<!-- $PostgreSQL: pgsql/doc/src/sgml/wal.sgml,v 1.29 2004/08/08 04:34:43 tgl Exp $ -->

<!-- $PostgreSQL: pgsql/doc/src/sgml/wal.sgml,v 1.30 2004/11/15 04:35:57 neilc Exp $ -->

<chapter id="wal">

<title>Write-Ahead Logging (<acronym>WAL</acronym>)</title>

</para>

<para>

<firstterm>Checkpoints</firstterm><indexterm><primary>checkpoint</></> are points in the sequence of

transactions at which it is guaranteed that the data files have

been updated with all information logged before the checkpoint. At

checkpointtime, all dirty data pages are flushed to disk and a

special checkpoint record is written to the log file. Asresult, in

the event of a crash, therecoverer knows from what point in the

log (known as the redo record) it should start the REDO operation,

since any changes made to data files before thatrecord are already

on disk. After a checkpoint has been made, any log segments written

before the redo record are no longer needed and can be recycled or

removed. (When <acronym>WAL</acronym> archiving is being done, the

<firstterm>Checkpoints</firstterm><indexterm><primary>checkpoint</></>

are points in the sequence oftransactions at which it is guaranteed

that the data files havebeen updated with all information logged before

thecheckpoint. At checkpointtime, all dirty data pages are flushed to

disk and aspecial checkpoint record is written to the log file. Asa

result, inthe event of a crash, thecrash recovery procedure knows from

what point in thelog (known as the redo record) it should start the

REDO operation,since any changes made to data files before thatpoint

are alreadyon disk. After a checkpoint has been made, any log segments

writtenbefore the redo record are no longer needed and can be recycled

orremoved. (When <acronym>WAL</acronym> archiving is being done, the

log segments must be archived before being recycled or removed.)

</para>