To: pgsql-hackers@lists.postgresql.org
Cc: Grigory Smolkin <g.smolkin@postgrespro.ru>
From: Arthur Zakirov <a.zakirov@postgrespro.ru>
Subject: [PATCH] xlogreader: do not read a file block twice
Message-ID: <2ddf4a32-517e-d6f4-d992-4a63b6035bfd@postgrespro.ru>
Date: Mon, 11 Feb 2019 19:25:39 +0300
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.5.0
MIME-Version: 1.0
Content-Type: multipart/mixed;
 boundary="------------1F92CF77D83EF3F8A4B8F2C7"
Content-Language: en-US
Precedence: bulk

This is a multi-part message in MIME format.
--------------1F92CF77D83EF3F8A4B8F2C7
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit

Hello hackers,

Grigory noticed that one of our utilities has very slow performance when 
xlogreader reads zlib archives. We found out that xlogreader sometimes 
reads a WAL file block twice.

zlib has slow performance when you read an archive not in sequential 
order. I think reading a block twice in same position isn't sequential, 
because gzread() moves current position forward and next call gzseek() 
to the same position moves it back.

It seems that the attached patch solves the issue. I think when reqLen 
== state->readLen the requested block already is in the xlogreader's buffer.

What do you think?

-- 
Arthur Zakirov
Postgres Professional: http://www.postgrespro.com
Russian Postgres Company

--------------1F92CF77D83EF3F8A4B8F2C7
Content-Type: text/x-patch;
 name="xlogreader-reqlen.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="xlogreader-reqlen.patch"

diff --git a/src/backend/access/transam/xlogreader.c b/src/backend/access/transam/xlogreader.c
index 69b6226f8f..cbc7e4e7ea 100644
--- a/src/backend/access/transam/xlogreader.c
+++ b/src/backend/access/transam/xlogreader.c
@@ -536,7 +536,7 @@ ReadPageInternal(XLogReaderState *state, XLogRecPtr pageptr, int reqLen)
 
 	/* check whether we have all the requested data already */
 	if (targetSegNo == state->readSegNo && targetPageOff == state->readOff &&
-		reqLen < state->readLen)
+		reqLen <= state->readLen)
 		return state->readLen;
 
 	/*

--------------1F92CF77D83EF3F8A4B8F2C7--