add memory usage reporting to VACUUM VERBOSE

From: Tatsuya Kawata <kawatatatsuya0913(at)gmail(dot)com>
To: Chao Li <li(dot)evan(dot)chao(at)gmail(dot)com>
Cc: Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com>, pgsql-hackers(at)postgresql(dot)org
Subject: Re: [PATCH] Add memory usage reporting to VACUUM VERBOSE
Date: 2025-12-30 03:23:24
Message-ID: CAHza6qdibhqf17uuAUZ_w67Q3U9a4-nGYCkseLYiRWGF0tDhrg@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi Sawada-san and Chao-san,

Thank you both for your continued reviews and feedback on this patch.I
agree with both points that could be misleading.
I've updated the message as follows:

memory usage: dead item storage %.2f MB accumulated across %d resets (limit
%.2f MB each)

This wording:
- Use both "memory usage:" prefix and "dead item storage" to be consistent
with other existing log lines (buffer usage:, WAL usage:, system usage:)
and to clarify what kind of memory we're reporting
- Clearly indicate that the value is a cumulative sum of snapshots, not
actual memory allocation
- Use "across %d resets" to show the relationship between the accumulated
value and reset count
- Keep "limit %.2f MB each" to indicate the per-reset memory limit

Please find attached the v11 patch.

Regards,
Tatsuya Kawata

Attachment Content-Type Size
v11-0001-Add-dead-items-memory-usage-to-VACUUM-VERBOSE-an.patch application/octet-stream 3.9 KB

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Chao Li 2025-12-30 03:24:37 Re: Implement waiting for wal lsn replay: reloaded
Previous Message Álvaro Herrera 2025-12-30 03:14:42 Re: Implement waiting for wal lsn replay: reloaded