<appendix id="glossary">
 <title>Glossary</title>
 <para>
  This is a list of terms and their meaning in the context of
  <productname>PostgreSQL</productname> and relational database
  systems in general.
 </para>

 <glosslist>
  <glossentry id="glossary-acid">
   <glossterm>ACID</glossterm>
   <glossdef>
    <para>
     <glossterm linkend="glossary-atomicity">Atomicity</glossterm>,
     <glossterm linkend="glossary-consistency">Consistency</glossterm>,
     <glossterm linkend="glossary-isolation">Isolation</glossterm>, and
     <glossterm linkend="glossary-durability">Durability</glossterm>.
     This set of properties of database transactions is intended to
     guarantee validity in concurrent operation and even in event of
     errors, power failures, etc.
    </para>
   </glossdef>
  </glossentry>

  <glossentry id="glossary-aggregate">
   <glossterm>Aggregate function (routine)</glossterm>
   <glossdef>
    <para>
     A <glossterm linkend="glossary-function">function</glossterm> that
     combines (<firstterm>aggregates</firstterm>) multiple input values,
     for example by counting, averaging or adding,
     yielding a single output value.
    </para>
    <para>
     For more information, see
     <xref linkend="functions-aggregate"/>.
    </para>
    <glossseealso otherterm="glossary-window-function" />
   </glossdef>
  </glossentry>

  <glossentry id="glossary-am">
   <glossterm>Access Method</glossterm>
   <glossdef>
    <para>
     Interfaces which <productname>PostgreSQL</productname> use in order to
     access data in tables and indexes.  This abstraction allows for adding
     support for new types of data storage.
    </para>
    <para>
     For more information, see <xref linkend="tableam" /> and
     <xref linkend="indexam" />.
    </para>
   </glossdef>
  </glossentry>

  <glossentry>
   <glossterm>Analytic function</glossterm>
   <glosssee otherterm="glossary-window-function" />
  </glossentry>

  <glossentry id="glossary-analyze">
   <glossterm>Analyze (operation)</glossterm>
   <glossdef>
    <para>
     The act of collecting statistics from data in
     <glossterm linkend="glossary-table">tables</glossterm>
     and other <glossterm linkend="glossary-relation">relations</glossterm>
     to help the <glossterm linkend="glossary-planner">query planner</glossterm>
     to make decisions about how to execute
     <glossterm linkend="glossary-query">queries</glossterm>.
    </para>
    <para>
     (Don't confuse this term with the <literal>ANALYZE</literal> option
     to the <xref linkend="sql-explain"/> command.)
    </para>
    <para>
     For more information, see
     <xref linkend="sql-analyze"/>.
    </para>
   </glossdef>
  </glossentry>

  <glossentry id="glossary-aio">
   <glossterm>Asynchronous <acronym>I/O</acronym></glossterm>
   <acronym>AIO</acronym>
   <indexterm>
    <primary>Asynchronous <acronym>I/O</acronym></primary>
   </indexterm>
   <glossdef>
    <para>
     Asynchronous <acronym>I/O</acronym> (<acronym>AIO</acronym>) describes
     performing <acronym>I/O</acronym> in a non-blocking way (asynchronously),
     in contrast to synchronous <acronym>I/O</acronym>, which blocks for the
     entire duration of the <acronym>I/O</acronym>.
    </para>
    <para>
     With <acronym>AIO</acronym>, starting an <acronym>I/O</acronym> operation
     is separated from waiting for the result of the operation, allowing
     multiple <acronym>I/O</acronym> operations to be initiated concurrently,
     as well as performing <acronym>CPU</acronym> heavy operations
     concurrently with <acronym>I/O</acronym>. The price for that increased
     concurrency is increased complexity.
    </para>
    <glossseealso otherterm="glossary-io" />
   </glossdef>
  </glossentry>

  <glossentry id="glossary-atomic">
   <glossterm>Atomic</glossterm>
   <glossdef>
    <para>
     In reference to a <glossterm linkend="glossary-datum">datum</glossterm>:
     the fact that its value cannot be broken down into smaller