path: root/lib/kernel/doc/src/disk_log.xml

<?xml version="1.0" encoding="utf-8" ?>
<!DOCTYPE erlref SYSTEM "erlref.dtd">

<erlref>
  <header>
    <copyright>
      <year>1997</year>
      <year>2013</year>
      <holder>Ericsson AB, All Rights Reserved</holder>
    </copyright>
    <legalnotice>
  Licensed under the Apache License, Version 2.0 (the "License");
  you may not use this file except in compliance with the License.
  You may obtain a copy of the License at
 
      http://www.apache.org/licenses/LICENSE-2.0

  Unless required by applicable law or agreed to in writing, software
  distributed under the License is distributed on an "AS IS" BASIS,
  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  See the License for the specific language governing permissions and
  limitations under the License.

  The Initial Developer of the Original Code is Ericsson AB.
    </legalnotice>

    <title>disk_log</title>
    <prepared>Claes Wikstr&ouml;m</prepared>
    <responsible>Claes Wikstr&ouml;m</responsible>
    <docno></docno>
    <approved>nobody</approved>
    <checked>no</checked>
    <date>1999-10-10</date>
    <rev>D</rev>
    <file>disk_log.sgml</file>
  </header>
  <module>disk_log</module>
  <modulesummary>A disk based term logging facility</modulesummary>
  <description>
    <p><c>disk_log</c> is a disk based term logger which makes
      it possible to efficiently log items on files. 
      Two types of logs are supported,
      <em>halt logs</em> and <em>wrap logs</em>. A halt log
      appends items to a single file, the size of which may or may
      not be limited by the disk log module, whereas a wrap log utilizes
      a sequence of wrap log files of limited size. As a wrap log file
      has been filled up, further items are logged onto to the next 
      file in the sequence, starting all over with the first file when 
      the last file has been filled up. For the sake of efficiency, 
      items are always written to files as binaries.
      </p>
    <p>Two formats of the log files are supported, the <em>internal format</em> and the <em>external format</em>. The internal
      format supports automatic repair of log files that have not been
      properly closed, and makes it possible to efficiently read
      logged items in <em>chunks</em> using a set of functions defined
      in this module. In fact, this is the only way to read internally
      formatted logs. The external format leaves it up to the user to
      read the logged deep byte lists. The disk log module cannot
      repair externally formatted logs. An item logged to an
      internally formatted log must not occupy more than 4 GB of disk
      space (the size must fit in 4 bytes).
      </p>
    <p>For each open disk log there is one process that handles requests 
      made to the disk log; the disk log process is created when <c>open/1</c>
      is called, provided there exists no process handling the disk log.
      A process that opens a disk log can either be an <em>owner</em> 
      or an anonymous <em>user</em> of the disk log. Each owner is 
      linked to the disk log 
      process, and the disk log is closed by the owner should the 
      owner terminate. Owners can subscribe to <em>notifications</em>, 
      messages of the form <c>{disk_log, Node, Log, Info}</c> that are sent
      from the disk log process when certain events occur, see 
      the commands below and in particular the <c>open/1</c> option 
      <seealso marker="#notify">notify</seealso>. 
      There can be several owners of a log, but a process cannot own a 
      log more than once. One and the same process may, however,
      open the log 
      as a user more than once. For a disk log process to properly close 
      its file and terminate, it must be closed by its owners and once by 
      some non-owner process for each time the log was used anonymously;
      the users are counted, and there must not be any users left when the
      disk log process terminates.
      </p>
    <p>Items can be logged <em>synchronously</em> by using the functions
      <c>log/2</c>, <c>blog/2</c>, <c>log_terms/2</c> and 
      <c>blog_terms/2</c>. For each of these functions, the caller is put
      on hold until the items have been logged (but not necessarily 
      written, use <c>sync/1</c> to ensure that). By adding an <c>a</c>
      to each of the mentioned function names we get functions that log
      items <em>asynchronously</em>. Asynchronous functions do not wait for
      the disk log process to actually write the items to the file, but
      return the control to the caller more or less immediately.
      </p>
    <p>When using the internal format for logs, the functions
      <c>log/2</c>, <c>log_terms/2</c>, <c>alog/2</c>, and
      <c>alog_terms/2</c> should be used. These functions log one or
      more Erlang terms. By prefixing each of the functions with
      a <c>b</c> (for "binary") we get the corresponding <c>blog</c> 
      functions for the external format. These functions log one or 
      more deep lists of bytes or, alternatively, binaries of deep lists 
      of bytes. 
      For example, to log the string <c>"hello"</c> in ASCII format, we
      can use <c>disk_log:blog(Log, "hello")</c>, or
      <c>disk_log:blog(Log, list_to_binary("hello"))</c>. The two
      alternatives are equally efficient. The <c>blog</c> functions
      can be used for internally formatted logs as well, but in
      this case they must be called with binaries constructed with
      calls to <c>term_to_binary/1</c>. There is no check to ensure
      this, it is entirely the responsibility of the caller. If these
      functions are called with binaries that do not correspond to
      Erlang terms, the <c>chunk/2,3</c> and automatic repair
      functions will fail. The corresponding terms (not the binaries)
      will be returned when <c>chunk/2,3</c> is called.
      </p>
    <p>A collection of open disk logs with the same name running on
      different nodes is said to be a <em>a distributed disk log</em>
      if requests made to any one of the logs are automatically made to
      the other logs as well. The members of such a collection will be
      called individual distributed disk logs, or just distributed
      disk logs if there is no risk of confusion. There is no order
      between the members of such a collection. For instance, logged
      terms are not necessarily written onto the node where the
      request was made before written onto the other nodes. One could
      note here that there are a few functions that do not make
      requests to all members of distributed disk logs, namely
      <c>info</c>, <c>chunk</c>, <c>bchunk</c>, <c>chunk_step</c> and
      <c>lclose</c>. An open disk log that is not a distributed disk
      log is said to be a <em>local disk log</em>. A local disk log is
      accessible only from the node where the disk log process runs,
      whereas a distributed disk log is accessible from all nodes in
      the Erlang system, with exception for those nodes where a local
      disk log with the same name as the distributed disk log exists.
      All processes on nodes that have access to a local or
      distributed disk log can log items or otherwise change, inspect
      or close the log.
      </p>
    <p>It is not guaranteed that all log files of a distributed disk log 
      contain the same log items; there is no attempt made to synchronize
      the contents of the files. However, as long as at least one of 
      the involved nodes is alive at each time, all items will be logged. 
      When logging items to a distributed log, or otherwise trying to
      change the log, the replies from individual logs are
      ignored. If all nodes are down, the disk log functions
      reply with a <c>nonode</c> error.
      </p>
    <note>
      <p>In some applications it may not be acceptable that
        replies from individual logs are ignored. An alternative in such
        situations is to use several local disk logs instead of one
        distributed disk log, and implement the distribution without use
        of the disk log module.</p>
    </note>
    <p>Errors are reported differently for asynchronous log attempts
      and other uses of the disk log module. When used synchronously
      the disk log module replies with an error message, but when called
      asynchronously, the disk log module does not know where to send 
      the error message. Instead owners subscribing to notifications will 
      receive an <c>error_status</c> message. 
      </p>
    <p>The disk log module itself does not report errors to the
      <c>error_logger</c> module; it is up to the caller to decide
      whether the error logger should be employed or not. The function
      <c>format_error/1</c> can be used to produce readable messages
      from error replies. Information events are however sent to the
      error logger in two situations, namely when a log is repaired,
      or when a file is missing while reading chunks.
      </p>
    <p>The error message <c>no_such_log</c> means that the given
      disk log is not currently open. Nothing is said about 
      whether the disk log files exist or not.
      </p>
    <note>
      <p>If an attempt to reopen or truncate a log fails (see
        <c>reopen</c> and <c>truncate</c>) the disk log process
        immediately terminates. Before the process terminates links to
        to owners and blocking processes (see <c>block</c>) are removed.
        The effect is that the links work in one direction only; any
        process using a disk log has to check for the error message
        <c>no_such_log</c> if some other process might truncate or
        reopen the log simultaneously.</p>
    </note>
  </description>
  <datatypes>
    <datatype>
      <name name="log"/>
    </datatype>
    <datatype>
      <name name="dlog_size"/>
    </datatype>
    <datatype>
      <name name="dlog_format"/>
    </datatype>
    <datatype>
      <name name="dlog_head_opt"/>
    </datatype>
    <datatype>
      <name name="dlog_byte"/>
    </datatype>
    <datatype>
      <name name="dlog_mode"/>
    </datatype>
    <datatype>
      <name name="dlog_type"/>
    </datatype>
    <datatype>
      <name name="continuation"/>
      <desc><p>Chunk continuation returned by
         <c>chunk/2,3</c>, <c>bchunk/2,3</c>, or <c>chunk_step/3</c>.</p>
      </desc>
    </datatype>
    <datatype>
      <name name="bytes"/>
    </datatype>
    <datatype>
      <name name="invalid_header"/>
    </datatype>
    <datatype>
      <name name="file_error"/>
    </datatype>
  </datatypes>
  <funcs>
    <func>
      <name name="accessible_logs" arity="0"/>
      <fsummary>Return the accessible disk logs on the  current node.</fsummary>
      <desc>
        <p>The <c>accessible_logs/0</c> function returns
          the names of the disk logs accessible on the current node.
          The first list contains local disk logs, and the
          second list contains distributed disk logs.
          </p>
      </desc>
    </func>
    <func>
      <name name="alog" arity="2"/>
      <name name="balog" arity="2"/>
      <fsummary>Asynchronously log an item onto a disk log.</fsummary>
      <type variable="Log"/>
      <type variable="Term" name_i="1"/>
      <type variable="Bytes"/>
      <type name="notify_ret"/>
      <desc>
        <p>The <c>alog/2</c> and <c>balog/2</c> functions asynchronously
          append an item to a disk log. The function <c>alog/2</c> is 
          used for internally formatted logs, and the function <c>balog/2</c>
          for externally formatted logs. <c>balog/2</c> can be used
          for internally formatted logs as well provided the binary was
          constructed with a call to <c>term_to_binary/1</c>.
          </p>
        <p>The owners that subscribe to notifications will receive the 
          message <c>read_only</c>, <c>blocked_log</c> 
          or <c>format_external</c> in case the item cannot be written
          on the log, and possibly one of the messages <c>wrap</c>,
          <c>full</c> and <c>error_status</c> if an item was written 
          on the log. The message <c>error_status</c> is sent if there
          is something wrong with the header function or a file error
          occurred.
          </p>
      </desc>
    </func>
    <func>
      <name name="alog_terms" arity="2"/>
      <name name="balog_terms" arity="2"/>
      <fsummary>Asynchronously log several items onto a disk log.</fsummary>
      <type variable="Log"/>
      <type variable="TermList" name_i="1"/>
      <type variable="ByteList"/>
      <type name="notify_ret"/>
      <desc>
        <p>The <c>alog_terms/2</c> and <c>balog_terms/2</c> functions 
          asynchronously append a list of items to a disk log. 
          The function <c>alog_terms/2</c> is used for internally 
          formatted logs, and the function <c>balog_terms/2</c>
          for externally formatted logs. <c>balog_terms/2</c> can be used
          for internally formatted logs as well provided the binaries were
          constructed with calls to <c>term_to_binary/1</c>.
          </p>
        <p>The owners that subscribe to notifications will receive the 
          message <c>read_only</c>, <c>blocked_log</c> 
          or <c>format_external</c> in case the items cannot be written
          on the log, and possibly one or more of the messages <c>wrap</c>,
          <c>full</c> and <c>error_status</c> if items were written 
          on the log. The message <c>error_status</c> is sent if there
          is something wrong with the header function or a file error 
          occurred.
          </p>
      </desc>
    </func>
    <func>
      <name name="block" arity="1"/>
      <name name="block" arity="2"/>
      <fsummary>Block a disk log.</fsummary>
      <type name="block_error_rsn"/>
      <desc>
        <p>With a call to <c>block/1,2</c> a process can block a log.
          If the blocking process is not an owner of the log, a temporary
          link is created between the disk log process and the blocking
          process. The link is used to ensure that the disk log is 
          unblocked should the blocking process terminate without 
          first closing or unblocking the log.
          </p>
        <p>Any process can probe a blocked log with <c>info/1</c> or
          close it with <c>close/1</c>. The blocking process can also
          use the functions <c>chunk/2,3</c>, <c>bchunk/2,3</c>,
          <c>chunk_step/3</c>, and <c>unblock/1</c> without being
          affected by the block. Any other attempt than those hitherto
          mentioned to update or read a blocked log suspends the
          calling process until the log is unblocked or returns an
          error message <c>{blocked_log, <anno>Log</anno>}</c>, depending on
          whether the value of <c><anno>QueueLogRecords</anno></c> is <c>true</c>
          or <c>false</c>. The default value of <c><anno>QueueLogRecords</anno></c>
          is <c>true</c>, which is used by <c>block/1</c>.
          </p>
      </desc>
    </func>
    <func>
      <name name="change_header" arity="2"/>
      <fsummary>Change the head or head_func option for an owner of  a disk log.</fsummary>
      <desc>
        <p>The <c>change_header/2</c> function changes the value of 
          the <c>head</c> or <c>head_func</c> option of a disk log.</p>
      </desc>
    </func>
    <func>
      <name name="change_notify" arity="3"/>
      <fsummary>Change the notify option for an owner of a  disk log.</fsummary>
      <desc>
        <p>The <c>change_notify/3</c> function changes the value of the
          <c>notify</c> option for an owner of a disk log. </p>
      </desc>
    </func>
    <func>
      <name name="change_size" arity="2"/>
      <fsummary>Change the size of an open disk log.</fsummary>
      <desc>
        <p>The <c>change_size/2</c> function changes the size of an open log. 
          For a halt log it is always possible to increase the size, 
          but it is not possible to decrease the size to something less than 
          the current size of the file.
          </p>
        <p>For a wrap log it is always possible to increase both the
          size and number of files, as long as the number of files does not 
          exceed 65000. If the maximum number of files is decreased, the
          change will not be valid until the current file is full and the 
          log wraps to the next file.
          The redundant files will be removed next time the log wraps around, 
          i.e. starts to log to file number 1.
          </p>
        <p>As an example, assume that the old maximum number of files 
          is 10 and that the new maximum number of files is 6. If 
          the current file number is not greater than the new maximum number 
          of files, the files 7 to 10 will be removed when file number 6 
          is full and the log starts to write to file number 1 again. 
          Otherwise the files greater than the current 
          file will be removed when the current file is full (e.g. if  
          the current file is 8, the files 9 and 10); the files between 
          new maximum number of files and the current 
          file (i.e. files 7 and 8) will be removed next time file number 6
          is full.
          </p>
        <p>If the size of the files is decreased the change will immediately
          affect the current log. It will not of course change the 
          size of log files already full until next time they are used.
          </p>
        <p>If the log size is decreased for instance to save space, 
          the function <c>inc_wrap_file/1</c> can be used to force the log 
          to wrap.
          </p>
      </desc>
    </func>
    <func>
      <name name="chunk" arity="2"/>
      <name name="chunk" arity="3"/>
      <name name="bchunk" arity="2"/>
      <name name="bchunk" arity="3"/>
      <fsummary>Read a chunk of items written to a disk log.</fsummary>
      <type variable="Log"/>
      <type variable="Continuation"/>
      <type variable="N"/>
      <type name="chunk_ret"/>
      <type name="bchunk_ret"/>
      <type name="chunk_error_rsn"/>
      <desc>
        <p>The <c>chunk/2,3</c> and <c>bchunk/2,3</c> functions make
          it possible to efficiently read the terms which have been
          appended to an internally formatted log. It minimizes disk
          I/O by reading 64 kilobyte chunks from the file. The
          <c>bchunk/2,3</c> functions return the binaries read from
          the file; they do not call <c>binary_to_term</c>. Otherwise
          the work just like <c>chunk/2,3</c>.
          </p>
        <p>The first time <c>chunk</c> (or <c>bchunk</c>) is called,
          an initial continuation, the atom <c>start</c>, must be
          provided. If there is a disk log process running on the
          current node, terms are read from that log, otherwise an
          individual distributed log on some other node is chosen, if
          such a log exists.
          </p>
        <p>When <c>chunk/3</c> is called, <c><anno>N</anno></c> controls the
          maximum number of terms that are read from the log in each
          chunk. Default is <c>infinity</c>, which means that all the
          terms contained in the 64 kilobyte chunk are read. If less than
          <c><anno>N</anno></c> terms are returned, this does not necessarily mean
          that the end of the file has been reached.
          </p>
        <p>The <c>chunk</c> function returns a tuple
          <c>{<anno>Continuation2</anno>, <anno>Terms</anno>}</c>, where <c><anno>Terms</anno></c> is a list
          of terms found in the log. <c><anno>Continuation2</anno></c> is yet
          another continuation which must be passed on to any
          subsequent calls to <c>chunk</c>. With a series of calls to
          <c>chunk</c> it is possible to extract all terms from a log.
          </p>
        <p>The <c>chunk</c> function returns a tuple
          <c>{<anno>Continuation2</anno>, <anno>Terms</anno>, <anno>Badbytes</anno>}</c> if the log is opened
          in read-only mode and the read chunk is corrupt. <c><anno>Badbytes</anno></c>
          is the number of bytes in the file which were found not to be
          Erlang terms in the chunk. Note also that the log is not repaired.
          When trying to read chunks from a log opened in read-write mode,
          the tuple <c>{corrupt_log_file, <anno>FileName</anno>}</c> is returned if the
          read chunk is corrupt.
          </p>
        <p><c>chunk</c> returns <c>eof</c> when the end of the log is
          reached, or <c>{error, <anno>Reason</anno>}</c> if an error occurs. Should
          a wrap log file be missing, a message is output on the error log.
          </p>
        <p>When <c>chunk/2,3</c> is used with wrap logs, the returned
          continuation may or may not be valid in the next call to
          <c>chunk</c>. This is because the log may wrap and delete
          the file into which the continuation points. To make sure
          this does not happen, the log can be blocked during the
          search.
          </p>
      </desc>
    </func>
    <func>
      <name name="chunk_info" arity="1"/>
      <fsummary>Return information about a chunk continuation of  a disk log.</fsummary>
      <desc>
        <p>The <c>chunk_info/1</c> function returns the following pair
          describing the chunk continuation returned by
          <c>chunk/2,3</c>, <c>bchunk/2,3</c>, or <c>chunk_step/3</c>:
          </p>
        <list type="bulleted">
          <item>
            <p><c>{node, <anno>Node</anno>}</c>. Terms are read from
              the disk log running on <c><anno>Node</anno></c>.</p>
          </item>
        </list>
      </desc>
    </func>
    <func>
      <name name="chunk_step" arity="3"/>
      <fsummary>Step forward or backward among the wrap log files  of a disk log.</fsummary>
      <desc>
        <p>The function <c>chunk_step</c> can be used in conjunction
          with <c>chunk/2,3</c> and <c>bchunk/2,3</c> to search
          through an internally formatted wrap log. It takes as
          argument a continuation as returned by <c>chunk/2,3</c>,
          <c>bchunk/2,3</c>, or <c>chunk_step/3</c>, and steps forward
          (or backward) <c><anno>Step</anno></c> files in the wrap log. The
          continuation returned points to the first log item in the
          new current file.
          </p>
        <p>If the atom <c>start</c> is given as continuation, a disk log
          to read terms from is chosen. A local or distributed disk log
          on the current node is preferred to an
          individual distributed log on some other node.
          </p>
        <p>If the wrap log is not full because all files have not been
          used yet, <c>{error, end_of_log}</c> is returned if trying to
          step outside the log.
          </p>
      </desc>
    </func>
    <func>
      <name name="close" arity="1"/>
      <fsummary>Close a disk log.</fsummary>
      <type name="close_error_rsn"/>
      <desc>
        <p><marker id="close_1"></marker>The function <c>close/1</c> closes a
          local or distributed disk log properly. An internally
          formatted log must be closed before the Erlang system is
          stopped, otherwise the log is regarded as unclosed and the
          automatic repair procedure will be activated next time the
          log is opened.
          </p>
        <p>The disk log process in not terminated as long as there are
          owners or users of the log. It should be stressed that each
          and every owner must close the log, possibly by terminating,
          and that any other process - not only the processes that have 
          opened the log anonymously - can decrement the <c>users</c>
          counter by closing the log. 
          Attempts to close a log by a process that is 
          not an owner are simply ignored if there are no users.
          </p>
        <p>If the log is blocked by the closing process, the log is also
          unblocked.
          </p>
      </desc>
    </func>
    <func>
      <name name="format_error" arity="1"/>
      <fsummary>Return an English description of a disk log  error reply.</fsummary>
      <desc>
        <p>Given the error returned by any function in this module, 
          the function <c>format_error</c> returns a descriptive string
          of the error in English. For file errors, the function 
          <c>format_error/1</c> in the <c>file</c> module is called.</p>
      </desc>
    </func>
    <func>
      <name name="inc_wrap_file" arity="1"/>
      <fsummary>Change to the next wrap log file of a disk log.</fsummary>
      <type name="inc_wrap_error_rsn"/>
      <type name="invalid_header"/>
      <desc>
        <p>The <c>inc_wrap_file/1</c> function forces the internally formatted 
          disk log to start logging to the 
          next log file. It can be used, for instance, in conjunction with
          <c>change_size/2</c> to reduce the amount of disk space allocated
          by the disk log.
          </p>
        <p>The owners that subscribe to notifications will normally 
          receive a <c>wrap</c> message, but in case of 
          an error with a reason tag of <c>invalid_header</c> or 
          <c>file_error</c> an <c>error_status</c> message will be sent.</p>
      </desc>
    </func>
    <func>
      <name name="info" arity="1"/>
      <fsummary>Return information about a disk log.</fsummary>
      <type name="dlog_info"/>
      <desc>
        <p>The <c>info/1</c> function returns a list of <c>{Tag, Value}</c> 
          pairs describing the log. If there is a disk log process running
          on the current node, that log is used as source of information,
          otherwise an individual distributed log on
          some other node is chosen, if such a log exists.
          </p>
        <p>The following pairs are returned for all logs:
          </p>
        <list type="bulleted">
          <item>
            <p><c>{name, <anno>Log</anno>}</c>, where <c><anno>Log</anno></c> is the name of
              the log as given by the <c>open/1</c> option <c>name</c>.</p>
          </item>
          <item>
            <p><c>{file, <anno>File</anno>}</c>. For halt logs <c><anno>File</anno></c> is the
              filename, and for wrap logs <c><anno>File</anno></c> is the base name.</p>
          </item>
          <item>
            <p><c>{type, <anno>Type</anno>}</c>, where <c><anno>Type</anno></c> is the type of
              the log as given by the <c>open/1</c> option <c>type</c>.</p>
          </item>
          <item>
            <p><c>{format, <anno>Format</anno>}</c>, where <c><anno>Format</anno></c> is the format
              of the log as given by the <c>open/1</c> option <c>format</c>.</p>
          </item>
          <item>
            <p><c>{size, <anno>Size</anno>}</c>, where <c><anno>Size</anno></c> is the size
              of the log as given by the <c>open/1</c> option <c>size</c>,
              or the size set by <c>change_size/2</c>. The value set by 
              <c>change_size/2</c> is reflected immediately.</p>
          </item>
          <item>
            <p><c>{mode, <anno>Mode</anno>}</c>, where <c><anno>Mode</anno></c> is the mode
              of the log as given by the <c>open/1</c> option <c>mode</c>.</p>
          </item>
          <item>
            <p><c>{owners, [{pid(), <anno>Notify</anno>}]}</c> where <c><anno>Notify</anno></c>
              is the value set by the <c>open/1</c> option <c>notify</c>
              or the function <c>change_notify/3</c> for the owners of 
              the log.</p>
          </item>
          <item>
            <p><c>{users, <anno>Users</anno>}</c> where <c><anno>Users</anno></c> is the number
              of anonymous users of the log, see the <c>open/1</c> option
              <seealso marker="#linkto">linkto</seealso>.</p>
          </item>
          <item>
            <p><c>{status, <anno>Status</anno>}</c>, where <c><anno>Status</anno></c> is <c>ok</c>
              or <c>{blocked, <anno>QueueLogRecords</anno>}</c> as set by the functions
              <c>block/1,2</c> and <c>unblock/1</c>.</p>
          </item>
          <item>
            <p><c>{node, <anno>Node</anno>}</c>. The information returned by the
              current invocation of the <c>info/1</c> function has been
              gathered from the disk log process running on <c><anno>Node</anno></c>.</p>
          </item>
          <item>
            <p><c>{distributed, <anno>Dist</anno>}</c>. If the log is local on
              the current node, then <c><anno>Dist</anno></c> has the value <c>local</c>,
              otherwise all nodes where the log is distributed
              are returned as a list.</p>
          </item>
        </list>
        <p>The following pairs are returned for all logs opened in 
          <c>read_write</c> mode:
          </p>
        <list type="bulleted">
          <item>
            <p><c>{head, <anno>Head</anno>}</c>. Depending of the value of
              the <c>open/1</c> options <c>head</c> and <c>head_func</c>
              or set by the function <c>change_header/2</c>, the value
              of <c><anno>Head</anno></c> is <c>none</c> (default),
              <c>{head, H}</c> (<c>head</c> option) or <c>{M,F,A}</c> 
              (<c>head_func</c> option).</p>
          </item>
          <item>
            <p><c>{no_written_items, <anno>NoWrittenItems</anno>}</c>, where
              <c><anno>NoWrittenItems</anno></c> is the number of items
              written to the log since the disk log process was created.</p>
          </item>
        </list>
        <p>The following pair is returned for halt logs opened in 
          <c>read_write</c> mode:
          </p>
        <list type="bulleted">
          <item>
            <p><c>{full, <anno>Full</anno>}</c>, where <c><anno>Full</anno></c> is <c>true</c> or
              <c>false</c> depending on whether the halt log is full or not.</p>
          </item>
        </list>
        <p>The following pairs are returned for wrap logs opened in 
          <c>read_write</c> mode:
          </p>
        <list type="bulleted">
          <item>
            <p><c>{no_current_bytes, integer() >= 0}</c> is the number
              of bytes written to the current wrap log file.</p>
          </item>
          <item>
            <p><c>{no_current_items, integer() >= 0}</c> is the number
              of items written to the current wrap log file, header
              inclusive.</p>
          </item>
          <item>
            <p><c>{no_items, integer() >= 0}</c> is the total number
              of items in all wrap log files.</p>
          </item>
          <item>
            <p><c>{current_file, integer()}</c> is the ordinal for
              the current wrap log file in the range <c>1..MaxNoFiles</c>,
              where <c>MaxNoFiles</c> is given by the <c>open/1</c> option
              <c>size</c> or set by <c>change_size/2</c>.</p>
          </item>
          <item>
            <p><c>{no_overflows, {<anno>SinceLogWasOpened</anno>, <anno>SinceLastInfo</anno>}}</c>,
              where <c><anno>SinceLogWasOpened</anno></c> (<c><anno>SinceLastInfo</anno></c>) is
              the number of times a wrap log file has been filled up and a 
              new one opened or <c>inc_wrap_file/1</c> has been called since 
              the disk log was last opened (<c>info/1</c> 
              was last called). The first time <c>info/2</c> is called 
              after a log was (re)opened or truncated, the two values 
              are equal.</p>
          </item>
        </list>
        <p>Note that the <c>chunk/2,3</c>, <c>bchunk/2,3</c>, and
          <c>chunk_step/3</c> functions do not affect any value
          returned by <c>info/1</c>.
          </p>
      </desc>
    </func>
    <func>
      <name name="lclose" arity="1"/>
      <name name="lclose" arity="2"/>
      <fsummary>Close a disk log on one node.</fsummary>
      <type name="lclose_error_rsn"/>
      <desc>
        <p>The function <c>lclose/1</c> closes a local log or an
          individual distributed log on the current node.
          The function <c>lclose/2</c> closes an individual
          distributed log on the specified node if the node
          is not the current one.
          <c>lclose(<anno>Log</anno>)</c> is equivalent to
          <c>lclose(<anno>Log</anno>,&nbsp;node())</c>.
          See also <seealso marker="#close_1">close/1</seealso>.
          </p>
        <p>If there is no log with the given name
          on the specified node, <c>no_such_log</c> is returned. 
          </p>
      </desc>
    </func>
    <func>
      <name name="log" arity="2"/>
      <name name="blog" arity="2"/>
      <fsummary>Log an item onto a disk log.</fsummary>
      <type variable="Log"/>
      <type variable="Term" name_i="1"/>
      <type variable="Bytes"/>
      <type name="log_error_rsn"/>
      <desc>
        <p>The <c>log/2</c> and <c>blog/2</c> functions synchronously
          append a term to a disk log. They return <c>ok</c> or
          <c>{error, <anno>Reason</anno>}</c> when the term has been written to
          disk. If the log is distributed, <c>ok</c> is always
          returned, unless all nodes are down. Terms are written by
          means of the ordinary <c>write()</c> function of the
          operating system. Hence, there is no guarantee that the term
          has actually been written to the disk, it might linger in
          the operating system kernel for a while. To make sure the
          item is actually written to disk, the <c>sync/1</c> function
          must be called.
          </p>
        <p>The <c>log/2</c> function is used for internally formatted logs,
          and <c>blog/2</c> for externally formatted logs. 
          <c>blog/2</c> can be used
          for internally formatted logs as well provided the binary was
          constructed with a call to <c>term_to_binary/1</c>.
          </p>
        <p>The owners that subscribe to notifications will be notified 
          of an error with an <c>error_status</c> message if the error 
          reason tag is <c>invalid_header</c> or <c>file_error</c>.
          </p>
      </desc>
    </func>
    <func>
      <name name="log_terms" arity="2"/>
      <name name="blog_terms" arity="2"/>
      <fsummary>Log several items onto a disk log.</fsummary>
      <type variable="Log"/>
      <type variable="TermList" name_i="1"/>
      <type variable="BytesList"/>
      <type name="log_error_rsn"/>
      <desc>
        <p>The <c>log_terms/2</c> and <c>blog_terms/2</c> functions
          synchronously append a list of items to the log. The benefit
          of using these functions rather than the <c>log/2</c> and
          <c>blog/2</c> functions is that of efficiency: the given
          list is split into as large sublists as possible (limited by
          the size of wrap log files), and each sublist is logged as
          one single item, which reduces the overhead.
          </p>
        <p>The <c>log_terms/2</c> function is used for internally formatted
          logs, and <c>blog_terms/2</c> for externally formatted logs.
          <c>blog_terms/2</c> can be used
          for internally formatted logs as well provided the binaries were
          constructed with calls to <c>term_to_binary/1</c>.
          </p>
        <p>The owners that subscribe to notifications will be notified 
          of an error with an <c>error_status</c> message if the error 
          reason tag is <c>invalid_header</c> or <c>file_error</c>.
          </p>
      </desc>
    </func>
    <func>
      <name name="open" arity="1"/>
      <fsummary>Open a disk log file.</fsummary>
      <type name="dlog_options"/>
      <type name="dlog_option"/>
      <type name="open_ret"/>
      <type name="ret"/>
      <type name="dist_open_ret"/>
      <type name="dist_error_rsn"/>
      <type name="open_error_rsn"/>
      <type name="dlog_optattr"/>
      <type name="dlog_size"/>
      <desc>
        <p>The <c><anno>ArgL</anno></c> parameter is a list of options which have
          the following meanings:</p>
        <list type="bulleted">
          <item>
            <p><c>{name, <anno>Log</anno>}</c> specifies the name of the log.
              This is the name which must be passed on as a parameter in
              all subsequent logging operations. A name must always
              be supplied.
              </p>
          </item>
          <item>
            <p><c>{file, <anno>FileName</anno>}</c> specifies the name of the
              file which will be used for logged terms. If this value is
              omitted and the name of the log is either an atom or a string, 
              the file name will default to <c>lists:concat([<anno>Log</anno>, ".LOG"])</c> for halt logs. For wrap logs, this will be
              the base name of the files. Each file in a wrap log
              will be called <c><![CDATA[<base_name>.N]]></c>, where <c>N</c> is an
              integer. Each wrap log will also have two files called
              <c><![CDATA[<base_name>.idx]]></c> and <c><![CDATA[<base_name>.siz]]></c>.
              </p>
          </item>
          <item>
            <p><c>{linkto, <anno>LinkTo</anno>}</c>.               <marker id="linkto"></marker>
If
              <c><anno>LinkTo</anno></c> is a pid, that pid becomes an owner of the
              log. If <c><anno>LinkTo</anno></c> is <c>none</c> the log records
              that it is used anonymously by some process by
              incrementing the <c>users</c> counter. By default, the
              process which calls <c>open/1</c> owns the log.
              </p>
          </item>
          <item>
            <p><c>{repair, <anno>Repair</anno>}</c>. If <c><anno>Repair</anno></c> is <c>true</c>,
              the current log file will be repaired, if needed. As the
              restoration is initiated, a message is output on the error log.
              If <c>false</c> is given,
              no automatic repair will be attempted. Instead, the
              tuple <c>{error, {need_repair, <anno>Log</anno>}}</c> is returned if an
              attempt is made to open a corrupt log file. 
              If <c>truncate</c> is given, the log file will
              be truncated, creating an empty log. Default is 
              <c>true</c>, which has no effect on logs opened in 
              read-only mode.
              </p>
          </item>
          <item>
            <p><c>{type, <anno>Type</anno>}</c> is the type of the log. Default
              is <c>halt</c>.
              </p>
          </item>
          <item>
            <p><c>{format, <anno>Format</anno>}</c> specifies the format of the
              disk log. Default is <c>internal</c>. 
              </p>
          </item>
          <item>
            <p><c>{size, <anno>Size</anno>}</c> specifies the size of the log.
              When a halt log has reached its maximum size, all attempts to 
              log more items are rejected. The default size is 
              <c>infinity</c>, which for halt implies that there is no
              maximum size. For wrap logs, the <c><anno>Size</anno></c> parameter
              may be either a pair
              <c>{<anno>MaxNoBytes</anno>, <anno>MaxNoFiles</anno>}</c> or <c>infinity</c>. In the
              latter case, if the files of an already existing wrap log
              with the same name can be found, the size is read
              from the existing wrap log, otherwise an error is returned. 
              Wrap logs write at most <c><anno>MaxNoBytes</anno></c> bytes on each file
              and use <c><anno>MaxNoFiles</anno></c> files before starting all over with
              the first wrap log file. Regardless of <c><anno>MaxNoBytes</anno></c>,
              at least the header (if there is one) and one
              item is written on each wrap log file before
              wrapping to the next file.
              When opening an existing wrap log, it is not
              necessary to supply a value for the option <c>Size</c>, but any
              supplied value must equal the current size of the log, otherwise
              the tuple <c>{error, {size_mismatch, <anno>CurrentSize</anno>, <anno>NewSize</anno>}}</c>
              is returned.
              </p>
          </item>
          <item>
            <p><c>{distributed, <anno>Nodes</anno>}</c>. This option can be used for
              adding members to a distributed disk log. The
              default value is <c>[]</c>, which means that
              the log is local on the current node.
              </p>
          </item>
          <item>
            <marker id="notify"></marker>
            <p><c>{notify, bool()}</c>. If <c>true</c>, the owners of the
              log are notified when certain events occur in the log.
              Default is <c>false</c>. The owners are sent one of the
              following messages when an event occurs:
              </p>
            <list type="bulleted">
              <item>
                <p><c>{disk_log, Node, Log, {wrap, NoLostItems}}</c> is sent when a wrap log has
                  filled up one of its files and a new file is
                  opened. <c>NoLostItems</c> is the number of
                  previously logged items that have been lost when
                  truncating existing files.
                  </p>
              </item>
              <item>
                <p><c>{disk_log, Node, Log, {truncated, NoLostItems}}</c> is sent when a log has been
                  truncated or reopened. For halt logs <c>NoLostItems</c>
                  is the number of items written on the log since the 
                  disk log process was created. For wrap logs 
                  <c>NoLostItems</c> is the number of items on all 
                  wrap log files.
                  </p>
              </item>
              <item>
                <p><c>{disk_log, Node, Log, {read_only, Items}}</c> 
                  is sent when an asynchronous log attempt is made to 
                  a log file opened in read-only mode. 
                  <c>Items</c> is the items from the log attempt.
                  </p>
              </item>
              <item>
                <p><c>{disk_log, Node, Log, {blocked_log, Items}}</c> 
                  is sent when an asynchronous log attempt is made to
                  a blocked log that does not queue log attempts.
                  <c>Items</c> is the items from the log attempt.
                  </p>
              </item>
              <item>
                <p><c>{disk_log, Node, Log, {format_external, Items}}</c> 
                  is sent when <c>alog/2</c> or <c>alog_terms/2</c> is 
                  used for internally formatted logs. <c>Items</c> is the 
                  items from the log attempt.
                  </p>
              </item>
              <item>
                <p><c>{disk_log, Node, Log, full}</c> is sent when
                  an attempt to log items to a wrap log would write more
                  bytes than the limit set by the <c>size</c> option.
                  </p>
              </item>
              <item>
                <p><c>{disk_log, Node, Log, {error_status, Status}}</c> 
                  is sent when the error status changes. The error status
                  is defined by the outcome of the last attempt to log
                  items to a the log or to truncate the log or the last
                  use of <c>sync/1</c>, <c>inc_wrap_file/1</c> or
                  <c>change_size/2</c>. <c>Status</c> is one of <c>ok</c> and
                  <c>{error, Error}</c>, the former being the initial value.
                  </p>
              </item>
            </list>
          </item>
          <item>
            <p><c>{head, <anno>Head</anno>}</c> specifies a header to be
              written first on the log file. If the log is a wrap
              log, the item <c><anno>Head</anno></c> is written first in each new file.
              <c><anno>Head</anno></c> should be a term if the format is
              <c>internal</c>, and a deep list of bytes (or a binary)
              otherwise. Default is <c>none</c>, which means that
              no header is written first on the file.
              </p>
          </item>
          <item>
            <p><c>{head_func, {M,F,A}}</c> specifies a function
              to be called each time a new log file is opened.
              The call <c>M:F(A)</c> is assumed to return <c>{ok, Head}</c>.
              The item <c>Head</c> is written first in each file.
              <c>Head</c> should be a term if the format is
              <c>internal</c>, and a deep list of bytes (or a binary)
              otherwise.
              </p>
          </item>
          <item>
            <p><c>{mode, <anno>Mode</anno>}</c> specifies if the log is to be
              opened in read-only or read-write mode. It defaults to
              <c>read_write</c>.
              </p>
          </item>
        </list>
        <p>The <c>open/1</c> function returns <c>{ok, <anno>Log</anno>}</c> if the
          log file was successfully opened. If the file was
          successfully repaired, the tuple <c>{repaired, <anno>Log</anno>, {recovered, <anno>Rec</anno>}, {badbytes, <anno>Bad</anno>}}</c> is returned, where
          <c><anno>Rec</anno></c> is the number of whole Erlang terms found in the
          file and <c><anno>Bad</anno></c> is the number of bytes in the file which
          were non-Erlang terms. If the <c>distributed</c> parameter
          was given, <c>open/1</c> returns a list of
          successful replies and a list of erroneous replies. Each
          reply is tagged with the node name.
          </p>
        <p>When a disk log is opened in read-write mode, any existing
          log file is checked for. If there is none a new empty
          log is created, otherwise the existing file is opened at the
          position after the last logged item, and the logging of items
          will commence from there. If the format is <c>internal</c>
          and the existing file is not recognized as an internally
          formatted log, a tuple <c>{error, {not_a_log_file, <anno>FileName</anno>}}</c>
          is returned.
          </p>
        <p>The <c>open/1</c> function cannot be used for changing the
          values of options of an already open log; when there are prior
          owners or users of a log, all option values except <c>name</c>, 
          <c>linkto</c> and <c>notify</c> are just checked against 
          the values that have been supplied before as option values
          to <c>open/1</c>, <c>change_header/2</c>, <c>change_notify/3</c>
          or <c>change_size/2</c>. As a consequence,
          none of the options except <c>name</c> is mandatory. If some
          given value differs from the current value, a tuple
          <c>{error, {arg_mismatch, <anno>OptionName</anno>, <anno>CurrentValue</anno>, <anno>Value</anno>}}</c>
          is returned. Caution: an owner's attempt to open a log 
          as owner once again is acknowledged with the return value 
          <c>{ok, <anno>Log</anno>}</c>, but the state of the disk log is not
          affected in any way.
          </p>
        <p>If a log with a given name is local on some node,
          and one tries to open the log distributed on the same node,
          then the tuple <c>{error, {node_already_open, <anno>Log</anno>}}</c> is
          returned. The same tuple is returned if the log is distributed on
          some node, and one tries to open the log locally on the same node.
          Opening individual distributed disk logs for the first time
          adds those logs to a (possibly empty) distributed disk log.
          The option values supplied are used 
          on all nodes mentioned by the <c>distributed</c> option. 
          Individual distributed logs know nothing
          about each other's option values, so each node can be
          given unique option values by creating a distributed
          log with several calls to <c>open/1</c>.
          </p>
        <p>It is possible to open a log file more than once by giving
          different values to the option <c>name</c> or by using the
          same file when distributing a log on different nodes.
          It is up to the user of the <c>disk_log</c>
          module to ensure that no more than one
          disk log process has write access to any file, or the 
          the file may be corrupted.
          </p>
        <p>If an attempt to open a log file for the first time fails, 
          the disk log process terminates with the EXIT message 
          <c>{{failed,Reason},[{disk_log,open,1}]}</c>.
          The function returns <c>{error, Reason}</c> for all other errors.
          </p>
      </desc>
    </func>
    <func>
      <name name="pid2name" arity="1"/>
      <fsummary>Return the name of the disk log handled by a pid.</fsummary>
      <desc>
        <p>The <c>pid2name/1</c> function returns the name of the log
          given the pid of a disk log process on the current node, or
          <c>undefined</c> if the given pid is not a disk log process.
          </p>
        <p>This function is meant to be used for debugging only.
          </p>
      </desc>
    </func>
    <func>
      <name name="reopen" arity="2"/>
      <name name="reopen" arity="3"/>
      <name name="breopen" arity="3"/>
      <fsummary>Reopen a disk log and save the old log.</fsummary>
      <type variable="Log"/>
      <type variable="File" name_i="1"/>
      <type variable="Head" name_i="2"/>
      <type variable="BHead"/>
      <type name="reopen_error_rsn"/>
      <desc>
        <p>The <c>reopen</c> functions first rename the log file 
          to <c><anno>File</anno></c> and then re-create a new log file.
          In case of a wrap log, <c><anno>File</anno></c> is used as the base name
          of the renamed files.
          By default the header given to <c>open/1</c> is written first in
          the newly opened log file, but if the <c><anno>Head</anno></c> or the
          <c><anno>BHead</anno></c> argument is given, this item is used instead.
          The header argument is used once only; next time a wrap log file
          is opened, the header given to <c>open/1</c> is used.
          </p>
        <p>The <c>reopen/2,3</c> functions are used for internally formatted 
          logs, and <c>breopen/3</c> for externally formatted logs.
          </p>
        <p>The owners that subscribe to notifications will receive 
          a <c>truncate</c> message.
          </p>
        <p>Upon failure to reopen the log, the disk log process terminates
          with the EXIT message <c>{{failed,Error},[{disk_log,Fun,Arity}]}</c>,
          and other processes that have requests queued receive the message 
          <c>{disk_log, Node, {error, disk_log_stopped}}</c>.
          </p>
      </desc>
    </func>
    <func>
      <name name="sync" arity="1"/>
      <fsummary>Flush the contents of a disk log to the disk.</fsummary>
      <type name="sync_error_rsn"/>
      <desc>
        <p>The <c>sync/1</c> function ensures that the contents of the 
          log are actually written to the disk. 
          This is usually a rather expensive operation.
          </p>
      </desc>
    </func>
    <func>
      <name name="truncate" arity="1"/>
      <name name="truncate" arity="2"/>
      <name name="btruncate" arity="2"/>
      <fsummary>Truncate a disk log.</fsummary>
      <type variable="Log"/>
      <type variable="Head" name_i="2"/>
      <type variable="BHead"/>
      <type name="trunc_error_rsn"/>
      <desc>
        <p>The <c>truncate</c> functions remove all items from a disk log.
          If the <c><anno>Head</anno></c> or the <c><anno>BHead</anno></c> argument is
          given, this item is written first in the newly truncated
          log, otherwise the header given to <c>open/1</c> is used.
          The header argument is only used once; next time a wrap log file 
          is opened, the header given to <c>open/1</c> is used.
          </p>
        <p>The <c>truncate/1,2</c> functions are used for internally 
          formatted logs, and <c>btruncate/2</c> for externally formatted
          logs.
          </p>
        <p>The owners that subscribe to notifications will receive 
          a <c>truncate</c> message.
          </p>
        <p>If the attempt to truncate the log fails, the disk log process
          terminates with the EXIT message 
          <c>{{failed,Reason},[{disk_log,Fun,Arity}]}</c>, and 
          other processes that have requests queued receive the message 
          <c>{disk_log, Node, {error, disk_log_stopped}}</c>.
          </p>
      </desc>
    </func>
    <func>
      <name name="unblock" arity="1"/>
      <fsummary>Unblock a disk log.</fsummary>
      <type name="unblock_error_rsn"/>
      <desc>
        <p>The <c>unblock/1</c> function unblocks a log. 
          A log can only be unblocked by the blocking process.
          </p>
      </desc>
    </func>
  </funcs>

  <section>
    <title>See Also</title>
    <p><seealso marker="file">file(3)</seealso>, 
      <seealso marker="pg2">pg2(3)</seealso>, 
      <seealso marker="wrap_log_reader">wrap_log_reader(3)</seealso></p>
  </section>
</erlref>