Message-ID: <gh-pgjdbc-pgjdbc-3870-c3575480786@github.com>
From: "vlsi (@vlsi)" <noreply+vlsi@github.com>
To: "pgjdbc/pgjdbc" <noreply+pgjdbc-pgjdbc@github.com>
Date: Tue, 25 Nov 2025 12:46:12 +0000
Subject: Re: [pgjdbc/pgjdbc] PR #3870: feat: Add INFO log when fetchSize is ignored under autocommit=true
In-Reply-To: <gh-pgjdbc-pgjdbc-3870@github.com>
References: <gh-pgjdbc-pgjdbc-3870@github.com>
List-Id: <gh-pgjdbc-pgjdbc.github.com>
X-GitHub-Author-Login: vlsi
X-GitHub-Comment-Id: 3575480786
X-GitHub-Comment-Type: issue_comment
X-GitHub-Edited-At: 2025-11-25T12:46:32Z
X-GitHub-Issue: 3870
X-GitHub-Repo: pgjdbc/pgjdbc
X-GitHub-Type: comment
X-GitHub-Url: https://github.com/pgjdbc/pgjdbc/pull/3870#issuecomment-3575480786
Content-Type: text/plain; charset=utf-8

I am afraid it might result in logging lots of extra log messages (the code would generate a message for every execution), and it will be useless if the query produces just a few rows. It would be false positives.

In the meantime, what do you think of the following instead?
a) Add a custom property for tracking performance or stability-related issues. For instance `stabilityIssues=ignore|warn|fail`. Then pgjdbc would either ignore (the default), or log a warning (`warn`), or throw an exception so the user knows which exact functionality uses a bad code pattern

b) Log a warning (~`INFO`) only in case the number of fetched rows exceeds certain threshold. It would reduce the number of false-positives. I don't like hard-coding the thresholds though.