wal.sgml 16.3 KB
Newer Older
Bruce Momjian's avatar
Bruce Momjian committed
1
<!-- $Header: /cvsroot/pgsql/doc/src/sgml/wal.sgml,v 1.19 2002/09/06 20:08:06 momjian Exp $ -->
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

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

 <note>
  <title>Author</title>
  <para>
   Vadim Mikheev and Oliver Elphick
  </para>
 </note>

 <sect1 id="wal-general">
  <title>General Description</Title>

  <para>
   <firstterm>Write Ahead Logging</firstterm> (<acronym>WAL</acronym>)
   is a standard approach to transaction logging.  Its detailed
   description may be found in most (if not all) books about
   transaction processing. Briefly, <acronym>WAL</acronym>'s central
21
   concept is that changes to data files (where tables and indexes
22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
   reside) must be written only after those changes have been logged -
   that is, when log records have been flushed to permanent
   storage. When 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 will first be redone from the log records (this is
   roll-forward recovery, also known as REDO) and then changes made by
   uncommitted transactions will be removed from the data pages
   (roll-backward recovery - UNDO).
  </para>

  <sect2 id="wal-benefits-now">
   <title>Immediate Benefits of <acronym>WAL</acronym></title>

   <para>
    The first obvious benefit of using <acronym>WAL</acronym> is a
    significantly reduced number of disk writes, since only the log
    file needs to be flushed to disk at the time of transaction
41
    commit; in multiuser environments, commits of many transactions
42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
    may be accomplished with a single <function>fsync()</function> of
    the log file. Furthermore, 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.
   </para>

   <para>
    The next benefit is consistency of the data pages. The truth is
    that, before <acronym>WAL</acronym>,
    <productname>PostgreSQL</productname> was never able to guarantee
    consistency in the case of a crash.  Before
    <acronym>WAL</acronym>, any crash during writing could result in:

    <orderedlist>
     <listitem>
Peter Eisentraut's avatar
Peter Eisentraut committed
57
      <simpara>index tuples pointing to nonexistent table rows</simpara>
58 59 60 61 62 63 64 65 66 67 68 69
     </listitem>

     <listitem>
      <simpara>index tuples lost in split operations</simpara>
     </listitem>

     <listitem>
      <simpara>totally corrupted table or index page content, because
      of partially written data pages</simpara>
     </listitem>
    </orderedlist>

70
    Problems with indexes (problems 1 and 2) could possibly have been
71 72
    fixed by additional <function>fsync()</function> calls, but it is
    not obvious how to handle the last case without
73 74
    <acronym>WAL</acronym>; <acronym>WAL</acronym> saves the entire data
    page content in the log if that is required to ensure page
75 76 77 78 79 80 81 82
    consistency for after-crash recovery.
   </para>
  </sect2>
  
  <sect2 id="wal-benefits-later">
   <title>Future Benefits</title>

   <para>
83
    UNDO operation is not implemented. This means that changes
84
    made by aborted transactions will still occupy disk space and that
85
    we still need a permanent <filename>pg_clog</filename> file to hold
86
    the status of transactions, since we are not able to re-use
87
    transaction identifiers. Once UNDO is implemented,
88
    <filename>pg_clog</filename> will no longer be required to be
89
    permanent; it will be possible to remove
90 91
    <filename>pg_clog</filename> at shutdown. (However, the urgency of
    this concern has decreased greatly with the adoption of a segmented
92 93 94
    storage method for <filename>pg_clog</filename> --- it is no longer
    necessary to keep old <filename>pg_clog</filename> entries around
    forever.)
95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120
   </para>

   <para>
    With UNDO, it will also be possible to implement
    <firstterm>savepoints</firstterm> to allow partial rollback of
    invalid transaction operations (parser errors caused by mistyping
    commands, insertion of duplicate primary/unique keys and so on)
    with the ability to continue or commit valid operations made by
    the transaction before the error.  At present, any error will
    invalidate the whole transaction and require a transaction abort.
   </para>

   <para>
    <acronym>WAL</acronym> offers the opportunity for a new method for
    database on-line backup and restore (<acronym>BAR</acronym>).  To
    use this method, one would have to make periodic saves of data
    files to another disk, a tape or another host and also archive the
    <acronym>WAL</acronym> log files.  The database file copy and the
    archived log files could be used to restore just as if one were
    restoring after a crash. Each time a new database file copy was
    made the old log files could be removed.  Implementing this
    facility will require the logging of data file and index creation
    and deletion; it will also require development of a method for
    copying the data files (operating system copy commands are not
    suitable).
   </para>
121 122

   <para>
123 124 125 126 127 128 129 130 131
    A difficulty standing in the way of realizing these benefits is that
    they require saving <acronym>WAL</acronym> entries for considerable
    periods of time (eg, as long as the longest possible transaction if
    transaction UNDO is wanted). The present <acronym>WAL</acronym>
    format is extremely bulky since it includes many disk page
    snapshots. This is not a serious concern at present, since the
    entries only need to be kept for one or two checkpoint intervals;
    but to achieve these future benefits some sort of compressed
    <acronym>WAL</acronym> format will be needed.
132
   </para>
133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
  </sect2>
 </sect1>

 <sect1 id="wal-implementation">
  <title>Implementation</title>

  <para>
   <acronym>WAL</acronym> is automatically enabled from release 7.1
   onwards. No action is required from the administrator with the
   exception of ensuring that the additional disk-space requirements
   of the <acronym>WAL</acronym> logs are met, and that any necessary
   tuning is done (see <xref linkend="wal-configuration">).
  </para>

  <para>
   <acronym>WAL</acronym> logs are stored in the directory
   <Filename><replaceable>$PGDATA</replaceable>/pg_xlog</Filename>, as
Peter Eisentraut's avatar
Peter Eisentraut committed
150 151
   a set of segment files, each 16 MB in size.  Each segment is
   divided into 8 kB pages. The log record headers are described in
152 153
   <filename>access/xlog.h</filename>; record content is dependent on
   the type of event that is being logged.  Segment files are given
154
   ever-increasing numbers as names, starting at
155 156 157 158 159 160 161 162
   <filename>0000000000000000</filename>.  The numbers do not wrap, at
   present, but it should take a very long time to exhaust the
   available stock of numbers.
  </para>

  <para>
   The <acronym>WAL</acronym> buffers and control structure are in
   shared memory, and are handled by the backends; they are protected
163
   by lightweight locks.  The demand on shared memory is dependent on the
164
   number of buffers.  The default size of the <acronym>WAL</acronym>
Peter Eisentraut's avatar
Peter Eisentraut committed
165
   buffers is 8 buffers of 8 kB each, or 64 kB total.
166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182
  </para>

  <para>
   It is of advantage if the log is located on another disk than the
   main database files.  This may be achieved by moving the directory,
   <filename>pg_xlog</filename>, to another location (while the
   postmaster is shut down, of course) and creating a symbolic link
   from the original location in <replaceable>$PGDATA</replaceable> to
   the new location.
  </para>

  <para>
   The aim of <acronym>WAL</acronym>, to ensure that the log is
   written before database records are altered, may be subverted by
   disk drives 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 may still lead to
183 184
   irrecoverable data corruption.  Administrators should try to ensure
   that disks holding <productname>PostgreSQL</productname>'s
185 186 187 188 189 190 191 192 193 194 195
   log files do not make such false reports.
  </para>

  <sect2 id="wal-recovery">
   <title>Database Recovery with <acronym>WAL</acronym></title>

   <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, when recovery is to be
    done, the backend first reads <filename>pg_control</filename> and
196 197 198 199 200 201
    then the checkpoint record; then it performs the REDO operation by
    scanning forward from the log position 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, all pages changed since
    the checkpoint will be restored to a consistent state.
202 203 204 205 206 207 208 209
   </para>

   <para>
    Using <filename>pg_control</filename> to get the checkpoint
    position speeds up the recovery process, but to handle possible
    corruption of <filename>pg_control</filename>, we should actually
    implement the reading of existing log segments in reverse order --
    newest to oldest -- in order to find the last checkpoint.  This has
Peter Eisentraut's avatar
Peter Eisentraut committed
210
    not been implemented, yet.
211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234
   </para>
  </sect2>
 </sect1>

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

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

  <para>
   There are two commonly used <acronym>WAL</acronym> functions:
   <function>LogInsert</function> and <function>LogFlush</function>.
   <function>LogInsert</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>LogInsert</function> will have
   to write (move to kernel cache) a few filled <acronym>WAL</acronym>
   buffers. This is undesirable because <function>LogInsert</function>
   is used on every database low level modification (for example,
   tuple insertion) at a time when an exclusive lock is held on
235 236 237
   affected data pages, so the operation needs to be as fast as
   possible.  What is worse, writing <acronym>WAL</acronym> buffers may
   also force the creation of a new log segment, which takes even more
238 239 240 241 242 243 244 245
   time. Normally, <acronym>WAL</acronym> buffers should be written
   and flushed by a <function>LogFlush</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>LogFlush</function> requests may
   not occur often enough to prevent <acronym>WAL</acronym> buffers
   being written by <function>LogInsert</function>. On such systems
   one should increase the number of <acronym>WAL</acronym> buffers by
246 247 248 249
   modifying the <filename>postgresql.conf</filename> <varname>
   WAL_BUFFERS</varname> parameter. The default number of <acronym>
   WAL</acronym> buffers is 8.  Increasing this value will 
   correspondingly increase shared memory usage.
250 251 252 253 254 255 256 257 258 259 260 261
  </para>

  <para>
   <firstterm>Checkpoints</firstterm> 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
   checkpoint time, all dirty data pages are flushed to disk and a
   special checkpoint record is written to the log file. As result, in
   the event of a crash, the recoverer knows from what record in the
   log (known as the redo record) it should start the REDO operation,
   since any changes made to data files before that record are already
   on disk. After a checkpoint has been made, any log segments written
262 263 264 265
   before the undo records are no longer needed and can be recycled or
   removed. (When <acronym>WAL</acronym>-based <acronym>BAR</acronym> is
   implemented, the log segments would be archived before being recycled
   or removed.)
266 267 268
  </para>

  <para>
269 270 271 272 273 274 275 276 277 278
   The checkpoint maker is also able to create a few log segments for
   future use, so as to avoid the need for
   <function>LogInsert</function> or <function>LogFlush</function> to
   spend time in creating them.  (If that happens, the entire database
   system will be delayed by the creation operation, so it's better if
   the files can be created in the checkpoint maker, which is not on
   anyone's critical path.)
   By default a new 16MB segment file is created only if more than 75% of
   the current segment has been used.  This is inadequate if the system
   generates more than 4MB of log output between checkpoints.
279 280 281
  </para>

  <para>
Tom Lane's avatar
Tom Lane committed
282 283 284 285 286 287
   The postmaster spawns a special backend process every so often
   to create the next checkpoint.  A checkpoint is created every
   <varname>CHECKPOINT_SEGMENTS</varname> log segments, or every
   <varname>CHECKPOINT_TIMEOUT</varname> seconds, whichever comes first.
   The default settings are 3 segments and 300 seconds respectively.
   It is also possible to force a checkpoint by using the SQL command
288 289 290
   <command>CHECKPOINT</command>.
  </para>

291 292
  <para>
   Reducing <varname>CHECKPOINT_SEGMENTS</varname> and/or
293 294 295 296 297 298 299 300 301
   <varname>CHECKPOINT_TIMEOUT</varname> causes checkpoints to be done
   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. In addition,
   to ensure data page consistency, the first modification of a data
   page after each checkpoint results in logging the entire page
   content. Thus a smaller checkpoint interval increases the volume of
   output to the log, partially negating the goal of using a smaller
   interval, and in any case causing more disk I/O.
302 303 304
  </para>

  <para>
305 306 307 308
   There will be at least one 16MB segment file, and will normally 
   not be more than <varname>CHECKPOINT_SEGMENTS</varname>)
   + 1 files.  You can use this to estimate space requirements for 
   WAL.  Ordinarily, when old log segment files are no longer needed, 
Bruce Momjian's avatar
Bruce Momjian committed
309 310
   they are recycled (renamed to become the next sequential future 
   segments). If, due to a short-term peak of log output rate, there 
Bruce Momjian's avatar
Bruce Momjian committed
311
   are more than <varname>CHECKPOINT_SEGMENTS</varname> + 1 segment files, 
Bruce Momjian's avatar
Bruce Momjian committed
312 313
   the unneeded segment files will be deleted instead of recycled until the
   system gets back under this limit.
314 315
  </para>

316
  <para>
317 318 319
   The <varname>COMMIT_DELAY</varname> parameter defines for how many
   microseconds the backend will sleep after writing a commit
   record to the log with <function>LogInsert</function> but before
320 321
   performing a <function>LogFlush</function>. This delay allows other
   backends to add their commit records to the log so as to have all
322
   of them flushed with a single log sync. No sleep will occur if <varname>fsync</varname>
323 324 325 326 327 328
   is not enabled or if fewer than <varname>COMMIT_SIBLINGS</varname>
   other backends are not currently in active transactions; this avoids
   sleeping when it's unlikely that any other backend will commit soon.
   Note that on most platforms, the resolution of a sleep request is
   ten milliseconds, so that any nonzero <varname>COMMIT_DELAY</varname>
   setting between 1 and 10000 microseconds will have the same effect.
329 330
   Good values for these parameters are not yet clear; experimentation
   is encouraged.
331
  </para>
332 333 334

  <para>
   The <varname>WAL_SYNC_METHOD</varname> parameter determines how
335 336
   <productname>PostgreSQL</productname> will ask the kernel to force
    WAL updates out to disk. 
337 338 339 340 341 342 343
   All the options should be the same as far as reliability goes,
   but it's quite platform-specific which one will be the fastest.
   Note that this parameter is irrelevant if <varname>FSYNC</varname>
   has been turned off.
  </para>

  <para>
Peter Eisentraut's avatar
Peter Eisentraut committed
344
   Setting the <varname>WAL_DEBUG</varname> parameter to any nonzero
345 346 347
   value will result in each <function>LogInsert</function> and
   <function>LogFlush</function> <acronym>WAL</acronym> call being
   logged to standard error.  At present, it makes no difference what
Peter Eisentraut's avatar
Peter Eisentraut committed
348
   the nonzero value is.  This option may be replaced by a more
349 350
   general mechanism in the future.
  </para>
351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369
 </sect1>
</chapter>

<!-- Keep this comment at the end of the file
Local variables:
mode:sgml
sgml-omittag:nil
sgml-shorttag:t
sgml-minimize-attributes:nil
sgml-always-quote-attributes:t
sgml-indent-step:1
sgml-indent-data:t
sgml-parent-document:nil
sgml-default-dtd-file:"./reference.ced"
sgml-exposed-tags:nil
sgml-local-catalogs:("/usr/lib/sgml/catalog")
sgml-local-ecat-files:nil
End:
-->