wal.sgml

<!-- doc/src/sgml/wal.sgml -->

<chapter id="wal">
 <title>Reliability and the Write-Ahead Log</title>

 <para>
  This chapter explains how the Write-Ahead Log is used to obtain
  efficient, reliable operation.
 </para>

 <sect1 id="wal-reliability">
  <title>Reliability</title>

  <para>
   Reliability is an important property of any serious database
   system, and <productname>PostgreSQL</productname> does everything possible to
   guarantee reliable operation. One aspect of reliable operation is
   that all data recorded by a committed transaction should be stored
   in a nonvolatile area that is safe from power loss, operating
   system failure, and hardware failure (except failure of the
   nonvolatile area itself, of course).  Successfully writing the data
   to the computer's permanent storage (disk drive or equivalent)
   ordinarily meets this requirement.  In fact, even if a computer is
   fatally damaged, if the disk drives survive they can be moved to
   another computer with similar hardware and all committed
   transactions will remain intact.
  </para>

  <para>
   While forcing data to the disk platters periodically might seem like
   a simple operation, it is not. Because disk drives are dramatically
   slower than main memory and CPUs, several layers of caching exist
   between the computer's main memory and the disk platters.
   First, there is the operating system's buffer cache, which caches
   frequently requested disk blocks and combines disk writes. Fortunately,
   all operating systems give applications a way to force writes from
   the buffer cache to disk, and <productname>PostgreSQL</productname> uses those
   features.  (See the <xref linkend="guc-wal-sync-method"/> parameter
   to adjust how this is done.)
  </para>

  <para>
   Next, there might be a cache in the disk drive controller; this is
   particularly common on <acronym>RAID</acronym> controller cards. Some of
   these caches are <firstterm>write-through</firstterm>, meaning writes are sent
   to the drive as soon as they arrive. Others are
   <firstterm>write-back</firstterm>, meaning data is sent to the drive at
   some later time. Such caches can be a reliability hazard because the
   memory in the disk controller cache is volatile, and will lose its
   contents in a power failure.  Better controller cards have
   <firstterm>battery-backup units</firstterm> (<acronym>BBU</acronym>s), meaning
   the card has a battery that
   maintains power to the cache in case of system power loss.  After power
   is restored the data will be written to the disk drives.
  </para>

  <para>
   And finally, most disk drives have caches. Some are write-through
   while some are write-back, and the same concerns about data loss
   exist for write-back drive caches as for disk controller
   caches.  Consumer-grade IDE and SATA drives are particularly likely
   to have write-back caches that will not survive a power failure.  Many
   solid-state drives (SSD) also have volatile write-back caches.
  </para>

  <para>
   These caches can typically be disabled; however, the method for doing
   this varies by operating system and drive type:
  </para>

  <itemizedlist>
    <listitem>
      <para>
        On <productname>Linux</productname>, IDE and SATA drives can be queried using
        <command>hdparm -I</command>; write caching is enabled if there is
        a <literal>*</literal> next to <literal>Write cache</literal>.  <command>hdparm -W 0</command>
        can be used to turn off write caching.  SCSI drives can be queried
        using <ulink url="http://sg.danny.cz/sg/sdparm.html"><application>sdparm</application></ulink>.
        Use <command>sdparm --get=WCE</command> to check
        whether the write cache is enabled and <command>sdparm --clear=WCE</command>
        to disable it.
      </para>
    </listitem>

    <listitem>
      <para>
        On <productname>FreeBSD</productname>, IDE drives can be queried using
        <command>atacontrol</command> and write caching turned off using
        <literal>hw.ata.wc=0</literal> in <filename>/boot/loader.conf</filename>;
        SCSI drives can be queried using <command>camcontrol identify</command>,
        and the write cache both queried and changed using
        <command>sdparm</command> when available.
      </para>
    </listitem>

    <listitem>
      <para>
        On <productname>Solaris</productname>, the disk write cache is controlled by
        <command>format -e</command>.
        (The Solaris <acronym>ZFS</acronym> file system is safe with disk write-cache
        enabled because it issues its own disk cache flush commands.)
      </para>
    </listitem>

    <listitem>
      <para>
        On <productname>Windows</productname>, if <varname>wal_sync_method</varname> is
        <literal>open_datasync</literal> (the default), write caching can be disabled
        by unchecking <literal>My Computer\Open\<replaceable>disk drive</replaceable>\Properties\Hardware\Properties\Policies\Enable write caching on the disk</literal>.
        Alternatively, set <varname>wal_sync_method</varname> to
        <literal>fsync</literal> or <literal>fsync_writethrough</literal>, which prevent
        write caching.
      </para>
    </listitem>

    <listitem>
      <para>
        On <productname>macOS</productname>, write caching can be prevented by
        setting <varname>wal_sync_method</varname> to <literal>fsync_writethrough</literal>.
      </para>
    </listitem>
  </itemizedlist>

  <para>
   Recent SATA drives (those following <acronym>ATAPI-6</acronym> or later)
   offer a drive cache flush command (<command>FLUSH CACHE EXT</command>),
   while SCSI drives have long supported a similar command
   <command>SYNCHRONIZE CACHE</command>.  These commands are not directly
   accessible to <productname>PostgreSQL</productname>, but some file systems
   (e.g., <acronym>ZFS</acronym>, <acronym>ext4</acronym>) can use them to flush
   data to the platters on write-back-enabled drives.  Unfortunately, such
   file systems behave suboptimally when combined with battery-backup unit
   (<acronym>BBU</acronym>) disk controllers.  In such setups, the synchronize
   command forces all data from the controller cache to the disks,
   eliminating much of the benefit of the BBU.  You can run the
   <xref linkend="pgtestfsync"/> program to see
   if you are affected.  If you are affected, the performance benefits
   of the BBU can be regained by turning off write barriers in
   the file system or reconfiguring the disk controller, if that is
   an option.  If write barriers are turned off, make sure the battery
   remains functional; a faulty battery can potentially lead to data loss.
   Hopefully file system and disk controller designers will eventually
   address this suboptimal behavior.
  </para>

  <para>
   When the operating system sends a write request to the storage hardware,
   there is little it can do to make sure the data has arrived at a truly
   non-volatile storage area. Rather, it is the
   administrator's responsibility to make certain that all storage components
   ensure integrity for both data and file-system metadata.
   Avoid disk controllers that have non-battery-backed write caches.
   At the drive level, disable write-back caching if the
   drive cannot guarantee the data will be written before shutdown.
   If you use SSDs, be aware that many of these do not honor cache flush
   commands by default.
   You can test for reliable I/O subsystem behavior using <ulink
   url="https://brad.livejournal.com/2116715.html"><filename>diskchecker.pl</filename></ulink>.
  </para>

  <para>
   Another risk of data loss is posed by the disk platter write
   operations themselves. Disk platters are divided into sectors,
   commonly 512 bytes each.  Every physical read or write operation
   processes a whole sector.
   When a write request arrives at the drive, it might be for some multiple
   of 512 bytes (<productname>PostgreSQL</productname> typically writes 8192 bytes, or
   16 sectors, at a time), and the process of writing could fail due
   to power loss at any time, meaning some of the 512-byte sectors were
   written while others were not.  To guard against such failures,
   <productname>PostgreSQL</productname> periodically writes full page images to
   permanent WAL storage <emphasis>before</emphasis> modifying the actual page on
   disk. By doing this, during crash recovery <productname>PostgreSQL</productname> can
   restore partially-written pages from WAL.  If you have file-system software
   that prevents partial page writes (e.g., ZFS),  you can turn off
   this page imaging by turning off the <xref
   linkend="guc-full-page-writes"/> parameter. Battery-Backed Unit
   (BBU) disk controllers do not prevent partial page writes unless
   they guarantee that data is written to the BBU as full (8kB) pages.
  </para>
  <para>
   <productname>PostgreSQL</productname> also protects against some kinds of data corruption
   on storage devices that may occur because of hardware errors or media failure over time,
   such as reading/writing garbage data.
   <itemizedlist>
    <listitem>
     <para>
      Each individual record in a WAL file is protected by a CRC-32 (32-bit) check
      that allows us to tell if record contents are correct. The CRC value
      is set when we write each WAL record and checked during crash recovery,
      archive recovery and replication.
     </para>
    </listitem>
    <listitem>
     <para>
      Data pages are not currently checksummed by default, though full page images
      recorded in WAL records will be protected; see <link
      linkend="app-initdb-data-checksums"><application>initdb</application></link>
      for details about enabling data checksums.
     </para>
    </listitem>
    <listitem>
     <para>
      Internal data structures such as <filename>pg_xact</filename>, <filename>pg_subtrans</filename>, <filename>pg_multixact</filename>,
      <filename>pg_serial</filename>, <filename>pg_notify</filename>, <filename>pg_stat</filename>, <filename>pg_snapshots</filename> are not directly
      checksummed, nor are pages protected by full page writes. However, where
      such data structures are persistent, WAL records are written that allow
      recent changes to be accurately rebuilt at crash recovery and those
      WAL records are protected as discussed above.
     </para>
    </listitem>
    <listitem>
     <para>
      Individual state files in <filename>pg_twophase</filename> are protected by CRC-32.
     </para>
    </listitem>
    <listitem>
     <para>
      Temporary data files used in larger SQL queries for sorts,
      materializations and intermediate results are not currently checksummed,
      nor will WAL records be written for changes to those files.
     </para>
    </listitem>
   </itemizedlist>
  </para>
  <para>
   <productname>PostgreSQL</productname> does not protect against correctable memory errors
   and it is assumed you will operate using RAM that uses industry standard
   Error Correcting Codes (ECC) or better protection.
  </para>
 </sect1>

 <sect1 id="checksums">
  <title>Data Checksums</title>
  <indexterm>
   <primary>checksums</primary>
  </indexterm>

  <para>
   By default, data pages are not protected by checksums, but this can
   optionally be enabled for a cluster. When enabled, each data page includes
   a checksum that is updated when the page is written and verified each time
   the page is read. Only data pages are protected by checksums; internal data
   structures and temporary files are not.
  </para>

  <para>
   Checksums are normally enabled when the cluster is initialized using <link
   linkend="app-initdb-data-checksums"><application>initdb</application></link>.
   They can also be enabled or disabled at a later time as an offline
   operation. Data checksums are enabled or disabled at the full cluster
   level, and cannot be specified individually for databases or tables.
  </para>

  <para>
   The current state of checksums in the cluster can be verified by viewing the
   value of the read-only configuration variable <xref
   linkend="guc-data-checksums" /> by issuing the command <command>SHOW
   data_checksums</command>.
  </para>

  <para>
   When attempting to recover from page corruptions, it may be necessary to
   bypass the checksum protection. To do this, temporarily set the
   configuration parameter <xref linkend="guc-ignore-checksum-failure" />.
  </para>

  <sect2 id="checksums-offline-enable-disable">
   <title>Off-line Enabling of Checksums</title>

   <para>
    The <link linkend="app-pgchecksums"><application>pg_checksums</application></link>
    application can be used to enable or disable data checksums, as well as
    verify checksums, on an offline cluster.
   </para>

  </sect2>
 </sect1>

  <sect1 id="wal-intro">
   <title>Write-Ahead Logging (<acronym>WAL</acronym>)</title>

   <indexterm zone="wal">
    <primary>WAL</primary>
   </indexterm>

   <indexterm>
    <primary>transaction log</primary>
    <see>WAL</see>
   </indexterm>

   <para>
    <firstterm>Write-Ahead Logging</firstterm> (<acronym>WAL</acronym>)
    is a standard method for ensuring data integrity.  A detailed
    description can be found in most (if not all) books about
    transaction processing. Briefly, <acronym>WAL</acronym>'s central
    concept is that changes to data files (where tables and indexes
    reside) must be written only after those changes have been logged,
    that is, after log records describing the changes have been flushed
    to permanent storage. If we follow this procedure, we do not need
    to flush data pages to disk on every transaction commit, because we
    know that in the event of a crash we will be able to recover the
    database using the log: any changes that have not been applied to
    the data pages can be redone from the log records.  (This is
    roll-forward recovery, also known as REDO.)
   </para>

   <tip>
    <para>
     Because <acronym>WAL</acronym> restores database file
     contents after a crash, journaled file systems are not necessary for
     reliable storage of the data files or WAL files.  In fact, journaling
     overhead can reduce performance, especially if journaling
     causes file system <emphasis>data</emphasis> to be flushed
     to disk.  Fortunately, data flushing during journaling can
     often be disabled with a file system mount option, e.g.,
     <literal>data=writeback</literal> on a Linux ext3 file system.
     Journaled file systems do improve boot speed after a crash.
    </para>
   </tip>


   <para>
    Using <acronym>WAL</acronym> results in a
    significantly reduced number of disk writes, because only the log
    file needs to be flushed to disk to guarantee that a transaction is
    committed, rather than every data file changed by the transaction.
    The log file is written sequentially,
    and so the cost of syncing the log is much less than the cost of
    flushing the data pages.  This is especially true for servers
    handling many small transactions touching different parts of the data
    store.  Furthermore, when the server is processing many small concurrent
    transactions, one <function>fsync</function> of the log file may
    suffice to commit many transactions.
   </para>

   <para>
    <acronym>WAL</acronym> also makes it possible to support on-line
    backup and point-in-time recovery, as described in <xref
    linkend="continuous-archiving"/>.  By archiving the WAL data we can support
    reverting to any time instant covered by the available WAL data:
    we simply install a prior physical backup of the database, and
    replay the WAL log just as far as the desired time.  What's more,
    the physical backup doesn't have to be an instantaneous snapshot
    of the database state &mdash; if it is made over some period of time,
    then replaying the WAL log for that period will fix any internal
    inconsistencies.
   </para>
  </sect1>

 <sect1 id="wal-async-commit">
  <title>Asynchronous Commit</title>

   <indexterm>
    <primary>synchronous commit</primary>
   </indexterm>

   <indexterm>
    <primary>asynchronous commit</primary>
   </indexterm>

  <para>
   <firstterm>Asynchronous commit</firstterm> is an option that allows transactions
   to complete more quickly, at the cost that the most recent transactions may
   be lost if the database should crash.  In many applications this is an
   acceptable trade-off.
  </para>

  <para>
   As described in the previous section, transaction commit is normally
   <firstterm>synchronous</firstterm>: the server waits for the transaction's
   <acronym>WAL</acronym> records to be flushed to permanent storage
   before returning a success indication to the client.  The client is
   therefore guaranteed that a transaction reported to be committed will
   be preserved, even in the event of a server crash immediately after.
   However, for short transactions this delay is a major component of the
   total transaction time.  Selecting asynchronous commit mode means that
   the server returns success as soon as the transaction is logically
   completed, before the <acronym>WAL</acronym> records it generated have
   actually made their way to disk.  This can provide a significant boost
   in throughput for small transactions.
  </para>

  <para>
   Asynchronous commit introduces the risk of data loss. There is a short
   time window between the report of transaction completion to the client
   and the time that the transaction is truly committed (that is, it is
   guaranteed not to be lost if the server crashes).  Thus asynchronous
   commit should not be used if the client will take external actions
   relying on the assumption that the transaction will be remembered.
   As an example, a bank would certainly not use asynchronous commit for
   a transaction recording an ATM's dispensing of cash.  But in many
   scenarios, such as event logging, there is no need for a strong
   guarantee of this kind.
  </para>

  <para>
   The risk that is taken by using asynchronous commit is of data loss,
   not data corruption.  If the database should crash, it will recover
   by replaying <acronym>WAL</acronym> up to the last record that was
   flushed.  The database will therefore be restored to a self-consistent
   state, but any transactions that were not yet flushed to disk will
   not be reflected in that state.  The net effect is therefore loss of
   the last few transactions.  Because the transactions are replayed in
   commit order, no inconsistency can be introduced &mdash; for example,
   if transaction B made changes relying on the effects of a previous
   transaction A, it is not possible for A's effects to be lost while B's
   effects are preserved.
  </para>

  <para>
   The user can select the commit mode of each transaction, so that
   it is possible to have both synchronous and asynchronous commit
   transactions running concurrently.  This allows flexible trade-offs
   between performance and certainty of transaction durability.
   The commit mode is controlled by the user-settable parameter
   <xref linkend="guc-synchronous-commit"/>, which can be changed in any of
   the ways that a configuration parameter can be set.  The mode used for
   any one transaction depends on the value of
   <varname>synchronous_commit</varname> when transaction commit begins.
  </para>

  <para>
   Certain utility commands, for instance <command>DROP TABLE</command>, are
   forced to commit synchronously regardless of the setting of
   <varname>synchronous_commit</varname>.  This is to ensure consistency
   between the server's file system and the logical state of the database.
   The commands supporting two-phase commit, such as <command>PREPARE
   TRANSACTION</command>, are also always synchronous.
  </para>

  <para>
   If the database crashes during the risk window between an
   asynchronous commit and the writing of the transaction's
   <acronym>WAL</acronym> records,
   then changes made during that transaction <emphasis>will</emphasis> be lost.
   The duration of the
   risk window is limited because a background process (the <quote>WAL
   writer</quote>) flushes unwritten <acronym>WAL</acronym> records to disk
   every <xref linkend="guc-wal-writer-delay"/> milliseconds.
   The actual maximum duration of the risk window is three times
   <varname>wal_writer_delay</varname> because the WAL writer is
   designed to favor writing whole pages at a time during busy periods.
  </para>

  <caution>
   <para>
    An immediate-mode shutdown is equivalent to a server crash, and will
    therefore cause loss of any unflushed asynchronous commits.
   </para>
  </caution>

  <para>
   Asynchronous commit provides behavior different from setting
   <xref linkend="guc-fsync"/> = off.
   <varname>fsync</varname> is a server-wide
   setting that will alter the behavior of all transactions.  It disables
   all logic within <productname>PostgreSQL</productname> that attempts to synchronize
   writes to different portions of the database, and therefore a system
   crash (that is, a hardware or operating system crash, not a failure of
   <productname>PostgreSQL</productname> itself) could result in arbitrarily bad
   corruption of the database state.  In many scenarios, asynchronous
   commit provides most of the performance improvement that could be
   obtained by turning off <varname>fsync</varname>, but without the risk
   of data corruption.
  </para>

  <para>
   <xref linkend="guc-commit-delay"/> also sounds very similar to
   asynchronous commit, but it is actually a synchronous commit method
   (in fact, <varname>commit_delay</varname> is ignored during an
   asynchronous commit).  <varname>commit_delay</varname> causes a delay
   just before a transaction flushes <acronym>WAL</acronym> to disk, in
   the hope that a single flush executed by one such transaction can also
   serve other transactions committing at about the same time.  The
   setting can be thought of as a way of increasing the time window in
   which transactions can join a group about to participate in a single
   flush, to amortize the cost of the flush among multiple transactions.
  </para>

 </sect1>

 <sect1 id="wal-configuration">
  <title><acronym>WAL</acronym> Configuration</title>

  <para>
   There are several <acronym>WAL</acronym>-related configuration parameters that
   affect database performance. This section explains their use.
   Consult <xref linkend="runtime-config"/> for general information about
   setting server configuration parameters.
  </para>

  <para>
   <firstterm>Checkpoints</firstterm><indexterm><primary>checkpoint</primary></indexterm>
   are points in the sequence of transactions at which it is guaranteed
   that the heap and index data files have been updated with all
   information written before that checkpoint.  At checkpoint time, all
   dirty data pages are flushed to disk and a special checkpoint record is
   written to the log file.  (The change records were previously flushed
   to the <acronym>WAL</acronym> files.)
   In the event of a crash, the crash recovery procedure looks at the latest
   checkpoint record to determine the point in the log (known as the redo
   record) from which it should start the REDO operation.  Any changes made to
   data files before that point are guaranteed to be already on disk.
   Hence, after a checkpoint, log segments preceding the one containing
   the redo record are no longer needed and can be recycled or removed. (When
   <acronym>WAL</acronym> archiving is being done, the log segments must be
   archived before being recycled or removed.)
  </para>

  <para>
   The checkpoint requirement of flushing all dirty data pages to disk
   can cause a significant I/O load.  For this reason, checkpoint
   activity is throttled so that I/O begins at checkpoint start and completes
   before the next checkpoint is due to start; this minimizes performance
   degradation during checkpoints.
  </para>

  <para>
   The server's checkpointer process automatically performs
   a checkpoint every so often.  A checkpoint is begun every <xref
   linkend="guc-checkpoint-timeout"/> seconds, or if
   <xref linkend="guc-max-wal-size"/> is about to be exceeded,
   whichever comes first.
   The default settings are 5 minutes and 1 GB, respectively.
   If no WAL has been written since the previous checkpoint, new checkpoints
   will be skipped even if <varname>checkpoint_timeout</varname> has passed.
   (If WAL archiving is being used and you want to put a lower limit on how
   often files are archived in order to bound potential data loss, you should
   adjust the <xref linkend="guc-archive-timeout"/> parameter rather than the
   checkpoint parameters.)
   It is also possible to force a checkpoint by using the SQL
   command <command>CHECKPOINT</command>.
  </para>

  <para>
   Reducing <varname>checkpoint_timeout</varname> and/or
   <varname>max_wal_size</varname> causes checkpoints to occur
   more often. This allows faster after-crash recovery, since less work
   will need to be redone. However, one must balance this against the
   increased cost of flushing dirty data pages more often. If
   <xref linkend="guc-full-page-writes"/> is set (as is the default), there is
   another factor to consider. To ensure data page consistency,
   the first modification of a data page after each checkpoint results in
   logging the entire page content. In that case,
   a smaller checkpoint interval increases the volume of output to the WAL log,
   partially negating the goal of using a smaller interval,
   and in any case causing more disk I/O.
  </para>

  <para>
   Checkpoints are fairly expensive, first because they require writing
   out all currently dirty buffers, and second because they result in
   extra subsequent WAL traffic as discussed above.  It is therefore
   wise to set the checkpointing parameters high enough so that checkpoints
   don't happen too often.  As a simple sanity check on your checkpointing
   parameters, you can set the <xref linkend="guc-checkpoint-warning"/>
   parameter.  If checkpoints happen closer together than
   <varname>checkpoint_warning</varname> seconds,
   a message will be output to the server log recommending increasing
   <varname>max_wal_size</varname>.  Occasional appearance of such
   a message is not cause for alarm, but if it appears often then the
   checkpoint control parameters should be increased. Bulk operations such
   as large <command>COPY</command> transfers might cause a number of such warnings
   to appear if you have not set <varname>max_wal_size</varname> high
   enough.
  </para>

  <para>
   To avoid flooding the I/O system with a burst of page writes,
   writing dirty buffers during a checkpoint is spread over a period of time.
   That period is controlled by
   <xref linkend="guc-checkpoint-completion-target"/>, which is
   given as a fraction of the checkpoint interval (configured by using
   <varname>checkpoint_timeout</varname>).
   The I/O rate is adjusted so that the checkpoint finishes when the
   given fraction of
   <varname>checkpoint_timeout</varname> seconds have elapsed, or before
   <varname>max_wal_size</varname> is exceeded, whichever is sooner.
   With the default value of 0.9,
   <productname>PostgreSQL</productname> can be expected to complete each checkpoint
   a bit before the next scheduled checkpoint (at around 90% of the last checkpoint's
   duration).  This spreads out the I/O as much as possible so that the checkpoint
   I/O load is consistent throughout the checkpoint interval.  The disadvantage of
   this is that prolonging checkpoints affects recovery time, because more WAL
   segments will need to be kept around for possible use in recovery.  A user
   concerned about the amount of time required to recover might wish to reduce
   <varname>checkpoint_timeout</varname> so that checkpoints occur more frequently
   but still spread the I/O across the checkpoint interval.  Alternatively,
   <varname>checkpoint_completion_target</varname> could be reduced, but this would
   result in times of more intense I/O (during the checkpoint) and times of less I/O
   (after the checkpoint completed but before the next scheduled checkpoint) and
   therefore is not recommended.
   Although <varname>checkpoint_completion_target</varname> could be set as high as
   1.0, it is typically recommended to set it to no higher than 0.9 (the default)
   since checkpoints include some other activities besides writing dirty buffers.
   A setting of 1.0 is quite likely to result in checkpoints not being
   completed on time, which would result in performance loss due to
   unexpected variation in the number of WAL segments needed.
  </para>

  <para>
   On Linux and POSIX platforms <xref linkend="guc-checkpoint-flush-after"/>
   allows to force the OS that pages written by the checkpoint should be
   flushed to disk after a configurable number of bytes.  Otherwise, these
   pages may be kept in the OS's page cache, inducing a stall when
   <literal>fsync</literal> is issued at the end of a checkpoint.  This setting will
   often help to reduce transaction latency, but it also can have an adverse
   effect on performance; particularly for workloads that are bigger than
   <xref linkend="guc-shared-buffers"/>, but smaller than the OS's page cache.
  </para>

  <para>
   The number of WAL segment files in <filename>pg_wal</filename> directory depends on
   <varname>min_wal_size</varname>, <varname>max_wal_size</varname> and
   the amount of WAL generated in previous checkpoint cycles. When old log
   segment files are no longer needed, they are removed or recycled (that is,
   renamed to become future segments in the numbered sequence). If, due to a
   short-term peak of log output rate, <varname>max_wal_size</varname> is
   exceeded, the unneeded segment files will be removed until the system
   gets back under this limit. Below that limit, the system recycles enough
   WAL files to cover the estimated need until the next checkpoint, and
   removes the rest. The estimate is based on a moving average of the number
   of WAL files used in previous checkpoint cycles. The moving average
   is increased immediately if the actual usage exceeds the estimate, so it
   accommodates peak usage rather than average usage to some extent.
   <varname>min_wal_size</varname> puts a minimum on the amount of WAL files
   recycled for future usage; that much WAL is always recycled for future use,
   even if the system is idle and the WAL usage estimate suggests that little
   WAL is needed.
  </para>

  <para>
   Independently of <varname>max_wal_size</varname>,
   the most recent <xref linkend="guc-wal-keep-size"/> megabytes of
   WAL files plus one additional WAL file are
   kept at all times. Also, if WAL archiving is used, old segments cannot be
   removed or recycled until they are archived. If WAL archiving cannot keep up
   with the pace that WAL is generated, or if <varname>archive_command</varname>
   fails repeatedly, old WAL files will accumulate in <filename>pg_wal</filename>
   until the situation is resolved. A slow or failed standby server that
   uses a replication slot will have the same effect (see
   <xref linkend="streaming-replication-slots"/>).
  </para>

  <para>
   In archive recovery or standby mode, the server periodically performs
   <firstterm>restartpoints</firstterm>,<indexterm><primary>restartpoint</primary></indexterm>
   which are similar to checkpoints in normal operation: the server forces
   all its state to disk, updates the <filename>pg_control</filename> file to
   indicate that the already-processed WAL data need not be scanned again,
   and then recycles any old log segment files in the <filename>pg_wal</filename>
   directory.
   Restartpoints can't be performed more frequently than checkpoints on the
   primary because restartpoints can only be performed at checkpoint records.
   A restartpoint is triggered when a checkpoint record is reached if at
   least <varname>checkpoint_timeout</varname> seconds have passed since the last
   restartpoint, or if WAL size is about to exceed
   <varname>max_wal_size</varname>. However, because of limitations on when a
   restartpoint can be performed, <varname>max_wal_size</varname> is often exceeded
   during recovery, by up to one checkpoint cycle's worth of WAL.
   (<varname>max_wal_size</varname> is never a hard limit anyway, so you should
   always leave plenty of headroom to avoid running out of disk space.)
  </para>

  <para>
   There are two commonly used internal <acronym>WAL</acronym> functions:
   <function>XLogInsertRecord</function> and <function>XLogFlush</function>.
   <function>XLogInsertRecord</function> is used to place a new record into
   the <acronym>WAL</acronym> buffers in shared memory. If there is no
   space for the new record, <function>XLogInsertRecord</function> will have
   to write (move to kernel cache) a few filled <acronym>WAL</acronym>
   buffers. This is undesirable because <function>XLogInsertRecord</function>
   is used on every database low level modification (for example, row
   insertion) at a time when an exclusive lock is held on affected
   data pages, so the operation needs to be as fast as possible.  What
   is worse, writing <acronym>WAL</acronym> buffers might also force the
   creation of a new log segment, which takes even more
   time. Normally, <acronym>WAL</acronym> buffers should be written
   and flushed by an <function>XLogFlush</function> request, which is
   made, for the most part, at transaction commit time to ensure that
   transaction records are flushed to permanent storage. On systems
   with high log output, <function>XLogFlush</function> requests might
   not occur often enough to prevent <function>XLogInsertRecord</function>
   from having to do writes.  On such systems
   one should increase the number of <acronym>WAL</acronym> buffers by
   modifying the <xref linkend="guc-wal-buffers"/> parameter.  When
   <xref linkend="guc-full-page-writes"/> is set and the system is very busy,
   setting <varname>wal_buffers</varname> higher will help smooth response times
   during the period immediately following each checkpoint.
  </para>

  <para>
   The <xref linkend="guc-commit-delay"/> parameter defines for how many
   microseconds a group commit leader process will sleep after acquiring a
   lock within <function>XLogFlush</function>, while group commit
   followers queue up behind the leader.  This delay allows other server
   processes to add their commit records to the WAL buffers so that all of
   them will be flushed by the leader's eventual sync operation.  No sleep
   will occur if <xref linkend="guc-fsync"/> is not enabled, or if fewer
   than <xref linkend="guc-commit-siblings"/> other sessions are currently
   in active transactions; this avoids sleeping when it's unlikely that
   any other session will commit soon.  Note that on some platforms, the
   resolution of a sleep request is ten milliseconds, so that any nonzero
   <varname>commit_delay</varname> setting between 1 and 10000
   microseconds would have the same effect.  Note also that on some
   platforms, sleep operations may take slightly longer than requested by
   the parameter.
  </para>

  <para>
   Since the purpose of <varname>commit_delay</varname> is to allow the
   cost of each flush operation to be amortized across concurrently
   committing transactions (potentially at the expense of transaction
   latency), it is necessary to quantify that cost before the setting can
   be chosen intelligently.  The higher that cost is, the more effective
   <varname>commit_delay</varname> is expected to be in increasing
   transaction throughput, up to a point.  The <xref
   linkend="pgtestfsync"/> program can be used to measure the average time
   in microseconds that a single WAL flush operation takes.  A value of
   half of the average time the program reports it takes to flush after a
   single 8kB write operation is often the most effective setting for
   <varname>commit_delay</varname>, so this value is recommended as the
   starting point to use when optimizing for a particular workload.  While
   tuning <varname>commit_delay</varname> is particularly useful when the
   WAL log is stored on high-latency rotating disks, benefits can be
   significant even on storage media with very fast sync times, such as
   solid-state drives or RAID arrays with a battery-backed write cache;
   but this should definitely be tested against a representative workload.
   Higher values of <varname>commit_siblings</varname> should be used in
   such cases, whereas smaller <varname>commit_siblings</varname> values
   are often helpful on higher latency media.  Note that it is quite
   possible that a setting of <varname>commit_delay</varname> that is too
   high can increase transaction latency by so much that total transaction
   throughput suffers.
  </para>

  <para>
   When <varname>commit_delay</varname> is set to zero (the default), it
   is still possible for a form of group commit to occur, but each group
   will consist only of sessions that reach the point where they need to
   flush their commit records during the window in which the previous
   flush operation (if any) is occurring.  At higher client counts a
   <quote>gangway effect</quote> tends to occur, so that the effects of group
   commit become significant even when <varname>commit_delay</varname> is
   zero, and thus explicitly setting <varname>commit_delay</varname> tends
   to help less.  Setting <varname>commit_delay</varname> can only help
   when (1) there are some concurrently committing transactions, and (2)
   throughput is limited to some degree by commit rate; but with high
   rotational latency this setting can be effective in increasing
   transaction throughput with as few as two clients (that is, a single
   committing client with one sibling transaction).
  </para>

  <para>
   The <xref linkend="guc-wal-sync-method"/> parameter determines how
   <productname>PostgreSQL</productname> will ask the kernel to force
   <acronym>WAL</acronym> updates out to disk.
   All the options should be the same in terms of reliability, with
   the exception of <literal>fsync_writethrough</literal>, which can sometimes
   force a flush of the disk cache even when other options do not do so.
   However, it's quite platform-specific which one will be the fastest.
   You can test the speeds of different options using the <xref
   linkend="pgtestfsync"/> program.
   Note that this parameter is irrelevant if <varname>fsync</varname>
   has been turned off.
  </para>

  <para>
   Enabling the <xref linkend="guc-wal-debug"/> configuration parameter
   (provided that <productname>PostgreSQL</productname> has been
   compiled with support for it) will result in each
   <function>XLogInsertRecord</function> and <function>XLogFlush</function>
   <acronym>WAL</acronym> call being logged to the server log. This
   option might be replaced by a more general mechanism in the future.
  </para>

  <para>
   There are two internal functions to write WAL data to disk:
   <function>XLogWrite</function> and <function>issue_xlog_fsync</function>.
   When <xref linkend="guc-track-wal-io-timing"/> is enabled, the total
   amounts of time <function>XLogWrite</function> writes and
   <function>issue_xlog_fsync</function> syncs WAL data to disk are counted as
   <literal>wal_write_time</literal> and <literal>wal_sync_time</literal> in
   <xref linkend="pg-stat-wal-view"/>, respectively.
   <function>XLogWrite</function> is normally called by
   <function>XLogInsertRecord</function> (when there is no space for the new
   record in WAL buffers), <function>XLogFlush</function> and the WAL writer,
   to write WAL buffers to disk and call <function>issue_xlog_fsync</function>.
   <function>issue_xlog_fsync</function> is normally called by
   <function>XLogWrite</function> to sync WAL files to disk.
   If <varname>wal_sync_method</varname> is either
   <literal>open_datasync</literal> or <literal>open_sync</literal>,
   a write operation in <function>XLogWrite</function> guarantees to sync written
   WAL data to disk and <function>issue_xlog_fsync</function> does nothing.
   If <varname>wal_sync_method</varname> is either <literal>fdatasync</literal>,
   <literal>fsync</literal>, or <literal>fsync_writethrough</literal>,
   the write operation moves WAL buffers to kernel cache and
   <function>issue_xlog_fsync</function> syncs them to disk. Regardless
   of the setting of <varname>track_wal_io_timing</varname>, the number
   of times <function>XLogWrite</function> writes and
   <function>issue_xlog_fsync</function> syncs WAL data to disk are also
   counted as <literal>wal_write</literal> and <literal>wal_sync</literal>
   in <structname>pg_stat_wal</structname>, respectively.
  </para>
 </sect1>

 <sect1 id="wal-internals">
  <title>WAL Internals</title>

  <indexterm zone="wal-internals">
   <primary>LSN</primary>
  </indexterm>

  <para>
   <acronym>WAL</acronym> is automatically enabled; no action is
   required from the administrator except ensuring that the
   disk-space requirements for the <acronym>WAL</acronym> logs are met,
   and that any necessary tuning is done (see <xref
   linkend="wal-configuration"/>).
  </para>

  <para>
   <acronym>WAL</acronym> records are appended to the <acronym>WAL</acronym>
   logs as each new record is written. The insert position is described by
   a Log Sequence Number (<acronym>LSN</acronym>) that is a byte offset into
   the logs, increasing monotonically with each new record.
   <acronym>LSN</acronym> values are returned as the datatype
   <link linkend="datatype-pg-lsn"><type>pg_lsn</type></link>. Values can be
   compared to calculate the volume of <acronym>WAL</acronym> data that
   separates them, so they are used to measure the progress of replication
   and recovery.
  </para>

  <para>
   <acronym>WAL</acronym> logs are stored in the directory
   <filename>pg_wal</filename> under the data directory, as a set of
   segment files, normally each 16 MB in size (but the size can be changed
   by altering the <option>--wal-segsize</option> <application>initdb</application> option).  Each segment is
   divided into pages, normally 8 kB each (this size can be changed via the
   <option>--with-wal-blocksize</option> configure option).  The log record headers
   are described in <filename>access/xlogrecord.h</filename>; the record
   content is dependent on the type of event that is being logged.  Segment
   files are given ever-increasing numbers as names, starting at
   <filename>000000010000000000000001</filename>.  The numbers do not wrap,
   but it will take a very, very long time to exhaust the
   available stock of numbers.
  </para>

  <para>
   It is advantageous if the log is located on a different disk from the
   main database files.  This can be achieved by moving the
   <filename>pg_wal</filename> directory to another location (while the server
   is shut down, of course) and creating a symbolic link from the
   original location in the main data directory to the new location.
  </para>

  <para>
   The aim of <acronym>WAL</acronym> is to ensure that the log is
   written before database records are altered, but this can be subverted by
   disk drives<indexterm><primary>disk drive</primary></indexterm> that falsely report a
   successful write to the kernel,
   when in fact they have only cached the data and not yet stored it
   on the disk.  A power failure in such a situation might lead to
   irrecoverable data corruption.  Administrators should try to ensure
   that disks holding <productname>PostgreSQL</productname>'s
   <acronym>WAL</acronym> log files do not make such false reports.
   (See <xref linkend="wal-reliability"/>.)
  </para>

  <para>
   After a checkpoint has been made and the log flushed, the
   checkpoint's position is saved in the file
   <filename>pg_control</filename>. Therefore, at the start of recovery,
   the server first reads <filename>pg_control</filename> and
   then the checkpoint record; then it performs the REDO operation by
   scanning forward from the log location indicated in the checkpoint
   record.  Because the entire content of data pages is saved in the
   log on the first page modification after a checkpoint (assuming
   <xref linkend="guc-full-page-writes"/> is not disabled), all pages
   changed since the checkpoint will be restored to a consistent
   state.
  </para>

  <para>
   To deal with the case where <filename>pg_control</filename> is
   corrupt, we should support the possibility of scanning existing log
   segments in reverse order &mdash; newest to oldest &mdash; in order to find the
   latest checkpoint.  This has not been implemented yet.
   <filename>pg_control</filename> is small enough (less than one disk page)
   that it is not subject to partial-write problems, and as of this writing
   there have been no reports of database failures due solely to the inability
   to read <filename>pg_control</filename> itself.  So while it is
   theoretically a weak spot, <filename>pg_control</filename> does not
   seem to be a problem in practice.
  </para>
 </sect1>
</chapter>