Message-ID: <gh-pgjdbc-pgjdbc-4070@github.com>
From: "davecramer (@davecramer)" <noreply+davecramer@github.com>
To: "pgjdbc/pgjdbc" <noreply+pgjdbc-pgjdbc@github.com>
Date: Thu, 14 May 2026 14:28:12 +0000
Subject: [pgjdbc/pgjdbc] PR #4070: feat: add pipelineMode connection property for protocol-level pipelining
List-Id: <gh-pgjdbc-pgjdbc.github.com>
X-GitHub-Author-Id: 406518
X-GitHub-Author-Login: davecramer
X-GitHub-Issue: 4070
X-GitHub-Repo: pgjdbc/pgjdbc
X-GitHub-State: closed
X-GitHub-Type: pull_request
X-GitHub-Url: https://github.com/pgjdbc/pgjdbc/pull/4070
Content-Type: text/plain; charset=utf-8

Implement synchronous pipelining (Option C) that sends multiple extended protocol queries before reading responses, reducing round-trip latency for batch operations.

Changes:
- Add PIPELINE_MODE connection property (default false)
- Add isPipelineModeEnabled() to QueryExecutor interface
- Raise flushIfDeadlockRisk() threshold from 64KB to 8MB when pipeline mode is enabled, allowing more queries to be buffered before forcing an intermediate Sync+read cycle
- Add functional and performance tests

This approach avoids the socket-sharing race conditions of the previous reader thread design by keeping all I/O on the main thread.