From 377cc45194f4f1bd6eb68bd6eb42402e46e326d7 Mon Sep 17 00:00:00 2001 From: Daniel Gustafsson Date: Mon, 25 May 2026 12:15:29 +0200 Subject: [PATCH] doc: Clarify CHECKPOINT handling of unlogged buffers The CHECKPOINT reference page still described checkpoints as flushing all data files, which could be misleading as it depends on the value of FLUSH_UNLOGGED option. Update the description to make it clearer that only data files of permanent relations are flushed by default. Author: Chao Li Reviewed-by: Daniel Gustafsson Discussion: https://postgr.es/m/4855807D-F1CA-44E6-9B58-406691832848@gmail.com --- doc/src/sgml/ref/checkpoint.sgml | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/doc/src/sgml/ref/checkpoint.sgml b/doc/src/sgml/ref/checkpoint.sgml index cd981cf2cab..08dbe62c612 100644 --- a/doc/src/sgml/ref/checkpoint.sgml +++ b/doc/src/sgml/ref/checkpoint.sgml @@ -35,8 +35,10 @@ CHECKPOINT [ ( option [, ...] ) ] A checkpoint is a point in the write-ahead log sequence at which - all data files have been updated to reflect the information in the - log. All data files will be flushed to disk. Refer to + all data files for permanent relations have been updated to reflect the + information in the log. All data for permanent relations files will be + flushed to disk, dirty buffers of unlogged relations are not flushed unless + FLUSH_UNLOGGED is specified. Refer to for more details about what happens during a checkpoint.