rt4-doc-html 4.4.2-2 / usr / share / doc / rt4-doc-html / RT_Config.html

<ul id="index">
  <li><a href="#NAME">NAME</a></li>
  <li><a href="#Base-configuration">Base configuration</a></li>
  <li><a href="#Database-connection">Database connection</a></li>
  <li><a href="#Logging">Logging</a></li>
  <li><a href="#Incoming-mail-gateway">Incoming mail gateway</a></li>
  <li><a href="#Outgoing-mail">Outgoing mail</a>
    <ul>
      <li><a href="#Email-dashboards">Email dashboards</a></li>
      <li><a href="#Sendmail-configuration">Sendmail configuration</a></li>
      <li><a href="#Other-mailers">Other mailers</a></li>
    </ul>
  </li>
  <li><a href="#Web-interface">Web interface</a>
    <ul>
      <li><a href="#Home-page">Home page</a></li>
      <li><a href="#Ticket-search">Ticket search</a></li>
      <li><a href="#Ticket-display">Ticket display</a></li>
      <li><a href="#Self-Service-Interface">Self Service Interface</a></li>
      <li><a href="#Articles">Articles</a></li>
    </ul>
  </li>
  <li><a href="#Assets">Assets</a>
    <ul>
      <li><a href="#Message-box-properties">Message box properties</a></li>
      <li><a href="#Transaction-display">Transaction display</a></li>
    </ul>
  </li>
  <li><a href="#Application-logic">Application logic</a></li>
  <li><a href="#Extra-security">Extra security</a></li>
  <li><a href="#Authorization-and-user-configuration">Authorization and user configuration</a>
    <ul>
      <li><a href="#External-Authentication-and-Authorization">External Authentication and Authorization</a></li>
      <li><a href="#Syncing-Users-and-Groups-with-LDAP-or-AD">Syncing Users and Groups with LDAP or AD</a></li>
    </ul>
  </li>
  <li><a href="#Internationalization">Internationalization</a></li>
  <li><a href="#Date-and-time-handling">Date and time handling</a></li>
  <li><a href="#Cryptography">Cryptography</a>
    <ul>
      <li><a href="#SMIME-configuration">SMIME configuration</a></li>
      <li><a href="#GnuPG-configuration">GnuPG configuration</a></li>
    </ul>
  </li>
  <li><a href="#External-storage">External storage</a></li>
  <li><a href="#Initialdata-Formats">Initialdata Formats</a></li>
  <li><a href="#Lifecycles">Lifecycles</a>
    <ul>
      <li><a href="#Lifecycle-definitions">Lifecycle definitions</a></li>
      <li><a href="#Default-values">Default values</a></li>
      <li><a href="#Transitions-between-statuses-and-UI-actions">Transitions between statuses and UI actions</a>
        <ul>
          <li><a href="#Statuses-available-during-ticket-creation">Statuses available during ticket creation</a></li>
          <li><a href="#Protecting-status-changes-with-rights">Protecting status changes with rights</a></li>
          <li><a href="#Labeling-and-defining-actions">Labeling and defining actions</a></li>
        </ul>
      </li>
      <li><a href="#Moving-tickets-between-queues-with-different-lifecycles">Moving tickets between queues with different lifecycles</a></li>
    </ul>
  </li>
  <li><a href="#SLA">SLA</a></li>
  <li><a href="#Administrative-interface">Administrative interface</a></li>
  <li><a href="#Development-options">Development options</a></li>
</ul>

<h1 id="NAME"><a href="#___top">NAME</a></h1>

<p>RT::Config</p>

<h1 id="Base-configuration"><a href="#___top">Base configuration</a></h1>

<dl>

<dt id="rtname"><code>$rtname</code></dt>
<dd>

<p><code>$rtname</code> is the string that RT will look for in mail messages to figure out what ticket a new piece of mail belongs to.</p>

<p>Your domain name is recommended, so as not to pollute the namespace. Once you start using a given tag, you should probably never change it; otherwise, mail for existing tickets won&#39;t get put in the right place.</p>

</dd>
<dt id="Organization"><code>$Organization</code></dt>
<dd>

<p>You should set this to your organization&#39;s DNS domain. For example, <i>fsck.com</i> or <i>asylum.arkham.ma.us</i>. It is used by the linking interface to guarantee that ticket URIs are unique and easy to construct. Changing it after you have created tickets in the system will <b>break</b> all existing ticket links!</p>

</dd>
<dt id="CorrespondAddress-CommentAddress"><code>$CorrespondAddress</code>, <code>$CommentAddress</code></dt>
<dd>

<p>RT is designed such that any mail which already has a ticket-id associated with it will get to the right place automatically.</p>

<p><code>$CorrespondAddress</code> and <code>$CommentAddress</code> are the default addresses that will be listed in From: and Reply-To: headers of correspondence and comment mail tracked by RT, unless overridden by a queue-specific address. They should be set to email addresses which have been configured as aliases for <i><a href="rt-mailgate.html">rt-mailgate</a></i>.</p>

</dd>
<dt id="WebDomain"><code>$WebDomain</code></dt>
<dd>

<p>Domain name of the RT server, e.g. &#39;www.example.com&#39;. It should not contain anything except the server name.</p>

</dd>
<dt id="WebPort"><code>$WebPort</code></dt>
<dd>

<p>If we&#39;re running as a superuser, run on port 80. Otherwise, pick a high port for this user.</p>

<p>443 is default port for https protocol.</p>

</dd>
<dt id="WebPath"><code>$WebPath</code></dt>
<dd>

<p>If you&#39;re putting the web UI somewhere other than at the root of your server, you should set <code>$WebPath</code> to the path you&#39;ll be serving RT at.</p>

<p><code>$WebPath</code> requires a leading / but no trailing /, or it can be blank.</p>

<p>In most cases, you should leave <code>$WebPath</code> set to &quot;&quot; (an empty value).</p>

</dd>
<dt id="Timezone"><code>$Timezone</code></dt>
<dd>

<p><code>$Timezone</code> is the default timezone, used to convert times entered by users into GMT, as they are stored in the database, and back again; users can override this. It should be set to a timezone recognized by your server.</p>

</dd>
<dt id="Plugins"><code>@Plugins</code></dt>
<dd>

<p>Once a plugin has been downloaded and installed, use <code>Plugin()</code> to add to the enabled <code>@Plugins</code> list:</p>

<pre><code>    Plugin( &quot;RT::Extension::JSGantt&quot; );</code></pre>

<p>RT will also accept the distribution name (i.e. <code>RT-Extension-JSGantt</code>) instead of the package name (<code>RT::Extension::JSGantt</code>).</p>

</dd>
<dt id="StaticRoots"><code>@StaticRoots</code></dt>
<dd>

<p>Set <code>@StaticRoots</code> to serve extra paths with a static handler. The contents of each hashref should be the the same arguments as <a href="http://metacpan.org/module/Plack::Middleware::Static">Plack::Middleware::Static</a> takes. These paths will be checked before any plugin or core static paths.</p>

<p>Example:</p>

<pre><code>    Set( @StaticRoots,
        {
            path =&gt; qr{^/static/},
            root =&gt; &#39;/local/path/to/static/parent&#39;,
        },
    );</code></pre>

</dd>
</dl>

<h1 id="Database-connection"><a href="#___top">Database connection</a></h1>

<dl>

<dt id="DatabaseType"><code>$DatabaseType</code></dt>
<dd>

<p>Database driver being used; case matters. Valid types are &quot;mysql&quot;, &quot;Oracle&quot;, and &quot;Pg&quot;. &quot;SQLite&quot; is also available for non-production use.</p>

</dd>
<dt id="DatabaseHost-DatabaseRTHost"><code>$DatabaseHost</code>, <code>$DatabaseRTHost</code></dt>
<dd>

<p>The domain name of your database server. If you&#39;re running MySQL and on localhost, leave it blank for enhanced performance.</p>

<p><code>DatabaseRTHost</code> is the fully-qualified hostname of your RT server, for use in granting ACL rights on MySQL.</p>

</dd>
<dt id="DatabasePort"><code>$DatabasePort</code></dt>
<dd>

<p>The port that your database server is running on. Ignored unless it&#39;s a positive integer. It&#39;s usually safe to leave this blank; RT will choose the correct default.</p>

</dd>
<dt id="DatabaseUser"><code>$DatabaseUser</code></dt>
<dd>

<p>The name of the user to connect to the database as.</p>

</dd>
<dt id="DatabasePassword"><code>$DatabasePassword</code></dt>
<dd>

<p>The password the <code>$DatabaseUser</code> should use to access the database.</p>

</dd>
<dt id="DatabaseName"><code>$DatabaseName</code></dt>
<dd>

<p>The name of the RT database on your database server. For Oracle, the SID and database objects are created in <code>$DatabaseUser</code>&#39;s schema.</p>

</dd>
<dt id="DatabaseExtraDSN"><code>%DatabaseExtraDSN</code></dt>
<dd>

<p>Allows additional properties to be passed to the database connection step. Possible properties are specific to the database-type; see https://metacpan.org/pod/DBI#connect</p>

<p>For PostgreSQL, for instance, the following enables SSL (but does no certificate checking, providing data hiding but no MITM protection):</p>

<pre><code>   # See https://metacpan.org/pod/DBD::Pg#connect
   # and http://www.postgresql.org/docs/8.4/static/libpq-ssl.html
   Set( %DatabaseExtraDSN, sslmode =&gt; &#39;require&#39; );</code></pre>

<p>For MySQL, the following acts similarly if the server has enabled SSL. Otherwise, it provides no protection; MySQL provides no way to <i>force</i> SSL connections:</p>

<pre><code>   # See https://metacpan.org/pod/DBD::mysql#connect
   # and http://dev.mysql.com/doc/refman/5.1/en/ssl-options.html
   Set( %DatabaseExtraDSN, mysql_ssl =&gt; 1 );</code></pre>

</dd>
<dt id="DatabaseAdmin"><code>$DatabaseAdmin</code></dt>
<dd>

<p>The name of the database administrator to connect to the database as during upgrades.</p>

</dd>
</dl>

<h1 id="Logging"><a href="#___top">Logging</a></h1>

<p>The default is to log anything except debugging information to syslog. Check the <a href="http://metacpan.org/module/Log::Dispatch">Log::Dispatch</a> POD for information about how to get things by syslog, mail or anything else, get debugging info in the log, etc.</p>

<p>It might generally make sense to send error and higher by email to some administrator. If you do this, be careful that this email isn&#39;t sent to this RT instance. Mail loops will generate a critical log message.</p>

<dl>

<dt id="LogToSyslog-LogToSTDERR"><code>$LogToSyslog</code>, <code>$LogToSTDERR</code></dt>
<dd>

<p>The minimum level error that will be logged to the specific device. From lowest to highest priority, the levels are:</p>

<pre><code>    debug info notice warning error critical alert emergency</code></pre>

<p>Many syslogds are configured to discard or file debug messages away, so if you&#39;re attempting to debug RT you may need to reconfigure your syslogd or use one of the other logging options.</p>

<p>Logging to your screen affects scripts run from the command line as well as the STDERR sent to your webserver (so these logs will usually show up in your web server&#39;s error logs).</p>

</dd>
<dt id="LogToFile-LogDir-LogToFileNamed"><code>$LogToFile</code>, <code>$LogDir</code>, <code>$LogToFileNamed</code></dt>
<dd>

<p>Logging to a standalone file is also possible. The file needs to both exist and be writable by all direct users of the RT API. This generally includes the web server and whoever rt-crontool runs as. Note that rt-mailgate and the RT CLI go through the webserver, so their users do not need to have write permissions to this file. If you expect to have multiple users of the direct API, Best Practical recommends using syslog instead of direct file logging.</p>

<p>You should set <code>$LogToFile</code> to one of the levels documented above.</p>

</dd>
<dt id="LogStackTraces"><code>$LogStackTraces</code></dt>
<dd>

<p>If set to a log level then logging will include stack traces for messages with level equal to or greater than specified.</p>

<p>NOTICE: Stack traces include parameters supplied to functions or methods. It is possible for stack trace logging to reveal sensitive information such as passwords or ticket content in your logs.</p>

</dd>
<dt id="LogToSyslogConf"><code>@LogToSyslogConf</code></dt>
<dd>

<p>Additional options to pass to <a href="http://metacpan.org/module/Log::Dispatch::Syslog">Log::Dispatch::Syslog</a>; the most interesting flags include <code>facility</code>, <code>logopt</code>, and possibly <code>ident</code>. See the <a href="http://metacpan.org/module/Log::Dispatch::Syslog">Log::Dispatch::Syslog</a> documentation for more information.</p>

</dd>
</dl>

<h1 id="Incoming-mail-gateway"><a href="#___top">Incoming mail gateway</a></h1>

<dl>

<dt id="EmailSubjectTagRegex"><code>$EmailSubjectTagRegex</code></dt>
<dd>

<p>This regexp controls what subject tags RT recognizes as its own. If you&#39;re not dealing with historical <code>$rtname</code> values, or historical queue-specific subject tags, you&#39;ll likely never have to change this configuration.</p>

<p>Be <b>very careful</b> with it. Note that it overrides <code>$rtname</code> for subject token matching.</p>

<p>The setting below would make RT behave exactly as it does without the setting enabled.</p>

</dd>
<dt id="OwnerEmail"><code>$OwnerEmail</code></dt>
<dd>

<p><code>$OwnerEmail</code> is the address of a human who manages RT. RT will send errors generated by the mail gateway to this address; it will also be displayed as the contact person on the RT&#39;s login page. Because RT sends errors to this address, it should <i>not</i> be an address that&#39;s managed by your RT instance, to avoid mail loops.</p>

</dd>
<dt id="LoopsToRTOwner"><code>$LoopsToRTOwner</code></dt>
<dd>

<p>If <code>$LoopsToRTOwner</code> is defined, RT will send mail that it believes might be a loop to <code>$OwnerEmail</code>.</p>

</dd>
<dt id="StoreLoops"><code>$StoreLoops</code></dt>
<dd>

<p>If <code>$StoreLoops</code> is defined, RT will record messages that it believes to be part of mail loops. As it does this, it will try to be careful not to send mail to the sender of these messages.</p>

</dd>
<dt id="MaxAttachmentSize"><code>$MaxAttachmentSize</code></dt>
<dd>

<p><code>$MaxAttachmentSize</code> sets the maximum size (in bytes) of attachments stored in the database. This setting is irrelevant unless one of $TruncateLongAttachments or $DropLongAttachments (below) are set, <b>OR</b> the database is stored in Oracle. On Oracle, attachments larger than this can be fully stored, but will be truncated to this length when read.</p>

</dd>
<dt id="TruncateLongAttachments"><code>$TruncateLongAttachments</code></dt>
<dd>

<p>If this is set to a non-undef value, RT will truncate attachments longer than <code>$MaxAttachmentSize</code>.</p>

</dd>
<dt id="DropLongAttachments"><code>$DropLongAttachments</code></dt>
<dd>

<p>If this is set to a non-undef value, RT will silently drop attachments longer than <code>MaxAttachmentSize</code>. <code>$TruncateLongAttachments</code>, above, takes priority over this.</p>

</dd>
<dt id="RTAddressRegexp"><code>$RTAddressRegexp</code></dt>
<dd>

<p><code>$RTAddressRegexp</code> is used to make sure RT doesn&#39;t add itself as a ticket CC if <code>$ParseNewMessageForTicketCcs</code>, above, is enabled. It is important that you set this to a regular expression that matches all addresses used by your RT. This lets RT avoid sending mail to itself. It will also hide RT addresses from the list of &quot;One-time Cc&quot; and Bcc lists on ticket reply.</p>

<p>If you have a number of addresses configured in your RT database already, you can generate a naive first pass regexp by using:</p>

<pre><code>    perl etc/upgrade/generate-rtaddressregexp</code></pre>

<p>If left blank, RT will compare each address to your configured <code>$CorrespondAddress</code> and <code>$CommentAddress</code> before searching for a Queue configured with a matching &quot;Reply Address&quot; or &quot;Comment Address&quot; on the Queue Admin page.</p>

</dd>
<dt id="CanonicalizeEmailAddressMatch-CanonicalizeEmailAddressReplace"><code>$CanonicalizeEmailAddressMatch</code>, <code>$CanonicalizeEmailAddressReplace</code></dt>
<dd>

<p>RT provides functionality which allows the system to rewrite incoming email addresses, using <a href="RT/User.html#CanonicalizeEmailAddress">&quot;CanonicalizeEmailAddress&quot; in RT::User</a>. The default implementation replaces all occurrences of the regular expression in <code>CanonicalizeEmailAddressMatch</code> with <code>CanonicalizeEmailAddressReplace</code>, via <code>s/$Match/$Replace/gi</code>. The most common use of this is to replace <code>@something.example.com</code> with <code>@example.com</code>. If more complex noramlization is required, <a href="RT/User.html#CanonicalizeEmailAddress">&quot;CanonicalizeEmailAddress&quot; in RT::User</a> can be overridden to provide it.</p>

</dd>
<dt id="ValidateUserEmailAddresses"><code>$ValidateUserEmailAddresses</code></dt>
<dd>

<p>By default <code>$ValidateUserEmailAddresses</code> is 1, and RT will refuse to create users with an invalid email address (as specified in RFC 2822) or with an email address made of multiple email addresses.</p>

<p>Set this to 0 to skip any email address validation. Doing so may open up vulnerabilities.</p>

</dd>
<dt id="MailPlugins"><code>@MailPlugins</code></dt>
<dd>

<p><code>@MailPlugins</code> is a list of authentication plugins for <a href="RT/Interface/Email.html">RT::Interface::Email</a> to use; see <a href="rt-mailgate.html">rt-mailgate</a></p>

</dd>
<dt id="ExtractSubjectTagMatch-ExtractSubjectTagNoMatch"><code>$ExtractSubjectTagMatch</code>, <code>$ExtractSubjectTagNoMatch</code></dt>
<dd>

<p>The default &quot;extract remote tracking tags&quot; scrip settings; these detect when your RT is talking to another RT, and adjust the subject accordingly.</p>

</dd>
<dt id="CheckMoreMSMailHeaders"><code>$CheckMoreMSMailHeaders</code></dt>
<dd>

<p>Some email clients create a plain text version of HTML-formatted email to help other clients that read only plain text. Unfortunately, the plain text parts sometimes end up with doubled newlines and these can then end up in RT. This is most often seen in MS Outlook.</p>

<p>Enable this option to have RT check for additional mail headers and attempt to identify email from MS Outlook. When detected, RT will then clean up double newlines. Note that it may clean up intentional double newlines as well.</p>

</dd>
</dl>

<h1 id="Outgoing-mail"><a href="#___top">Outgoing mail</a></h1>

<dl>

<dt id="MailCommand"><code>$MailCommand</code></dt>
<dd>

<p><code>$MailCommand</code> defines which method RT will use to try to send mail. We know that &#39;sendmailpipe&#39; works fairly well. If &#39;sendmailpipe&#39; doesn&#39;t work well for you, try &#39;sendmail&#39;. &#39;qmail&#39; is also a supported value.</p>

<p>For testing purposes, or to simply disable sending mail out into the world, you can set <code>$MailCommand</code> to &#39;mbox&#39; which logs all mail, in mbox format, to files in <i>/opt/rt4/var/</i> based in the process start time. The &#39;testfile&#39; option is similar, but the files that it creates (under /tmp) are temporary, and removed upon process completion; the format is also not mbox-compatable.</p>

</dd>
<dt id="SetOutgoingMailFrom"><code>$SetOutgoingMailFrom</code></dt>
<dd>

<p><code>$SetOutgoingMailFrom</code> tells RT to set the sender envelope to the Correspond mail address of the ticket&#39;s queue.</p>

<p>Warning: If you use this setting, bounced mails will appear to be incoming mail to the system, thus creating new tickets.</p>

<p>If the value contains an <code>@</code>, it is assumed to be an email address and used as a global envelope sender. Expected usage in this case is to simply set the same envelope sender on all mail from RT, without defining <code>$OverrideOutgoingMailFrom</code>. If you do define <code>$OverrideOutgoingMailFrom</code>, anything specified there overrides the global value (including Default).</p>

<p>This option only works if <code>$MailCommand</code> is set to &#39;sendmailpipe&#39;.</p>

</dd>
<dt id="OverrideOutgoingMailFrom"><code>$OverrideOutgoingMailFrom</code></dt>
<dd>

<p><code>$OverrideOutgoingMailFrom</code> is used for overwriting the Correspond address of the queue as it is handed to sendmail -f. This helps force the From_ header away from www-data or other email addresses that show up in the &quot;Sent by&quot; line in Outlook.</p>

<p>The option is a hash reference of queue id/name to email address. If there is no ticket involved, then the value of the <code>Default</code> key will be used.</p>

<p>This option only works if <code>$SetOutgoingMailFrom</code> is enabled and <code>$MailCommand</code> is set to &#39;sendmailpipe&#39;.</p>

</dd>
<dt id="DefaultMailPrecedence"><code>$DefaultMailPrecedence</code></dt>
<dd>

<p><code>$DefaultMailPrecedence</code> is used to control the default Precedence level of outgoing mail where none is specified. By default it is <code>bulk</code>, but if you only send mail to your staff, you may wish to change it.</p>

<p>Note that you can set the precedence of individual templates by including an explicit Precedence header.</p>

<p>If you set this value to <code>undef</code> then we do not set a default Precedence header to outgoing mail. However, if there already is a Precedence header, it will be preserved.</p>

</dd>
<dt id="OverrideMailPrecedence"><code>$OverrideMailPrecedence</code></dt>
<dd>

<p><code>$OverrideMailPrecedence</code> is used for overwriting the <code>$DefaultMailPrecedence</code> value for a queue.</p>

<p>The option is a hash reference of queue id/name to precedence. If you set the precedence to <code>undef</code>, a Precedence header will not be added to the mail.</p>

<p>This option only works if <code>$DefaultMailPrecedence</code> is enabled.</p>

</dd>
<dt id="DefaultErrorMailPrecedence"><code>$DefaultErrorMailPrecedence</code></dt>
<dd>

<p><code>$DefaultErrorMailPrecedence</code> is used to control the default Precedence level of outgoing mail that indicates some kind of error condition. By default it is <code>bulk</code>, but if you only send mail to your staff, you may wish to change it.</p>

<p>If you set this value to <code>undef</code> then we do not add a Precedence header to error mail.</p>

</dd>
<dt id="UseOriginatorHeader"><code>$UseOriginatorHeader</code></dt>
<dd>

<p><code>$UseOriginatorHeader</code> is used to control the insertion of an RT-Originator Header in every outgoing mail, containing the mail address of the transaction creator.</p>

</dd>
<dt id="UseFriendlyFromLine"><code>$UseFriendlyFromLine</code></dt>
<dd>

<p>By default, RT sets the outgoing mail&#39;s &quot;From:&quot; header to &quot;SenderName via RT&quot;. Setting <code>$UseFriendlyFromLine</code> to 0 disables it.</p>

</dd>
<dt id="FriendlyFromLineFormat"><code>$FriendlyFromLineFormat</code></dt>
<dd>

<p><code>sprintf()</code> format of the friendly &#39;From:&#39; header; its arguments are SenderName and SenderEmailAddress.</p>

</dd>
<dt id="UseFriendlyToLine"><code>$UseFriendlyToLine</code></dt>
<dd>

<p>RT can optionally set a &quot;Friendly&quot; &#39;To:&#39; header when sending messages to Ccs or AdminCcs (rather than having a blank &#39;To:&#39; header.</p>

<p>This feature DOES NOT WORK WITH SENDMAIL[tm] BRAND SENDMAIL. If you are using sendmail, rather than postfix, qmail, exim or some other MTA, you _must_ disable this option.</p>

</dd>
<dt id="FriendlyToLineFormat"><code>$FriendlyToLineFormat</code></dt>
<dd>

<p><code>sprintf()</code> format of the friendly &#39;To:&#39; header; its arguments are WatcherType and TicketId.</p>

</dd>
<dt id="NotifyActor"><code>$NotifyActor</code></dt>
<dd>

<p>By default, RT doesn&#39;t notify the person who performs an update, as they already know what they&#39;ve done. If you&#39;d like to change this behavior, Set <code>$NotifyActor</code> to 1</p>

</dd>
<dt id="RecordOutgoingEmail"><code>$RecordOutgoingEmail</code></dt>
<dd>

<p>By default, RT records each message it sends out to its own internal database. To change this behavior, set <code>$RecordOutgoingEmail</code> to 0</p>

<p>If this is disabled, users&#39; digest mail delivery preferences (i.e. EmailFrequency) will also be ignored.</p>

</dd>
<dt id="VERPPrefix-VERPDomain"><code>$VERPPrefix</code>, <code>$VERPDomain</code></dt>
<dd>

<p>Setting these options enables VERP support <a href="http://cr.yp.to/proto/verp.txt">http://cr.yp.to/proto/verp.txt</a>.</p>

<p>Uncomment the following two directives to generate envelope senders of the form <code>${VERPPrefix}${originaladdress}@${VERPDomain}</code> (i.e. rt-jesse=fsck.com@rt.example.com ).</p>

<p>This currently only works with sendmail and sendmailpipe.</p>

</dd>
<dt id="ForwardFromUser"><code>$ForwardFromUser</code></dt>
<dd>

<p>By default, RT forwards a message using queue&#39;s address and adds RT&#39;s tag into subject of the outgoing message, so recipients&#39; replies go into RT as correspondents.</p>

<p>To change this behavior, set <code>$ForwardFromUser</code> to 1 and RT will use the address of the current user and remove RT&#39;s subject tag.</p>

</dd>
<dt id="HTMLFormatter"><code>$HTMLFormatter</code></dt>
<dd>

<p>RT&#39;s default pure-perl formatter may fail to successfully convert even on some relatively simple HTML; this will result in blank <code>text/plain</code> parts, which is particuarly unfortunate if HTML templates are not in use.</p>

<p>If the optional dependency <a href="http://metacpan.org/module/HTML::FormatExternal">HTML::FormatExternal</a> is installed, RT will use external programs to render HTML to plain text. The default is to try, in order, <code>w3m</code>, <code>elinks</code>, <code>html2text</code>, <code>links</code>, <code>lynx</code>, and then fall back to the <code>core</code> pure-perl formatter if none are installed.</p>

<p>Set <code>$HTMLFormatter</code> to one of the above programs (or the full path to such) to use a different program than the above would choose by default. Setting this requires that <a href="http://metacpan.org/module/HTML::FormatExternal">HTML::FormatExternal</a> be installed.</p>

<p>If the chosen formatter is not in the webserver&#39;s $PATH, you may set this option the full path to one of the aforementioned executables.</p>

</dd>
</dl>

<h2 id="Email-dashboards"><a href="#___top">Email dashboards</a></h2>

<dl>

<dt id="DashboardAddress"><code>$DashboardAddress</code></dt>
<dd>

<p>The email address from which RT will send dashboards. If none is set, then <code>$OwnerEmail</code> will be used.</p>

</dd>
<dt id="DashboardSubject"><code>$DashboardSubject</code></dt>
<dd>

<p>Lets you set the subject of dashboards. Arguments are the frequency (Daily, Weekly, Monthly) of the dashboard and the dashboard&#39;s name.</p>

</dd>
<dt id="EmailDashboardRemove"><code>@EmailDashboardRemove</code></dt>
<dd>

<p>A list of regular expressions that will be used to remove content from mailed dashboards.</p>

</dd>
<dt id="EmailDashboardLanguageOrder"><code>@EmailDashboardLanguageOrder</code></dt>
<dd>

<p>A list that specifies which language to use for dashboard subscription email. There are several special keys:</p>

<p>* _subscription: the language chosen on the dashboard subscription page * _recipient: the recipient&#39;s language, as chosen on their &quot;About Me&quot; page * _subscriber: the subscriber&#39;s language, as chosen on their &quot;About Me&quot; page</p>

<p>The first key that produces a value is used for the email. Be aware that users may not actually have a language set on their &quot;About Me&quot; page, since RT falls back to the language their web browser specifies (and of course in a scheduled email dashboard, there is no web browser).</p>

<p>You may also include a specific language as a fallback when there is no language specified otherwise. Using a specific language never fails to produce a value, so subsequent values in the list will never be considered.</p>

<p>By default, RT examines the subscription, then the recipient, then subscriber, then finally falls back to English.</p>

<p>See also <a href="#LexiconLanguages">&quot;@LexiconLanguages&quot;</a>.</p>

</dd>
</dl>

<h2 id="Sendmail-configuration"><a href="#___top">Sendmail configuration</a></h2>

<p>These options only take effect if <code>$MailCommand</code> is &#39;sendmail&#39; or &#39;sendmailpipe&#39;</p>

<dl>

<dt id="SendmailArguments"><code>$SendmailArguments</code></dt>
<dd>

<p><code>$SendmailArguments</code> defines what flags to pass to <code>$SendmailPath</code> These options are good for most sendmail wrappers and work-a-likes.</p>

<p>These arguments are good for sendmail brand sendmail 8 and newer: <code>Set($SendmailArguments,&quot;-oi -ODeliveryMode=b -OErrorMode=m&quot;);</code></p>

</dd>
<dt id="SendmailBounceArguments"><code>$SendmailBounceArguments</code></dt>
<dd>

<p><code>$SendmailBounceArguments</code> defines what flags to pass to <code>$Sendmail</code> assuming RT needs to send an error (i.e. bounce).</p>

</dd>
<dt id="SendmailPath"><code>$SendmailPath</code></dt>
<dd>

<p>If you selected &#39;sendmailpipe&#39; above, you MUST specify the path to your sendmail binary in <code>$SendmailPath</code>.</p>

</dd>
</dl>

<h2 id="Other-mailers"><a href="#___top">Other mailers</a></h2>

<dl>

<dt id="MailParams"><code>@MailParams</code></dt>
<dd>

<p><code>@MailParams</code> defines a list of options passed to $MailCommand if it is not &#39;sendmailpipe&#39; or &#39;sendmail&#39;;</p>

</dd>
</dl>

<h1 id="Web-interface"><a href="#___top">Web interface</a></h1>

<dl>

<dt id="WebDefaultStylesheet"><code>$WebDefaultStylesheet</code></dt>
<dd>

<p>This determines the default stylesheet the RT web interface will use. RT ships with several themes by default:</p>

<pre><code>  rudder          The default theme for RT 4.2
  aileron         The default layout for RT 4.0
  web2            The default layout for RT 3.8
  ballard         Theme which doesn&#39;t rely on JavaScript for menuing</code></pre>

<p>This value actually specifies a directory in <i>share/static/css/</i> from which RT will try to load the file main.css (which should @import any other files the stylesheet needs). This allows you to easily and cleanly create your own stylesheets to apply to RT. This option can be overridden by users in their preferences.</p>

</dd>
<dt id="DefaultQueue"><code>$DefaultQueue</code></dt>
<dd>

<p>Use this to select the default queue name that will be used for creating new tickets. You may use either the queue&#39;s name or its ID. This only affects the queue selection boxes on the web interface.</p>

</dd>
<dt id="RememberDefaultQueue"><code>$RememberDefaultQueue</code></dt>
<dd>

<p>When a queue is selected in the new ticket dropdown, make it the new default for the new ticket dropdown.</p>

</dd>
<dt id="EnableReminders"><code>$EnableReminders</code></dt>
<dd>

<p>Hide all links and portlets related to Reminders by setting this to 0</p>

</dd>
<dt id="CustomFieldValuesSources"><code>@CustomFieldValuesSources</code></dt>
<dd>

<p>Set <code>@CustomFieldValuesSources</code> to a list of class names which extend <a href="RT/CustomFieldValues/External.html">RT::CustomFieldValues::External</a>. This can be used to pull lists of custom field values from external sources at runtime.</p>

</dd>
<dt id="CustomFieldValuesCanonicalizers"><code>@CustomFieldValuesCanonicalizers</code></dt>
<dd>

<p>Set <code>@CustomFieldValuesCanonicalizers</code> to a list of class names which extend <a href="RT/CustomFieldValues/Canonicalizer.html">RT::CustomFieldValues::Canonicalizer</a>. This can be used to rewrite (canonicalize) values entered by users to fit some defined format.</p>

<p>See the documentation in <a href="RT/CustomFieldValues/Canonicalizer.html">RT::CustomFieldValues::Canonicalizer</a> for adding your own canonicalizers.</p>

</dd>
<dt id="CustomFieldGroupings"><code>%CustomFieldGroupings</code></dt>
<dd>

<p>This option affects the display of ticket, user, and asset custom fields in the web interface. It does not address the sorting of custom fields within the groupings; that ordering is controlled by the Ticket Custom Fields tab in Queue configuration in the Admin UI. Asset custom field ordering is found in the Asset Custom Fields tab in Catalog configuration.</p>

<p>A nested data structure defines how to group together custom fields under a mix of built-in and arbitrary headings (&quot;groupings&quot;).</p>

<p>Set <code>%CustomFieldGroupings</code> to a nested structure similar to the following:</p>

<pre><code>    Set(%CustomFieldGroupings,
        &#39;RT::Ticket&#39; =&gt; [
            &#39;Grouping Name&#39;     =&gt; [&#39;CF Name&#39;, &#39;Another CF&#39;],
            &#39;Another Grouping&#39;  =&gt; [&#39;Some CF&#39;],
            &#39;Dates&#39;             =&gt; [&#39;Shipped date&#39;],
        ],
        &#39;RT::User&#39; =&gt; [
            &#39;Phones&#39; =&gt; [&#39;Fax number&#39;],
        ],
        &#39;RT::Asset&#39; =&gt; [
            &#39;Asset Details&#39; =&gt; [&#39;Serial Number&#39;, &#39;Manufacturer&#39;, &#39;Type&#39;, &#39;Tracking Number&#39;],
            &#39;Dates&#39;         =&gt; [&#39;Support Expiration&#39;, &#39;Issue Date&#39;],
        ],
        &#39;RT::Group&#39; =&gt; [
            &#39;Basics&#39; =&gt; [&#39;Department&#39;],
        ],
    );</code></pre>

<p>The first level keys are record types for which CFs may be used, and the values are either hashrefs or arrayrefs -- if arrayrefs, then the ordering is preserved during display, otherwise groupings are displayed alphabetically. The second level keys are the grouping names and the values are array refs containing a list of CF names.</p>

<p>There are several special built-in groupings which RT displays in specific places (usually the collapsible box of the same title). The ordering of these standard groupings cannot be modified. You may also only append Custom Fields to the list in these boxes, not reorder or remove core fields.</p>

<p>For <code>RT::Ticket</code>, these groupings are: <code>Basics</code>, <code>Dates</code>, <code>Links</code>, <code>People</code></p>

<p>For <code>RT::User</code>: <code>Identity</code>, <code>Access control</code>, <code>Location</code>, <code>Phones</code></p>

<p>For <code>RT::Group</code>: <code>Basics</code></p>

<p>For <code>RT::Asset</code>: <code>Basics</code>, <code>Dates</code>, <code>People</code>, <code>Links</code></p>

<p>Extensions may also add their own built-in groupings, refer to the individual extension documentation for those.</p>

</dd>
<dt id="CanonicalizeRedirectURLs"><code>$CanonicalizeRedirectURLs</code></dt>
<dd>

<p>Set <code>$CanonicalizeRedirectURLs</code> to 1 to use <code>$WebURL</code> when redirecting rather than the one we get from <code>%ENV</code>.</p>

<p>Apache&#39;s UseCanonicalName directive changes the hostname that RT finds in <code>%ENV</code>. You can read more about what turning it On or Off means in the documentation for your version of Apache.</p>

<p>If you use RT behind a reverse proxy, you almost certainly want to enable this option.</p>

</dd>
<dt id="CanonicalizeURLsInFeeds"><code>$CanonicalizeURLsInFeeds</code></dt>
<dd>

<p>Set <code>$CanonicalizeURLsInFeeds</code> to 1 to use <code>$WebURL</code> in feeds rather than the one we get from request.</p>

<p>If you use RT behind a reverse proxy, you almost certainly want to enable this option.</p>

</dd>
<dt id="JSFiles"><code>@JSFiles</code></dt>
<dd>

<p>A list of additional JavaScript files to be included in head.</p>

</dd>
<dt id="CSSFiles"><code>@CSSFiles</code></dt>
<dd>

<p>A list of additional CSS files to be included in head.</p>

<p>If you&#39;re a plugin author, refer to RT-&gt;AddStyleSheets.</p>

</dd>
<dt id="UsernameFormat"><code>$UsernameFormat</code></dt>
<dd>

<p>This determines how user info is displayed. &#39;concise&#39; will show the first of RealName, Name or EmailAddress that has a value. &#39;verbose&#39; will show EmailAddress, and the first of RealName or Name which is defined. The default, &#39;role&#39;, uses &#39;verbose&#39; for unprivileged users, and the Name followed by the RealName for privileged users.</p>

</dd>
<dt id="UserSearchResultFormat"><code>$UserSearchResultFormat</code></dt>
<dd>

<p>This controls the display of lists of users returned from the User Summary Search. The display of users in the Admin interface is controlled by <code>%AdminSearchResultFormat</code>.</p>

</dd>
<dt id="UserSummaryPortlets"><code>@UserSummaryPortlets</code></dt>
<dd>

<p>A list of portlets to be displayed on the User Summary page. By default, we show all of the available portlets. Extensions may provide their own portlets for this page.</p>

</dd>
<dt id="UserSummaryExtraInfo"><code>$UserSummaryExtraInfo</code></dt>
<dd>

<p>This controls what information is displayed on the User Summary portal. By default the user&#39;s Real Name, Email Address and Username are displayed. You can remove these or add more as needed. This expects a Format string of user attributes. Please note that not all the attributes are supported in this display because we&#39;re not building a table.</p>

</dd>
<dt id="UserSummaryTicketListFormat"><code>$UserSummaryTicketListFormat</code></dt>
<dd>

<p>Control the appearance of the Active and Inactive ticket lists in the User Summary.</p>

</dd>
<dt id="WebBaseURL-WebURL"><code>$WebBaseURL</code>, <code>$WebURL</code></dt>
<dd>

<p>Usually you don&#39;t want to set these options. The only obvious reason is if RT is accessible via https protocol on a non standard port, e.g. &#39;https://rt.example.com:9999&#39;. In all other cases these options are computed using <code>$WebDomain</code>, <code>$WebPort</code> and <code>$WebPath</code>.</p>

<p><code>$WebBaseURL</code> is the scheme, server and port (e.g. &#39;http://rt.example.com&#39;) for constructing URLs to the web UI. <code>$WebBaseURL</code> doesn&#39;t need a trailing /.</p>

<p><code>$WebURL</code> is the <code>$WebBaseURL</code>, <code>$WebPath</code> and trailing /, for example: &#39;http://www.example.com/rt/&#39;.</p>

</dd>
<dt id="WebImagesURL"><code>$WebImagesURL</code></dt>
<dd>

<p><code>$WebImagesURL</code> points to the base URL where RT can find its images. Define the directory name to be used for images in RT web documents.</p>

</dd>
<dt id="LogoURL"><code>$LogoURL</code></dt>
<dd>

<p><code>$LogoURL</code> points to the URL of the RT logo displayed in the web UI. This can also be configured via the web UI.</p>

</dd>
<dt id="LogoLinkURL"><code>$LogoLinkURL</code></dt>
<dd>

<p><code>$LogoLinkURL</code> is the URL that the RT logo hyperlinks to.</p>

</dd>
<dt id="LogoAltText"><code>$LogoAltText</code></dt>
<dd>

<p><code>$LogoAltText</code> is a string of text for the alt-text of the logo. It will be passed through <code>loc</code> for localization.</p>

</dd>
<dt id="WebNoAuthRegex"><code>$WebNoAuthRegex</code></dt>
<dd>

<p>What portion of RT&#39;s URL space should not require authentication. The default is almost certainly correct, and should only be changed if you are extending RT.</p>

</dd>
<dt id="WebFlushDbCacheEveryRequest"><code>$WebFlushDbCacheEveryRequest</code></dt>
<dd>

<p>By default, RT clears its database cache after every page view. This ensures that you&#39;ve always got the most current information when working in a multi-process (mod_perl or FastCGI) Environment. Setting <code>$WebFlushDbCacheEveryRequest</code> to 0 will turn this off, which will speed RT up a bit, at the expense of a tiny bit of data accuracy.</p>

</dd>
<dt id="ChartFont"><code>%ChartFont</code></dt>
<dd>

<p>The <a href="http://metacpan.org/module/GD">GD</a> module (which RT uses for graphs) ships with a built-in font that doesn&#39;t have full Unicode support. You can use a given TrueType font for a specific language by setting %ChartFont to (language =&gt; the absolute path of a font) pairs. Your GD library must have support for TrueType fonts to use this option. If there is no entry for a language in the hash then font with &#39;others&#39; key is used.</p>

<p>RT comes with two TrueType fonts covering most available languages.</p>

</dd>
<dt id="ChartsTimezonesInDB"><code>$ChartsTimezonesInDB</code></dt>
<dd>

<p>RT stores dates using the UTC timezone in the DB, so charts grouped by dates and time are not representative. Set <code>$ChartsTimezonesInDB</code> to 1 to enable timezone conversions using your DB&#39;s capabilities. You may need to do some work on the DB side to use this feature, read more in <i><a href="customizing/timezones_in_charts.html">docs/customizing/timezones_in_charts.pod</a></i>.</p>

<p>At this time, this feature only applies to MySQL and PostgreSQL.</p>

</dd>
<dt id="ChartColors"><code>@ChartColors</code></dt>
<dd>

<p>An array of 6-digit hexadecimal RGB color values used for chart series. By default there are 12 distinct colors.</p>

</dd>
</dl>

<h2 id="Home-page"><a href="#___top">Home page</a></h2>

<dl>

<dt id="DefaultSummaryRows"><code>$DefaultSummaryRows</code></dt>
<dd>

<p><code>$DefaultSummaryRows</code> is default number of rows displayed in for search results on the front page.</p>

</dd>
<dt id="RefreshIntervals"><code>@RefreshIntervals</code></dt>
<dd>

<p>This setting defines the possible homepage and search result refresh options. Each value is a number of seconds. You should not include a value of <code>0</code>, as that is always provided as an option.</p>

<p>See also <a href="#HomePageRefreshInterval">&quot;$HomePageRefreshInterval&quot;</a> and <a href="#SearchResultsRefreshInterval">&quot;$SearchResultsRefreshInterval&quot;</a>.</p>

</dd>
<dt id="HomePageRefreshInterval"><code>$HomePageRefreshInterval</code></dt>
<dd>

<p><code>$HomePageRefreshInterval</code> is default number of seconds to refresh the RT home page. Choose from any value in <a href="#RefreshIntervals">&quot;@RefreshIntervals&quot;</a>, or the default of <code>0</code> for no automatic refresh.</p>

</dd>
<dt id="HomepageComponents"><code>$HomepageComponents</code></dt>
<dd>

<p><code>$HomepageComponents</code> is an arrayref of allowed components on a user&#39;s customized homepage (&quot;RT at a glance&quot;).</p>

</dd>
</dl>

<h2 id="Ticket-search"><a href="#___top">Ticket search</a></h2>

<dl>

<dt id="UseSQLForACLChecks"><code>$UseSQLForACLChecks</code></dt>
<dd>

<p>Historically, ACLs were checked on display, which could lead to empty search pages and wrong ticket counts. Set <code>$UseSQLForACLChecks</code> to 0 to go back to this method; this will reduce the complexity of the generated SQL statements, at the cost of the aforementioned bugs.</p>

</dd>
<dt id="TicketsItemMapSize"><code>$TicketsItemMapSize</code></dt>
<dd>

<p>On the display page of a ticket from search results, RT provides links to the first, next, previous and last ticket from the results. In order to build these links, RT needs to fetch the full result set from the database, which can be resource-intensive.</p>

<p>Set <code>$TicketsItemMapSize</code> to number of tickets you want RT to examine to build these links. If the full result set is larger than this number, RT will omit the &quot;last&quot; link in the menu. Set this to zero to always examine all results.</p>

</dd>
<dt id="SearchResultsRefreshInterval"><code>$SearchResultsRefreshInterval</code></dt>
<dd>

<p><code>$SearchResultsRefreshInterval</code> is default number of seconds to refresh search results in RT. Choose from any value in <a href="#RefreshIntervals">&quot;@RefreshIntervals&quot;</a>, or the default of <code>0</code> for no automatic refresh.</p>

</dd>
<dt id="DefaultSearchResultFormat"><code>$DefaultSearchResultFormat</code></dt>
<dd>

<p><code>$DefaultSearchResultFormat</code> is the default format for RT search results</p>

</dd>
<dt id="DefaultSearchResultOrderBy"><code>$DefaultSearchResultOrderBy</code></dt>
<dd>

<p>What Tickets column should we order by for RT Ticket search results.</p>

</dd>
<dt id="DefaultSearchResultOrder"><code>$DefaultSearchResultOrder</code></dt>
<dd>

<p>When ordering RT Ticket search results by <code>$DefaultSearchResultOrderBy</code>, should the sort be ascending (ASC) or descending (DESC).</p>

</dd>
<dt id="ShowSearchResultCount"><code>$ShowSearchResultCount</code></dt>
<dd>

<p>Display search result count on ticket lists. Defaults to 0 (hide them).</p>

</dd>
<dt id="FullTextSearch"><code>%FullTextSearch</code></dt>
<dd>

<p>Full text search (FTS) without database indexing is a very slow operation, and is thus disabled by default.</p>

<p>Before setting <code>Indexed</code> to 1, read <i><a href="full_text_indexing.html">docs/full_text_indexing.pod</a></i> for the full details of FTS on your particular database.</p>

<p>It is possible to enable FTS without database indexing support, simply by setting the <code>Enable</code> key to 1, while leaving <code>Indexed</code> set to 0. This is not generally suggested, as unindexed full-text searching can cause severe performance problems.</p>

</dd>
<dt id="MaxFulltextAttachmentSize"><code>$MaxFulltextAttachmentSize</code></dt>
<dd>

<p>On some systems, very large attachments can cause memory and other performance issues for the indexer making it unable to complete indexing. Adding resources like memory and CPU will solve this issue, but in cases where that isn&#39;t possible, this option sets a maximum size in bytes on attachments to index. Attachments larger than this limit are skipped and will not be available to full text searches.</p>

</dd>
<dt id="DontSearchFileAttachments"><code>$DontSearchFileAttachments</code></dt>
<dd>

<p>If <code>$DontSearchFileAttachments</code> is set to 1, then uploaded files (attachments with file names) are not searched during content search.</p>

<p>Note that if you use indexed FTS then named attachments are still indexed by default regardless of this option.</p>

</dd>
<dt id="OnlySearchActiveTicketsInSimpleSearch"><code>$OnlySearchActiveTicketsInSimpleSearch</code></dt>
<dd>

<p>When query in simple search doesn&#39;t have status info, use this to only search active ones.</p>

</dd>
<dt id="SearchResultsAutoRedirect"><code>$SearchResultsAutoRedirect</code></dt>
<dd>

<p>When only one ticket is found in search, use this to redirect to the ticket display page automatically.</p>

</dd>
</dl>

<h2 id="Ticket-display"><a href="#___top">Ticket display</a></h2>

<dl>

<dt id="ShowMoreAboutPrivilegedUsers"><code>$ShowMoreAboutPrivilegedUsers</code></dt>
<dd>

<p>This determines if the &#39;More about requestor&#39; box on Ticket/Display.html is shown for Privileged Users.</p>

</dd>
<dt id="MoreAboutRequestorTicketList"><code>$MoreAboutRequestorTicketList</code></dt>
<dd>

<p>This can be set to Active, Inactive, All or None. It controls what ticket list will be displayed in the &#39;More about requestor&#39; box on Ticket/Display.html. This option can be controlled by users also.</p>

</dd>
<dt id="MoreAboutRequestorTicketListFormat"><code>$MoreAboutRequestorTicketListFormat</code></dt>
<dd>

<p>Control the appearance of the ticket lists in the &#39;More About Requestors&#39; box.</p>

</dd>
<dt id="MoreAboutRequestorExtraInfo"><code>$MoreAboutRequestorExtraInfo</code></dt>
<dd>

<p>By default, the &#39;More about requestor&#39; box on Ticket/Display.html shows the Requestor&#39;s name and ticket list. If you would like to see extra information about the user, this expects a Format string of user attributes. Please note that not all the attributes are supported in this display because we&#39;re not building a table.</p>

<p>Example: <code>Set($MoreAboutRequestorExtraInfo,&quot;Organization, Address1&quot;)</code></p>

</dd>
<dt id="MoreAboutRequestorGroupsLimit"><code>$MoreAboutRequestorGroupsLimit</code></dt>
<dd>

<p>By default, the &#39;More about requestor&#39; box on Ticket/Display.html shows all the groups of the Requestor. Use this to limit the number of groups; a value of undef removes the group display entirely.</p>

</dd>
<dt id="UseSideBySideLayout"><code>$UseSideBySideLayout</code></dt>
<dd>

<p>Should the ticket create and update forms use a more space efficient two column layout. This layout may not work in narrow browsers if you set a MessageBoxWidth (below).</p>

</dd>
<dt id="EditCustomFieldsSingleColumn"><code>$EditCustomFieldsSingleColumn</code></dt>
<dd>

<p>When displaying a list of Ticket Custom Fields for editing, RT defaults to a 2 column list. If you set this to 1, it will instead display the Custom Fields in a single column.</p>

</dd>
<dt id="ShowUnreadMessageNotifications"><code>$ShowUnreadMessageNotifications</code></dt>
<dd>

<p>If set to 1, RT will prompt users when there are new, unread messages on tickets they are viewing.</p>

</dd>
<dt id="AutocompleteOwners"><code>$AutocompleteOwners</code></dt>
<dd>

<p>If set to 1, the owner drop-downs for ticket update/modify and the query builder are replaced by text fields that autocomplete. This can alleviate the sometimes huge owner list for installations where many users have the OwnTicket right.</p>

<p>Autocompleter is automatically turned on if list contains more than 50 users, but penalty of executing potentially slow query is still paid.</p>

<p>Drop down doesn&#39;t show unprivileged users. If your setup allows unprivileged to own ticket then you have to enable autocompleting.</p>

</dd>
<dt id="AutocompleteOwnersForSearch"><code>$AutocompleteOwnersForSearch</code></dt>
<dd>

<p>If set to 1, the owner drop-downs for the query builder are always replaced by text field that autocomplete and <code>$AutocompleteOwners</code> is ignored. Helpful when owners list is huge in the query builder.</p>

</dd>
<dt id="AutocompleteQueues"><code>$AutocompleteQueues</code></dt>
<dd>

<p>If set to 1, any queue drop-downs are replaced by text fields that autocomplete. This can alleviate the sometimes huge queue list for installations with many queues, and can also increase page load times in some cases. A user can override this setting as a personal preference.</p>

</dd>
<dt id="UserSearchFields"><code>$UserSearchFields</code></dt>
<dd>

<p>Used by the User Autocompleter as well as the User Search.</p>

<p>Specifies which fields of <a href="RT/User.html">RT::User</a> to match against and how to match each field when autocompleting users. Valid match methods are LIKE, STARTSWITH, ENDSWITH, =, and !=. Valid search fields are the core User fields, as well as custom fields, which are specified as &quot;CF.1234&quot; or &quot;CF.Name&quot;</p>

</dd>
<dt id="TicketAutocompleteFields"><code>$TicketAutocompleteFields</code></dt>
<dd>

<p>Specifies which fields of <a href="RT/Ticket.html">RT::Ticket</a> to match against and how to match each field when autocompleting users. Valid match methods are LIKE, STARTSWITH, ENDSWITH, <code>=</code>, and <code>!=</code>.</p>

<p>Not all Ticket fields are publically accessible and hence won&#39;t work for autocomplete unless you override their accessibility using a local overlay or a plugin. Out of the box the following fields are public: id, Subject.</p>

</dd>
<dt id="DisplayTicketAfterQuickCreate"><code>$DisplayTicketAfterQuickCreate</code></dt>
<dd>

<p>Enable this to redirect to the created ticket display page automatically when using QuickCreate.</p>

</dd>
<dt id="WikiImplicitLinks"><code>$WikiImplicitLinks</code></dt>
<dd>

<p>Support implicit links in WikiText custom fields? Setting this to 1 causes InterCapped or ALLCAPS words in WikiText fields to automatically become links to searches for those words. If used on Articles, it links to the Article with that name.</p>

</dd>
<dt id="PreviewScripMessages"><code>$PreviewScripMessages</code></dt>
<dd>

<p>Set <code>$PreviewScripMessages</code> to 1 if the scrips preview on the ticket reply page should include the content of the messages to be sent.</p>

</dd>
<dt id="SimplifiedRecipients"><code>$SimplifiedRecipients</code></dt>
<dd>

<p>If <code>$SimplifiedRecipients</code> is set, a simple list of who will receive <b>any</b> kind of mail will be shown on the ticket reply page, instead of a detailed breakdown by scrip.</p>

</dd>
<dt id="SquelchedRecipients"><code>$SquelchedRecipients</code></dt>
<dd>

<p>If <code>$SquelchedRecipients</code> is set, the checkbox list of who will receive <b>any</b> kind of mail on the ticket reply page are displayed initially as <b>un</b>checked - which means nobody in that list would get any mail. It does not affect correspondence done via email yet.</p>

</dd>
<dt id="HideResolveActionsWithDependencies"><code>$HideResolveActionsWithDependencies</code></dt>
<dd>

<p>If set to 1, this option will skip ticket menu actions which can&#39;t be completed successfully because of outstanding active Depends On tickets.</p>

<p>By default, all ticket actions are displayed in the menu even if some of them can&#39;t be successful until all Depends On links are resolved or transitioned to another inactive status.</p>

</dd>
<dt id="HideUnsetFieldsOnDisplay"><code>$HideUnsetFieldsOnDisplay</code></dt>
<dd>

<p>This determines if we should hide unset fields on ticket display page. Set this to 1 to hide unset fields.</p>

</dd>
</dl>

<h2 id="Self-Service-Interface"><a href="#___top">Self Service Interface</a></h2>

<p>The Self Service Interface is a view automatically presented to Unprivileged users who have a password and log into the web UI. The following options modify the default behavior of the Self Service pages.</p>

<dl>

<dt id="SelfServiceCorrespondenceOnly"><code>$SelfServiceCorrespondenceOnly</code></dt>
<dd>

<p>On the ticket display page, show only correspondence transactions in the ticket history. This hides all ticket update transactions like status changes, custom field updates, updates to watchers, etc.</p>

</dd>
<dt id="HideTimeFieldsFromUnprivilegedUsers"><code>$HideTimeFieldsFromUnprivilegedUsers</code></dt>
<dd>

<p>This determines if we should hide Time Worked, Time Estimated, and Time Left for unprivileged users. Set this to 1 to hide those fields.</p>

</dd>
<dt id="AllowUserAutocompleteForUnprivileged"><code>$AllowUserAutocompleteForUnprivileged</code></dt>
<dd>

<p>Should unprivileged users (users of SelfService) be allowed to autocomplete users. Setting this option to 1 means unprivileged users will be able to search all your users.</p>

</dd>
<dt id="DefaultSelfServiceSearchResultFormat"><code>$DefaultSelfServiceSearchResultFormat</code></dt>
<dd>

<p><code>$DefaultSelfServiceSearchResultFormat</code> is the default format of searches displayed in the SelfService interface.</p>

</dd>
<dt id="SelfServiceRegex"><code>$SelfServiceRegex</code></dt>
<dd>

<p>What portion of RT&#39;s URLspace should be accessible to Unprivileged users This does not override the redirect from <i>/Ticket/Display.html</i> to <i>/SelfService/Display.html</i> when Unprivileged users attempt to access ticked displays.</p>

</dd>
</dl>

<h2 id="Articles"><a href="#___top">Articles</a></h2>

<dl>

<dt id="ArticleOnTicketCreate"><code>$ArticleOnTicketCreate</code></dt>
<dd>

<p>Set this to 1 to display the Articles interface on the Ticket Create page in addition to the Reply/Comment page.</p>

</dd>
<dt id="HideArticleSearchOnReplyCreate"><code>$HideArticleSearchOnReplyCreate</code></dt>
<dd>

<p>Set this to 1 to hide the search and include boxes from the Article UI. This assumes you have enabled Article Hotlist feature, otherwise you will have no access to Articles.</p>

</dd>
<dt id="LinkArticlesOnInclude"><code>$LinkArticlesOnInclude</code></dt>
<dd>

<p>Set this to 0 to suppress the default behavior of automatically linking to Articles when they are included in a message.</p>

</dd>
</dl>

<h1 id="Assets"><a href="#___top">Assets</a></h1>

<dl>

<dt id="AssetQueues"><code>@AssetQueues</code></dt>
<dd>

<p>This should be a list of names of queues whose tickets should always display the &quot;Assets&quot; box. This is useful for queues which deal primarily with assets, as it provides a ready box to link an asset to the ticket, even when the ticket has no related assets yet.</p>

</dd>
<dt id="DefaultCatalog"><code>$DefaultCatalog</code></dt>
<dd>

<p>This provides the default catalog after a user initially logs in. However, the default catalog is &quot;sticky,&quot; and so will remember the last-selected catalog thereafter.</p>

</dd>
<dt id="AssetSearchFields"><code>$AssetSearchFields</code></dt>
<dd>

<p>Specifies which fields of <a href="RT/Asset.html">RT::Asset</a> to match against and how to match each field when performing a quick search on assets. Valid match methods are LIKE, STARTSWITH, ENDSWITH, =, and !=. Valid search fields are id, Name, Description, or custom fields, which are specified as &quot;CF.1234&quot; or &quot;CF.Name&quot;</p>

</dd>
<dt id="AssetSearchFormat"><code>$AssetSearchFormat</code></dt>
<dd>

<p>The format that results of the asset search are displayed with. This is either a string, which will be used for all catalogs, or a hash reference, keyed by catalog&#39;s name/id. If a hashref and neither name or id is found therein, falls back to the key &#39;&#39;.</p>

<p>If you wish to use the multiple catalog format, your configuration would look something like:</p>

<pre><code>    Set($AssetSearchFormat, {
        &#39;General assets&#39; =&gt; q[Format String for the General Assets Catalog],
        8                =&gt; q[Format String for Catalog 8],
        &#39;&#39;               =&gt; q[Format String for any catalogs not listed explicitly],
    });</code></pre>

</dd>
<dt id="AssetSummaryFormat"><code>$AssetSummaryFormat</code></dt>
<dd>

<p>The information that is displayed on ticket display pages about assets related to the ticket. This is displayed in a table beneath the asset name.</p>

</dd>
<dt id="AssetSummaryRelatedTicketsFormat"><code>$AssetSummaryRelatedTicketsFormat</code></dt>
<dd>

<p>The information that is displayed on ticket display pages about tickets related to assets related to the ticket. This is displayed as a list of tickets underneath the asset properties.</p>

</dd>
<dt id="AssetBasicCustomFieldsOnCreate"><code>$AssetBasicCustomFieldsOnCreate</code></dt>
<dd>

<p>Specify a list of Asset custom fields to show in &quot;Basics&quot; widget on create.</p>

<p>e.g.</p>

<p>Set( $AssetBasicCustomFieldsOnCreate, [ &#39;foo&#39;, &#39;bar&#39; ] );</p>

</dd>
</dl>

<h2 id="Message-box-properties"><a href="#___top">Message box properties</a></h2>

<dl>

<dt id="MessageBoxWidth-MessageBoxHeight"><code>$MessageBoxWidth</code>, <code>$MessageBoxHeight</code></dt>
<dd>

<p>For message boxes, set the entry box width, height and what type of wrapping to use. These options can be overridden by users in their preferences.</p>

<p>When the width is set to undef, no column count is specified and the message box will take up 100% of the available width. Combining this with HARD messagebox wrapping (below) is not recommended, as it will lead to inconsistent width in transactions between browsers.</p>

<p>These settings only apply to the non-RichText message box. See below for Rich Text settings.</p>

</dd>
<dt id="MessageBoxRichText"><code>$MessageBoxRichText</code></dt>
<dd>

<p>Should &quot;rich text&quot; editing be enabled? This option lets your users send HTML email messages from the web interface.</p>

</dd>
<dt id="MessageBoxUseSystemContextMenu"><code>$MessageBoxUseSystemContextMenu</code></dt>
<dd>

<p>Should the &quot;rich text&quot; editor use the browser&#39;s right-click menu?</p>

<p>The rich text editor&#39;s custom menu can block system functionality like spell check, dictionary lookup and browser extension options. Enabling this option allows the browser&#39;s native right click menu to be used; However, enabling this option may also prevent pasting rich text from other applications.</p>

</dd>
<dt id="MessageBoxRichTextHeight"><code>$MessageBoxRichTextHeight</code></dt>
<dd>

<p>Height of rich text JavaScript enabled editing boxes (in pixels)</p>

</dd>
<dt id="MessageBoxIncludeSignature"><code>$MessageBoxIncludeSignature</code></dt>
<dd>

<p>Should your users&#39; signatures (from their Preferences page) be included in Comments and Replies.</p>

</dd>
<dt id="MessageBoxIncludeSignatureOnComment"><code>$MessageBoxIncludeSignatureOnComment</code></dt>
<dd>

<p>Should your users&#39; signatures (from their Preferences page) be included in Comments. Setting this to 0 overrides <code>$MessageBoxIncludeSignature</code>.</p>

</dd>
</dl>

<h2 id="Transaction-display"><a href="#___top">Transaction display</a></h2>

<dl>

<dt id="OldestTransactionsFirst"><code>$OldestTransactionsFirst</code></dt>
<dd>

<p>By default, RT shows newest transactions at the bottom of the ticket history page, if you want see them at the top set this to 0. This option can be overridden by users in their preferences.</p>

</dd>
<dt id="ShowHistory"><code>$ShowHistory</code></dt>
<dd>

<p>This option controls how history is shown on the ticket display page. It accepts one of three possible modes and is overrideable on a per-user preference level. If you regularly deal with long tickets and don&#39;t care much about the history, you may wish to change this option to <code>click</code>.</p>

<dl>

<dt id="delay-the-default"><code>delay</code> (the default)</dt>
<dd>

<p>When set to <code>delay</code>, history is loaded via javascript after the rest of the page has been loaded. This speeds up apparent page load times and generally provides a smoother experience. You may notice slight delays before the ticket history appears on very long tickets.</p>

</dd>
<dt id="click"><code>click</code></dt>
<dd>

<p>When set to <code>click</code>, history is loaded on demand when a placeholder link is clicked. This speeds up ticket display page loads and history is never loaded if not requested.</p>

</dd>
<dt id="always"><code>always</code></dt>
<dd>

<p>When set to <code>always</code>, history is loaded before showing the page. This ensures history is always available immediately, but at the expense of longer page load times. This behaviour was the default in RT 4.0.</p>

</dd>
<dt id="scroll"><code>scroll</code></dt>
<dd>

<p>When set to <code>scroll</code>, history is loaded via javascript after the rest of the page has been loaded, as you scroll down the page. Ten transactions are loaded initially, and then more are loaded ten at a time. This can dramatically speed up initial page load times on very long tickets.</p>

</dd>
</dl>

</dd>
<dt id="ShowBccHeader"><code>$ShowBccHeader</code></dt>
<dd>

<p>By default, RT hides from the web UI information about blind copies user sent on reply or comment.</p>

</dd>
<dt id="TrustHTMLAttachments"><code>$TrustHTMLAttachments</code></dt>
<dd>

<p>If <code>TrustHTMLAttachments</code> is not defined, we will display them as text. This prevents malicious HTML and JavaScript from being sent in a request (although there is probably more to it than that)</p>

</dd>
<dt id="AlwaysDownloadAttachments"><code>$AlwaysDownloadAttachments</code></dt>
<dd>

<p>Always download attachments, regardless of content type. If set, this overrides <code>TrustHTMLAttachments</code>.</p>

</dd>
<dt id="AttachmentListCount"><code>$AttachmentListCount</code></dt>
<dd>

<p>The number of attachments to display by default on ticket display and ticket reply pages. Attachments beyond this count will be displayed only after the user clicks a &quot;Show all&quot; link. The default value, <code>undef</code>, means always show all attachments. A value of <code>0</code> means show no attachments by default.</p>

</dd>
<dt id="PreferRichText"><code>$PreferRichText</code></dt>
<dd>

<p>By default, RT shows rich text (HTML) messages if possible. If <code>$PreferRichText</code> is set to 0, RT will show plain text messages in preference to any rich text alternatives.</p>

<p>As a security precaution, RT limits the HTML that is displayed to a known-good subset -- as allowing arbitrary HTML to be displayed exposes multiple vectors for XSS and phishing attacks. If <a href="#TrustHTMLAttachments">&quot;$TrustHTMLAttachments&quot;</a> is enabled, the original HTML is available for viewing via the &quot;Download&quot; link.</p>

<p>If the optional <a href="http://metacpan.org/module/HTML::Gumbo">HTML::Gumbo</a> dependency is installed, RT will leverage this to allow a broader set of HTML through, including tables.</p>

</dd>
<dt id="MaxInlineBody"><code>$MaxInlineBody</code></dt>
<dd>

<p><code>$MaxInlineBody</code> is the maximum textual attachment size that we want to see inline when viewing a transaction. RT will inline any text if the value is undefined or 0. This option can be overridden by users in their preferences. The default is 25k.</p>

</dd>
<dt id="ShowTransactionImages"><code>$ShowTransactionImages</code></dt>
<dd>

<p>By default, RT shows images attached to incoming (and outgoing) ticket updates inline. Set this variable to 0 if you&#39;d like to disable that behavior.</p>

</dd>
<dt id="ShowRemoteImages"><code>$ShowRemoteImages</code></dt>
<dd>

<p>By default, RT doesn&#39;t show remote images attached to incoming (and outgoing) ticket updates inline. Set this variable to 1 if you&#39;d like to enable remote image display. Showing remote images may allow spammers and other senders to track when messages are viewed and see referer information.</p>

<p>Note that this setting is independent of <a href="#ShowTransactionImages">&quot;$ShowTransactionImages&quot;</a> above.</p>

</dd>
<dt id="PlainTextMono"><code>$PlainTextMono</code></dt>
<dd>

<p>Normally plaintext attachments are displayed as HTML with line breaks preserved. This causes space- and tab-based formatting not to be displayed correctly. Set <code>$PlainTextMono</code> to 1 to use a monospaced font and preserve formatting.</p>

</dd>
<dt id="SuppressInlineTextFiles"><code>$SuppressInlineTextFiles</code></dt>
<dd>

<p>If <code>$SuppressInlineTextFiles</code> is set to 1, then uploaded text files (text-type attachments with file names) are prevented from being displayed in-line when viewing a ticket&#39;s history.</p>

</dd>
<dt id="Active_MakeClicky"><code>@Active_MakeClicky</code></dt>
<dd>

<p>MakeClicky detects various formats of data in headers and email messages, and extends them with supporting links. By default, RT provides two formats:</p>

<p>* &#39;httpurl&#39;: detects http:// and https:// URLs and adds &#39;[Open URL]&#39; link after the URL.</p>

<p>* &#39;httpurl_overwrite&#39;: also detects URLs as &#39;httpurl&#39; format, but replaces the URL with a link. Enabled by default.</p>

<p>See <i>share/html/Elements/MakeClicky</i> for documentation on how to add your own styles of link detection.</p>

</dd>
<dt id="QuoteFolding"><code>$QuoteFolding</code></dt>
<dd>

<p>Quote folding is the hiding of old replies in transaction history. It defaults to on. Set this to 0 to disable it.</p>

</dd>
</dl>

<h1 id="Application-logic"><a href="#___top">Application logic</a></h1>

<dl>

<dt id="ParseNewMessageForTicketCcs"><code>$ParseNewMessageForTicketCcs</code></dt>
<dd>

<p>If <code>$ParseNewMessageForTicketCcs</code> is set to 1, RT will attempt to divine Ticket &#39;Cc&#39; watchers from the To and Cc lines of incoming messages that create new Tickets. This option does not apply to replies or comments on existing Tickets. Be forewarned that if you have <i>any</i> addresses which forward mail to RT automatically and you enable this option without modifying <code>$RTAddressRegexp</code> below, you will get yourself into a heap of trouble.</p>

</dd>
<dt id="UseTransactionBatch"><code>$UseTransactionBatch</code></dt>
<dd>

<p>Set <code>$UseTransactionBatch</code> to 1 to execute transactions in batches, such that a resolve and comment (for example) would happen simultaneously, instead of as two transactions, unaware of each others&#39; existence.</p>

</dd>
<dt id="StrictLinkACL"><code>$StrictLinkACL</code></dt>
<dd>

<p>When this feature is enabled a user needs <i>ModifyTicket</i> rights on both tickets to link them together; otherwise, <i>ModifyTicket</i> rights on either of them is sufficient.</p>

</dd>
<dt id="RedistributeAutoGeneratedMessages"><code>$RedistributeAutoGeneratedMessages</code></dt>
<dd>

<p>Should RT redistribute correspondence that it identifies as machine generated? A 1 will do so; setting this to 0 will cause no such messages to be redistributed. You can also use &#39;privileged&#39; (the default), which will redistribute only to privileged users. This helps to protect against malformed bounces and loops caused by auto-created requestors with bogus addresses.</p>

</dd>
<dt id="ApprovalRejectionNotes"><code>$ApprovalRejectionNotes</code></dt>
<dd>

<p>Should rejection notes from approvals be sent to the requestors?</p>

</dd>
<dt id="ForceApprovalsView"><code>$ForceApprovalsView</code></dt>
<dd>

<p>Should approval tickets only be viewed and modified through the standard approval interface? With this setting enabled (by default), any attempt to use the normal ticket display and modify page for approval tickets will be redirected.</p>

<p>For example, with this option set to 1 and an approval ticket #123:</p>

<pre><code>    /Ticket/Display.html?id=123</code></pre>

<p>is redirected to</p>

<pre><code>    /Approval/Display.html?id=123</code></pre>

<p>With this option set to 0, the redirect won&#39;t happen.</p>

</dd>
</dl>

<h1 id="Extra-security"><a href="#___top">Extra security</a></h1>

<p>This is a list of extra security measures to enable that help keep your RT safe. If you don&#39;t know what these mean, you should almost certainly leave the defaults alone.</p>

<dl>

<dt id="DisallowExecuteCode"><code>$DisallowExecuteCode</code></dt>
<dd>

<p>If set to 1, the <code>ExecuteCode</code> right will be removed from all users, <b>including</b> the superuser. This is intended for when RT is installed into a shared environment where even the superuser should not be allowed to run arbitrary Perl code on the server via scrips.</p>

</dd>
<dt id="Framebusting"><code>$Framebusting</code></dt>
<dd>

<p>If set to 0, framekiller javascript will be disabled and the X-Frame-Options: DENY header will be suppressed from all responses. This disables RT&#39;s clickjacking protection.</p>

</dd>
<dt id="RestrictReferrer"><code>$RestrictReferrer</code></dt>
<dd>

<p>If set to 0, the HTTP <code>Referer</code> (sic) header will not be checked to ensure that requests come from RT&#39;s own domain. As RT allows for GET requests to alter state, disabling this opens RT up to cross-site request forgery (CSRF) attacks.</p>

</dd>
<dt id="RestrictLoginReferrer"><code>$RestrictLoginReferrer</code></dt>
<dd>

<p>If set to 0, RT will allow the user to log in from any link or request, merely by passing in <code>user</code> and <code>pass</code> parameters; setting it to 1 forces all logins to come from the login box, so the user is aware that they are being logged in. The default is off, for backwards compatability.</p>

</dd>
<dt id="ReferrerWhitelist"><code>@ReferrerWhitelist</code></dt>
<dd>

<p>This is a list of hostname:port combinations that RT will treat as being part of RT&#39;s domain. This is particularly useful if you access RT as multiple hostnames or have an external auth system that needs to redirect back to RT once authentication is complete.</p>

<pre><code> Set(@ReferrerWhitelist, qw(www.example.com:443  www3.example.com:80));</code></pre>

<p>If the &quot;RT has detected a possible cross-site request forgery&quot; error is triggered by a host:port sent by your browser that you believe should be valid, you can copy the host:port from the error message into this list.</p>

<p>Simple wildcards, similar to SSL certificates, are allowed. For example:</p>

<pre><code>    *.example.com:80    # matches foo.example.com
                        # but not example.com
                        #      or foo.bar.example.com

    www*.example.com:80 # matches www3.example.com
                        #     and www-test.example.com
                        #     and www.example.com</code></pre>

</dd>
<dt id="ReferrerComponents"><code>%ReferrerComponents</code></dt>
<dd>

<p><code>%ReferrerComponents</code> is the hash to customize referrer checking behavior when <code>$RestrictReferrer</code> is enabled, where you can whitelist or blacklist the components along with their query args. e.g.</p>

<pre><code>    Set( %ReferrerComponents,
        ( &#39;/Foo.html&#39; =&gt; 1, &#39;/Bar.html&#39; =&gt; 0, &#39;/Baz.html&#39; =&gt; [ &#39;id&#39;, &#39;results&#39; ] )
    );</code></pre>

<p>With this, &#39;/Foo.html&#39; will be whitelisted, and &#39;/Bar.html&#39; will be blacklisted. &#39;/Baz.html&#39; with id/results query arguments will be whitelisted but blacklisted if there are other query arguments.</p>

</dd>
<dt id="BcryptCost"><code>$BcryptCost</code></dt>
<dd>

<p>This sets the default cost parameter used for the <code>bcrypt</code> key derivation function. Valid values range from 4 to 31, inclusive, with higher numbers denoting greater effort.</p>

</dd>
</dl>

<h1 id="Authorization-and-user-configuration"><a href="#___top">Authorization and user configuration</a></h1>

<dl>

<dt id="WebRemoteUserAuth"><code>$WebRemoteUserAuth</code></dt>
<dd>

<p>If <code>$WebRemoteUserAuth</code> is defined, RT will defer to the environment&#39;s REMOTE_USER variable, which should be set by the webserver&#39;s authentication layer.</p>

</dd>
<dt id="WebRemoteUserContinuous"><code>$WebRemoteUserContinuous</code></dt>
<dd>

<p>If <code>$WebRemoteUserContinuous</code> is defined, RT will check for the REMOTE_USER on each access. If you would prefer this to only happen once (at initial login) set this to 0. The default setting will help ensure that if your webserver&#39;s authentication layer deauthenticates a user, RT notices as soon as possible.</p>

</dd>
<dt id="WebFallbackToRTLogin"><code>$WebFallbackToRTLogin</code></dt>
<dd>

<p>If <code>$WebFallbackToRTLogin</code> is defined, the user is allowed a chance of fallback to the login screen, even if REMOTE_USER failed.</p>

</dd>
<dt id="WebRemoteUserGecos"><code>$WebRemoteUserGecos</code></dt>
<dd>

<p><code>$WebRemoteUserGecos</code> means to match &#39;gecos&#39; field as the user identity; useful with <code>mod_auth_external</code>.</p>

</dd>
<dt id="WebRemoteUserAutocreate"><code>$WebRemoteUserAutocreate</code></dt>
<dd>

<p><code>$WebRemoteUserAutocreate</code> will create users under the same name as REMOTE_USER upon login, if they are missing from the Users table.</p>

</dd>
<dt id="UserAutocreateDefaultsOnLogin"><code>$UserAutocreateDefaultsOnLogin</code></dt>
<dd>

<p>If <code>$WebRemoteUserAutocreate</code> is set to 1, <code>$UserAutocreateDefaultsOnLogin</code> will be passed to <a href="RT/User.html#Create">&quot;Create&quot; in RT::User</a>. Use it to set defaults, such as creating unprivileged users with <code>&lt;{ Privileged =</code> 0 }&gt;&gt;. This must be a hashref.</p>

</dd>
<dt id="WebSessionClass"><code>$WebSessionClass</code></dt>
<dd>

<p><code>$WebSessionClass</code> is the class you wish to use for storing sessions. On MySQL, Pg, and Oracle it defaults to using your database, in other cases sessions are stored in files using <a href="http://metacpan.org/module/Apache::Session::File">Apache::Session::File</a>. Other installed Apache::Session::* modules can be used to store sessions.</p>

<pre><code>    Set($WebSessionClass, &quot;Apache::Session::File&quot;);</code></pre>

</dd>
<dt id="WebSessionProperties"><code>%WebSessionProperties</code></dt>
<dd>

<p><code>%WebSessionProperties</code> is the hash to configure class <a href="#WebSessionClass">&quot;$WebSessionClass&quot;</a> in case custom class is used. By default it&#39;s empty and values are picked depending on the class. Make sure that it&#39;s empty if you&#39;re using DB as session backend.</p>

</dd>
<dt id="AutoLogoff"><code>$AutoLogoff</code></dt>
<dd>

<p>By default, RT&#39;s user sessions persist until a user closes his or her browser. With the <code>$AutoLogoff</code> option you can setup session lifetime in minutes. A user will be logged out if he or she doesn&#39;t send any requests to RT for the defined time.</p>

</dd>
<dt id="LogoutRefresh"><code>$LogoutRefresh</code></dt>
<dd>

<p>The number of seconds to wait after logout before sending the user to the login page. By default, 1 second, though you may want to increase this if you display additional information on the logout page.</p>

</dd>
<dt id="WebSecureCookies"><code>$WebSecureCookies</code></dt>
<dd>

<p>By default, RT&#39;s session cookie isn&#39;t marked as &quot;secure&quot;. Some web browsers will treat secure cookies more carefully than non-secure ones, being careful not to write them to disk, only sending them over an SSL secured connection, and so on. To enable this behavior, set <code>$WebSecureCookies</code> to 1. NOTE: You probably don&#39;t want to turn this on <i>unless</i> users are only connecting via SSL encrypted HTTPS connections.</p>

</dd>
<dt id="WebHttpOnlyCookies"><code>$WebHttpOnlyCookies</code></dt>
<dd>

<p>Default RT&#39;s session cookie to not being directly accessible to javascript. The content is still sent during regular and AJAX requests, and other cookies are unaffected, but the session-id is less programmatically accessible to javascript. Turning this off should only be necessary in situations with odd client-side authentication requirements.</p>

</dd>
<dt id="MinimumPasswordLength"><code>$MinimumPasswordLength</code></dt>
<dd>

<p><code>$MinimumPasswordLength</code> defines the minimum length for user passwords. Setting it to 0 disables this check.</p>

</dd>
</dl>

<h2 id="External-Authentication-and-Authorization"><a href="#___top">External Authentication and Authorization</a></h2>

<p>RT has a built-in module for integrating with a directory service like LDAP or Active Directory for authentication (login) and authorization (enabling/disabling users and setting user attributes). The core configuration settings for the service are listed here. Additional details are available in the <a href="RT/Authen/ExternalAuth.html">RT::Authen::ExternalAuth</a> module documentation.</p>

<dl>

<dt id="ExternalSettings"><code>$ExternalSettings</code></dt>
<dd>

<p>This option, along with the following options, activate and configure authentication via a resource external to RT. All of the configuration for your external authentication service, like LDAP or Active Directory, are defined in a data structure in this option. You can find full details on the configuration options in the <a href="RT/Authen/ExternalAuth.html">RT::Authen::ExternalAuth</a> documentation.</p>

</dd>
<dt id="ExternalAuthPriority"><code>$ExternalAuthPriority</code></dt>
<dd>

<p>Sets the priority of authentication resources if you have multiple configured. RT will attempt authorization with each resource, in order, until one succeeds or no more remain. See <a href="RT/Authen/ExternalAuth.html">RT::Authen::ExternalAuth</a> for details.</p>

</dd>
<dt id="ExternalInfoPriority"><code>$ExternalInfoPriority</code></dt>
<dd>

<p>Sets the order of resources for querying user information if you have multiple configured. RT will query each resource, in order, until one succeeds or no more remain. See <a href="RT/Authen/ExternalAuth.html">RT::Authen::ExternalAuth</a> for details.</p>

</dd>
<dt id="UserAutocreateDefaultsOnLogin1"><code>$UserAutocreateDefaultsOnLogin</code></dt>
<dd>

<p>A hashref of options to set for users who are autocreated on login via ExternalAuth. For example, you can automatically make &quot;Privileged&quot; users who were authenticated and created from LDAP or Active Directory. See <a href="RT/Authen/ExternalAuth.html">RT::Authen::ExternalAuth</a> for details.</p>

</dd>
<dt id="AutoCreateNonExternalUsers"><code>$AutoCreateNonExternalUsers</code></dt>
<dd>

<p>Users should still be autocreated by RT as internal users if they fail to exist in an external service; this is so requestors who are not in LDAP can still be created when they email in. See <a href="RT/Authen/ExternalAuth.html">RT::Authen::ExternalAuth</a> for details.</p>

</dd>
</dl>

<h2 id="Syncing-Users-and-Groups-with-LDAP-or-AD"><a href="#___top">Syncing Users and Groups with LDAP or AD</a></h2>

<p>In addition to the authentication services described above, RT also has a utility you can schedule to periodicially sync from your directory service additional user attributes, new users, disabled users, and group membership. Options for the LDAPImport tool are listed here. Additional information is available in the <a href="RT/LDAPImport.html">RT::LDAPImport</a> documentation.</p>

<dl>

<dt id="LDAPHost"><code>$LDAPHost</code></dt>
<dd>

<p>Your LDAP server hostname.</p>

</dd>
<dt id="LDAPUser"><code>$LDAPUser</code></dt>
<dd>

<p>The LDAP user to log in with.</p>

</dd>
<dt id="LDAPPassword"><code>$LDAPPassword</code></dt>
<dd>

<p>Password for LDAPUser.</p>

</dd>
<dt id="LDAPFilter"><code>$LDAPFilter</code></dt>
<dd>

<p>The filter to use when querying LDAP for the set of users to sync.</p>

</dd>
<dt id="LDAPMapping"><code>$LDAPMapping</code></dt>
<dd>

<p>Mapping to apply between LDAP attributes retrieved and RT user record attributes. See the <a href="RT/LDAPImport.html">RT::LDAPImport</a> documentation for details.</p>

</dd>
<dt id="LDAPGroupBase"><code>$LDAPGroupBase</code></dt>
<dd>

<p>The base for the LDAP group search.</p>

</dd>
<dt id="LDAPGroupFilter"><code>$LDAPGroupFilter</code></dt>
<dd>

<p>The filter to use when querying LDAP for groups to sync.</p>

</dd>
<dt id="LDAPGroupMapping"><code>$LDAPGroupMapping</code></dt>
<dd>

<p>Mapping to apply between LDAP group member attributes retrieved and RT groups. See the <a href="RT/LDAPImport.html">RT::LDAPImport</a> documentation for details.</p>

</dd>
</dl>

<h1 id="Internationalization"><a href="#___top">Internationalization</a></h1>

<dl>

<dt id="LexiconLanguages"><code>@LexiconLanguages</code></dt>
<dd>

<p>An array that contains languages supported by RT&#39;s internationalization interface. Defaults to all *.po lexicons; setting it to <code>qw(en ja)</code> will make RT bilingual instead of multilingual, but will save some memory.</p>

</dd>
<dt id="EmailInputEncodings"><code>@EmailInputEncodings</code></dt>
<dd>

<p>An array that contains default encodings used to guess which charset an attachment uses, if it does not specify one explicitly. All options must be recognized by <a href="http://metacpan.org/module/Encode::Guess">Encode::Guess</a>. The first element may also be &#39;*&#39;, which enables encoding detection using <a href="http://metacpan.org/module/Encode::Detect::Detector">Encode::Detect::Detector</a>, if installed.</p>

</dd>
<dt id="EmailOutputEncoding"><code>$EmailOutputEncoding</code></dt>
<dd>

<p>The charset for localized email. Must be recognized by Encode.</p>

</dd>
</dl>

<h1 id="Date-and-time-handling"><a href="#___top">Date and time handling</a></h1>

<dl>

<dt id="DateTimeFormat"><code>$DateTimeFormat</code></dt>
<dd>

<p>You can choose date and time format. See the &quot;Output formatters&quot; section in perldoc <i><a href="RT/Date.html">lib/RT/Date.pm</a></i> for more options. This option can be overridden by users in their preferences.</p>

<p>Some examples:</p>

<p><code>Set($DateTimeFormat, &quot;LocalizedDateTime&quot;);</code> <code>Set($DateTimeFormat, { Format =</code> &quot;ISO&quot;, Seconds =&gt; 0 });&gt; <code>Set($DateTimeFormat, &quot;RFC2822&quot;);</code> <code>Set($DateTimeFormat, { Format =</code> &quot;RFC2822&quot;, Seconds =&gt; 0, DayOfWeek =&gt; 0 });&gt;</p>

</dd>
<dt id="DateDayBeforeMonth"><code>$DateDayBeforeMonth</code></dt>
<dd>

<p>Set this to 1 if your local date convention looks like &quot;dd/mm/yy&quot; instead of &quot;mm/dd/yy&quot;. Used only for parsing, not for displaying dates.</p>

</dd>
<dt id="AmbiguousDayInPast-AmbiguousDayInFuture"><code>$AmbiguousDayInPast</code>, <code>$AmbiguousDayInFuture</code></dt>
<dd>

<p>Should an unspecified day or year in a date refer to a future or a past value? For example, should a date of &quot;Tuesday&quot; default to mean the date for next Tuesday or last Tuesday? Should the date &quot;March 1&quot; default to the date for next March or last March?</p>

<p>Set <code>$AmbiguousDayInPast</code> for the last date, or <code>$AmbiguousDayInFuture</code> for the next date; the default is usually correct. If both are set, <code>$AmbiguousDayInPast</code> takes precedence.</p>

</dd>
<dt id="DefaultTimeUnitsToHours"><code>$DefaultTimeUnitsToHours</code></dt>
<dd>

<p>Use this to set the default units for time entry to hours instead of minutes. Note that this only effects entry, not display.</p>

</dd>
<dt id="SignatureAboveQuote"><code>$SignatureAboveQuote</code></dt>
<dd>

<p>By default RT places the signature at the bottom of the quoted text in the message box for ticket replies. Set this to 1 to place the signature above the quoted text.</p>

</dd>
<dt id="TimeInICal"><code>$TimeInICal</code></dt>
<dd>

<p>By default, events in the iCal feed on the ticket search page contain only dates, making them all day calendar events. Set <code>$TimeInICal</code> if you have start or due dates on tickets that have significant time values and you want those times to be included in the events in the iCal feed.</p>

<p>This option can also be set as an individual user preference.</p>

</dd>
<dt id="PreferDateTimeFormatNatural"><code>$PreferDateTimeFormatNatural</code></dt>
<dd>

<p>By default, RT parses an unknown date first with <a href="http://metacpan.org/module/Time::ParseDate">Time::ParseDate</a>, and if this fails with <a href="http://metacpan.org/module/DateTime::Format::Natural">DateTime::Format::Natural</a>. <code>$PreferDateTimeFormatNatural</code> changes this behavior to first parse with <a href="http://metacpan.org/module/DateTime::Format::Natural">DateTime::Format::Natural</a>, and if this fails with <a href="http://metacpan.org/module/Time::ParseDate">Time::ParseDate</a>. This gives you the possibility to use the more advanced features of <a href="http://metacpan.org/module/DateTime::Format::Natural">DateTime::Format::Natural</a>. For example with <a href="http://metacpan.org/module/Time::ParseDate">Time::ParseDate</a> it isn&#39;t possible to get the &#39;first day of the last month&#39;, where <a href="http://metacpan.org/module/DateTime::Format::Natural">DateTime::Format::Natural</a> supports this with &#39;last month&#39;.</p>

<p>Be aware that <a href="http://metacpan.org/module/Time::ParseDate">Time::ParseDate</a> and <a href="http://metacpan.org/module/DateTime::Format::Natural">DateTime::Format::Natural</a> have different definitions for the relative date and time syntax. <a href="http://metacpan.org/module/Time::ParseDate">Time::ParseDate</a> returns for &#39;last month&#39; this DayOfMonth from the last month. <a href="http://metacpan.org/module/DateTime::Format::Natural">DateTime::Format::Natural</a> returns for &#39;last month&#39; the first day of the last month. So changing this config option maybe changes the results of your saved searches.</p>

</dd>
</dl>

<h1 id="Cryptography"><a href="#___top">Cryptography</a></h1>

<p>A complete description of RT&#39;s cryptography capabilities can be found in <a href="RT/Crypt.html">RT::Crypt</a>. At this moment, GnuPG (PGP) and SMIME security protocols are supported.</p>

<dl>

<dt id="Crypt"><code>%Crypt</code></dt>
<dd>

<p>The following options apply to all cryptography protocols.</p>

<p>By default, all enabled security protocols will analyze each incoming email. You may set <code>Incoming</code> to a subset of this list, if some enabled protocols do not apply to incoming mail; however, this is usually unnecessary.</p>

<p>For outgoing emails, the first security protocol from the above list is used. Use the <code>Outgoing</code> option to set a security protocol that should be used in outgoing emails. At this moment, only one protocol can be used to protect outgoing emails.</p>

<p>Set <code>RejectOnMissingPrivateKey</code> to 0 if you don&#39;t want to reject emails encrypted for key RT doesn&#39;t have and can not decrypt.</p>

<p>Set <code>RejectOnBadData</code> to 0 if you don&#39;t want to reject letters with incorrect data.</p>

<p>If you want to allow people to encrypt attachments inside the DB then set <code>AllowEncryptDataInDB</code> to 1.</p>

<p>Set <code>Dashboards</code> to a hash with Encrypt and Sign keys to control whether dashboards should be encrypted and/or signed correspondingly. By default they are not encrypted or signed.</p>

</dd>
</dl>

<h2 id="SMIME-configuration"><a href="#___top">SMIME configuration</a></h2>

<p>A full description of the SMIME integration can be found in <a href="RT/Crypt/SMIME.html">RT::Crypt::SMIME</a>.</p>

<dl>

<dt id="SMIME"><code>%SMIME</code></dt>
<dd>

<p>Set <code>Enable</code> to 0 or 1 to disable or enable SMIME for encrypting and signing messages.</p>

<p>Set <code>OpenSSL</code> to path to <i>openssl</i> executable.</p>

<p>Set <code>Keyring</code> to directory with key files. Key and certificates should be stored in a PEM file in this directory named named, e.g., <i>email.address@example.com.pem</i>.</p>

<p>Set <code>CAPath</code> to either a PEM-formatted certificate of a single signing certificate authority, or a directory of such (including hash symlinks as created by the openssl tool <code>c_rehash</code>). Only SMIME certificates signed by these certificate authorities will be treated as valid signatures. If left unset (and <code>AcceptUntrustedCAs</code> is unset, as it is by default), no signatures will be marked as valid!</p>

<p>Set <code>AcceptUntrustedCAs</code> to allow arbitrary SMIME certificates, no matter their signing entities. Such mails will be marked as untrusted, but signed; <code>CAPath</code> will be used to mark which mails are signed by trusted certificate authorities. This configuration is generally insecure, as it allows the possibility of accepting forged mail signed by an untrusted certificate authority.</p>

<p>Setting <code>AcceptUntrustedCAs</code> also allows encryption to users with certificates created by untrusted CAs.</p>

<p>Set <code>Passphrase</code> to a scalar (to use for all keys), an anonymous function, or a hash (to look up by address). If the hash is used, the &#39;&#39; key is used as a default.</p>

<p>See <a href="RT/Crypt/SMIME.html">RT::Crypt::SMIME</a> for details.</p>

</dd>
</dl>

<h2 id="GnuPG-configuration"><a href="#___top">GnuPG configuration</a></h2>

<p>A full description of the (somewhat extensive) GnuPG integration can be found by running the command `perldoc <a href="RT/Crypt/GnuPG.html">RT::Crypt::GnuPG</a>` (or `perldoc lib/RT/Crypt/GnuPG.pm` from your RT install directory).</p>

<dl>

<dt id="GnuPG"><code>%GnuPG</code></dt>
<dd>

<p>Set <code>Enable</code> to 0 or 1 to disable or enable GnuPG interfaces for encrypting and signing outgoing messages.</p>

<p>Set <code>GnuPG</code> to the name or path of the gpg binary to use.</p>

<p>Set <code>Passphrase</code> to a scalar (to use for all keys), an anonymous function, or a hash (to look up by address). If the hash is used, the &#39;&#39; key is used as a default.</p>

<p>Set <code>OutgoingMessagesFormat</code> to &#39;inline&#39; to use inline encryption and signatures instead of &#39;RFC&#39; (GPG/MIME: RFC3156 and RFC1847) format.</p>

</dd>
<dt id="GnuPGOptions"><code>%GnuPGOptions</code></dt>
<dd>

<p>Options to pass to the GnuPG program.</p>

<p>If you override this in your RT_SiteConfig, you should be sure to include a homedir setting.</p>

<p>Note that options with &#39;-&#39; character MUST be quoted.</p>

</dd>
</dl>

<h1 id="External-storage"><a href="#___top">External storage</a></h1>

<p>By default, RT stores attachments in the database. ExternalStorage moves all attachments that RT does not need efficient access to (which include textual content and images) to outside of the database. This may either be on local disk, or to a cloud storage solution. This decreases the size of RT&#39;s database, in turn decreasing the burden of backing up RT&#39;s database, at the cost of adding additional locations which must be configured or backed up. Attachment storage paths are calculated based on file contents; this provides de-duplication.</p>

<p>A full description of external storage can be found by running the command `perldoc <a href="RT/ExternalStorage.html">RT::ExternalStorage</a>` (or `perldoc lib/RT/ExternalStorage.pm` from your RT install directory).</p>

<p>Note that simply configuring <a href="RT/ExternalStorage.html">RT::ExternalStorage</a> is insufficient; there are additional steps required (including setup of a regularly-scheduled upload job) to enable RT to make use of external storage.</p>

<dl>

<dt id="ExternalStorage"><code>%ExternalStorage</code></dt>
<dd>

<p>This selects which storage engine is used, as well as options for configuring that specific storage engine. RT ships with the following storage engines:</p>

<p><a href="RT/ExternalStorage/Disk.html">RT::ExternalStorage::Disk</a>, which stores files on directly onto disk.</p>

<p><a href="RT/ExternalStorage/AmazonS3.html">RT::ExternalStorage::AmazonS3</a>, which stores files on Amazon&#39;s S3 service.</p>

<p><a href="RT/ExternalStorage/Dropbox.html">RT::ExternalStorage::Dropbox</a>, which stores files in Dropbox.</p>

<p>See each storage engine&#39;s documentation for the configuration it requires and accepts.</p>

<pre><code>    Set(%ExternalStorage,
        Type =&gt; &#39;Disk&#39;,
        Path =&gt; &#39;/opt/rt4/var/attachments&#39;,
    );</code></pre>

</dd>
<dt id="ExternalStorageCutoffSize"><code>$ExternalStorageCutoffSize</code></dt>
<dd>

<p>Certain object types, like values for Binary (aka file upload) custom fields, are always put into external storage. However, for other object types, like images and text, there is a line in the sand where you want small objects in the database but large objects in external storage. By default, objects larger than 10 MiB will be put into external storage. <code>$ExternalStorageCutoffSize</code> adjusts that line in the sand.</p>

<p>Note that changing this setting does not affect existing attachments, only the new ones that <code>sbin/rt-externalize-attachments</code> hasn&#39;t seen yet.</p>

</dd>
<dt id="ExternalStorageDirectLink"><code>$ExternalStorageDirectLink</code></dt>
<dd>

<p>Certain ExternalStorage backends can serve files over HTTP. For such backends, RT can link directly to those files in external storage. This cuts down download time and relieves resource pressure because RT&#39;s web server is no longer involved in retrieving and then immediately serving each attachment.</p>

<p>Of the storage engines that RT ships, only <a href="RT/ExternalStorage/AmazonS3.html">RT::ExternalStorage::AmazonS3</a> supports this feature, and you must manually configure it to allow direct linking.</p>

<p>Set this to 1 to link directly to files in external storage.</p>

</dd>
</dl>

<h1 id="Initialdata-Formats"><a href="#___top">Initialdata Formats</a></h1>

<p>RT supports pluggable data format parsers for <i><a href="initialdata.html">initialdata</a></i> files.</p>

<p>If you add format handlers, note that you can remove the perl entry if you don&#39;t want it available. <b>Removing the default perl entry may cause problems installing plugins and RT updates</b>. If so, re-enable it temporarily.</p>

<dl>

<dt id="InitialdataFormatHandlers"><code>$InitialdataFormatHandlers</code></dt>
<dd>

<p>Set the <code>$InitialdataFormatHandlers</code> to an arrayref containing a list of format handler modules. The &#39;perl&#39; entry is the system default, and handles perl-style intialdata files.</p>

<pre><code>    Set( $InitialdataFormatHandlers,
         [
            &#39;perl&#39;,
            &#39;RT::Extension::Initialdata::Foo&#39;,
            ...
         [
       );</code></pre>

</dd>
</dl>

<h1 id="Lifecycles"><a href="#___top">Lifecycles</a></h1>

<h2 id="Lifecycle-definitions"><a href="#___top">Lifecycle definitions</a></h2>

<p>Each lifecycle is a list of possible statuses split into three logic sets: <b>initial</b>, <b>active</b> and <b>inactive</b>. Each status in a lifecycle must be unique. (Statuses may not be repeated across sets.) Each set may have any number of statuses.</p>

<p>For example:</p>

<pre><code>    default =&gt; {
        initial  =&gt; [&#39;new&#39;],
        active   =&gt; [&#39;open&#39;, &#39;stalled&#39;],
        inactive =&gt; [&#39;resolved&#39;, &#39;rejected&#39;, &#39;deleted&#39;],
        ...
    },</code></pre>

<p>Status names can be from 1 to 64 ASCII characters. Statuses are localized using RT&#39;s standard internationalization and localization system.</p>

<dl>

<dt id="initial">initial</dt>
<dd>

<p>You can define multiple <b>initial</b> statuses for tickets in a given lifecycle.</p>

<p>RT will automatically set its <b>Started</b> date when you change a ticket&#39;s status from an <b>initial</b> state to an <b>active</b> or <b>inactive</b> status.</p>

</dd>
<dt id="active">active</dt>
<dd>

<p><b>Active</b> tickets are &quot;currently in play&quot; - they&#39;re things that are being worked on and not yet complete.</p>

</dd>
<dt id="inactive">inactive</dt>
<dd>

<p><b>Inactive</b> tickets are typically in their &quot;final resting state&quot;.</p>

<p>While you&#39;re free to implement a workflow that ignores that description, typically once a ticket enters an inactive state, it will never again enter an active state.</p>

<p>RT will automatically set the <b>Resolved</b> date when a ticket&#39;s status is changed from an <b>Initial</b> or <b>Active</b> status to an <b>Inactive</b> status.</p>

<p><b>deleted</b> is still a special status and protected by the <b>DeleteTicket</b> right, unless you re-defined rights (read below). If you don&#39;t want to allow ticket deletion at any time simply don&#39;t include it in your lifecycle.</p>

</dd>
</dl>

<p>Statuses in each set are ordered and listed in the UI in the defined order.</p>

<p>Changes between statuses are constrained by transition rules, as described below.</p>

<h2 id="Default-values"><a href="#___top">Default values</a></h2>

<p>In some cases a default value is used to display in UI or in API when value is not provided. You can configure defaults using the following syntax:</p>

<pre><code>    default =&gt; {
        ...
        defaults =&gt; {
            on_create =&gt; &#39;new&#39;,
            on_resolve =&gt; &#39;resolved&#39;,
            ...
        },
    },</code></pre>

<p>The following defaults are used.</p>

<dl>

<dt id="on_create">on_create</dt>
<dd>

<p>If you (or your code) doesn&#39;t specify a status when creating a ticket, RT will use the this status. See also <a href="#Statuses-available-during-ticket-creation">&quot;Statuses available during ticket creation&quot;</a>.</p>

</dd>
<dt id="approved">approved</dt>
<dd>

<p>When an approval is accepted, the status of depending tickets will be changed to this value.</p>

</dd>
<dt id="denied">denied</dt>
<dd>

<p>When an approval is denied, the status of depending tickets will be changed to this value.</p>

</dd>
<dt id="reminder_on_open">reminder_on_open</dt>
<dd>

<p>When a reminder is opened, the status will be changed to this value.</p>

</dd>
<dt id="reminder_on_resolve">reminder_on_resolve</dt>
<dd>

<p>When a reminder is resolved, the status will be changed to this value.</p>

</dd>
</dl>

<h2 id="Transitions-between-statuses-and-UI-actions"><a href="#___top">Transitions between statuses and UI actions</a></h2>

<p>A <b>Transition</b> is a change of status from A to B. You should define all possible transitions in each lifecycle using the following format:</p>

<pre><code>    default =&gt; {
        ...
        transitions =&gt; {
            &#39;&#39;       =&gt; [qw(new open resolved)],
            new      =&gt; [qw(open resolved rejected deleted)],
            open     =&gt; [qw(stalled resolved rejected deleted)],
            stalled  =&gt; [qw(open)],
            resolved =&gt; [qw(open)],
            rejected =&gt; [qw(open)],
            deleted  =&gt; [qw(open)],
        },
        ...
    },</code></pre>

<p>The order of items in the listing for each transition line affects the order they appear in the drop-down. If you change the config for &#39;open&#39; state listing to:</p>

<pre><code>    open     =&gt; [qw(stalled rejected deleted resolved)],</code></pre>

<p>then the &#39;resolved&#39; status will appear as the last item in the drop-down.</p>

<h3 id="Statuses-available-during-ticket-creation"><a href="#___top">Statuses available during ticket creation</a></h3>

<p>By default users can create tickets with a status of new, open, or resolved, but cannot create tickets with a status of rejected, stalled, or deleted. If you want to change the statuses available during creation, update the transition from &#39;&#39; (empty string), like in the example above.</p>

<h3 id="Protecting-status-changes-with-rights"><a href="#___top">Protecting status changes with rights</a></h3>

<p>A transition or group of transitions can be protected by a specific right. Additionally, you can name new right names, which will be added to the system to control that transition. For example, if you wished to create a lesser right than ModifyTicket for rejecting tickets, you could write:</p>

<pre><code>    default =&gt; {
        ...
        rights =&gt; {
            &#39;* -&gt; deleted&#39;  =&gt; &#39;DeleteTicket&#39;,
            &#39;* -&gt; rejected&#39; =&gt; &#39;RejectTicket&#39;,
            &#39;* -&gt; *&#39;        =&gt; &#39;ModifyTicket&#39;,
        },
        ...
    },</code></pre>

<p>This would create a new <code>RejectTicket</code> right in the system which you could assign to whatever groups you choose.</p>

<p>On the left hand side you can have the following variants:</p>

<pre><code>    &#39;&lt;from&gt; -&gt; &lt;to&gt;&#39;
    &#39;* -&gt; &lt;to&gt;&#39;
    &#39;&lt;from&gt; -&gt; *&#39;
    &#39;* -&gt; *&#39;</code></pre>

<p>Valid transitions are listed in order of priority. If a user attempts to change a ticket&#39;s status from <b>new</b> to <b>open</b> then the lifecycle is checked for presence of an exact match, then for &#39;any to <b>open</b>&#39;, &#39;<b>new</b> to any&#39; and finally &#39;any to any&#39;.</p>

<p>If you don&#39;t define any rights, or there is no match for a transition, RT will use the <b>DeleteTicket</b> or <b>ModifyTicket</b> as appropriate.</p>

<h3 id="Labeling-and-defining-actions"><a href="#___top">Labeling and defining actions</a></h3>

<p>For each transition you can define an action that will be shown in the UI; each action annotated with a label and an update type.</p>

<p>Each action may provide a default update type, which can be <b>Comment</b>, <b>Respond</b>, or absent. For example, you may want your staff to write a reply to the end user when they change status from <b>new</b> to <b>open</b>, and thus set the update to <b>Respond</b>. Neither <b>Comment</b> nor <b>Respond</b> are mandatory, and user may leave the message empty, regardless of the update type.</p>

<p>This configuration can be used to accomplish what $ResolveDefaultUpdateType was used for in RT 3.8.</p>

<p>Use the following format to define labels and actions of transitions:</p>

<pre><code>    default =&gt; {
        ...
        actions =&gt; [
            &#39;new -&gt; open&#39;     =&gt; { label =&gt; &#39;Open it&#39;, update =&gt; &#39;Respond&#39; },
            &#39;new -&gt; resolved&#39; =&gt; { label =&gt; &#39;Resolve&#39;, update =&gt; &#39;Comment&#39; },
            &#39;new -&gt; rejected&#39; =&gt; { label =&gt; &#39;Reject&#39;,  update =&gt; &#39;Respond&#39; },
            &#39;new -&gt; deleted&#39;  =&gt; { label =&gt; &#39;Delete&#39; },

            &#39;open -&gt; stalled&#39;  =&gt; { label =&gt; &#39;Stall&#39;,   update =&gt; &#39;Comment&#39; },
            &#39;open -&gt; resolved&#39; =&gt; { label =&gt; &#39;Resolve&#39;, update =&gt; &#39;Comment&#39; },
            &#39;open -&gt; rejected&#39; =&gt; { label =&gt; &#39;Reject&#39;,  update =&gt; &#39;Respond&#39; },

            &#39;stalled -&gt; open&#39;  =&gt; { label =&gt; &#39;Open it&#39; },
            &#39;resolved -&gt; open&#39; =&gt; { label =&gt; &#39;Re-open&#39;, update =&gt; &#39;Comment&#39; },
            &#39;rejected -&gt; open&#39; =&gt; { label =&gt; &#39;Re-open&#39;, update =&gt; &#39;Comment&#39; },
            &#39;deleted -&gt; open&#39;  =&gt; { label =&gt; &#39;Undelete&#39; },
        ],
        ...
    },</code></pre>

<p>In addition, you may define multiple actions for the same transition. Alternately, you may use &#39;* -&gt; x&#39; to match more than one transition. For example:</p>

<pre><code>    default =&gt; {
        ...
        actions =&gt; [
            ...
            &#39;new -&gt; rejected&#39; =&gt; { label =&gt; &#39;Reject&#39;, update =&gt; &#39;Respond&#39; },
            &#39;new -&gt; rejected&#39; =&gt; { label =&gt; &#39;Quick Reject&#39; },
            ...
            &#39;* -&gt; deleted&#39; =&gt; { label =&gt; &#39;Delete&#39; },
            ...
        ],
        ...
    },</code></pre>

<h2 id="Moving-tickets-between-queues-with-different-lifecycles"><a href="#___top">Moving tickets between queues with different lifecycles</a></h2>

<p>Unless there is an explicit mapping between statuses in two different lifecycles, you can not move tickets between queues with these lifecycles -- even if both use the exact same set of statuses. Such a mapping is defined as follows:</p>

<pre><code>    __maps__ =&gt; {
        &#39;from lifecycle -&gt; to lifecycle&#39; =&gt; {
            &#39;status in left lifecycle&#39; =&gt; &#39;status in right lifecycle&#39;,
            ...
        },
        ...
    },</code></pre>

<h1 id="SLA"><a href="#___top">SLA</a></h1>

<dl>

<dt id="ServiceAgreements"><code>%ServiceAgreements</code></dt>
<dd>

<pre><code>    Set( %ServiceAgreements, (
        Default =&gt; &#39;4h&#39;,
        QueueDefault =&gt; {
            &#39;Incident&#39; =&gt; &#39;2h&#39;,
        },
        Levels =&gt; {
            &#39;2h&#39; =&gt; { Resolve =&gt; { RealMinutes =&gt; 60*2 } },
            &#39;4h&#39; =&gt; { Resolve =&gt; { RealMinutes =&gt; 60*4 } },
        },
    ));</code></pre>

<p>In this example <i>Incident</i> is the name of the queue, and <i>2h</i> is the name of the SLA which will be applied to this queue by default.</p>

<p>Each service level can be described using several options: <a href="#Starts-interval-first-business-minute">Starts</a>, <a href="#Resolve-and-Response-interval-no-defaults">Resolve</a>, <a href="#Resolve-and-Response-interval-no-defaults">Response</a>, <a href="#Keep-in-loop-interval-no-defaults">KeepInLoop</a>, <a href="#OutOfHours-struct-no-default">OutOfHours</a> and <a href="#Configuring-business-hours">ServiceBusinessHours</a>.</p>

<dl>

<dt id="Starts-interval-first-business-minute">Starts (interval, first business minute)</dt>
<dd>

<p>By default when a ticket is created Starts date is set to first business minute after time of creation. In other words if a ticket is created during business hours then Starts will be equal to Created time, otherwise Starts will be beginning of the next business day.</p>

<p>However, if you provide 24/7 support then you most probably would be interested in Starts to be always equal to Created time.</p>

<p>Starts option can be used to adjust behaviour. Format of the option is the same as format for deadlines which described later in details. RealMinutes, BusinessMinutes options and OutOfHours modifiers can be used here like for any other deadline. For example:</p>

<pre><code>    &#39;standard&#39; =&gt; {
        # give people 15 minutes
        Starts   =&gt; { BusinessMinutes =&gt; 15  },
    },</code></pre>

<p>You can still use old option StartImmediately to set Starts date equal to Created date.</p>

<p>Example:</p>

<pre><code>    &#39;24/7&#39; =&gt; {
        StartImmediately =&gt; 1,
        Response =&gt; { RealMinutes =&gt; 30 },
    },</code></pre>

<p>But it&#39;s the same as:</p>

<pre><code>    &#39;24/7&#39; =&gt; {
        Starts =&gt; { RealMinutes =&gt; 0 },
        Response =&gt; { RealMinutes =&gt; 30 },
    },</code></pre>

</dd>
<dt id="Resolve-and-Response-interval-no-defaults">Resolve and Response (interval, no defaults)</dt>
<dd>

<p>These two options define deadlines for resolve of a ticket and reply to customer(requestors) questions accordingly.</p>

<p>You can define them using real time, business or both. Read more about the latter <a href="#Using-both-Resolve-and-Response-in-the-same-level">below</a>.</p>

<p>The Due date field is used to store calculated deadlines.</p>

<dl>

<dt id="Resolve">Resolve</dt>
<dd>

<p>Defines deadline when a ticket should be resolved. This option is quite simple and straightforward when used without <a href="#Response">&quot;Response&quot;</a>.</p>

<p>Example:</p>

<pre><code>    # 8 business hours
    &#39;simple&#39; =&gt; { Resolve =&gt; 60*8 },
    ...
    # one real week
    &#39;hard&#39; =&gt; { Resolve =&gt; { RealMinutes =&gt; 60*24*7 } },</code></pre>

</dd>
<dt id="Response">Response</dt>
<dd>

<p>In many companies providing support service(s) resolve time of a ticket is less important than time of response to requestors from staff members.</p>

<p>You can use Response option to define such deadlines. The Due date is set when a ticket is created, unset when a worker replies, and re-set when the requestor replies again -- until the ticket is closed, when the ticket&#39;s Due date is unset.</p>

<p><b>NOTE</b> that this behaviour changes when Resolve and Response options are combined; see <a href="#Using-both-Resolve-and-Response-in-the-same-level">&quot;Using both Resolve and Response in the same level&quot;</a>.</p>

<p>Note that by default, only the requestors on the ticket are considered &quot;outside actors&quot; and thus require a Response due date; all other email addresses are treated as workers of the ticket, and thus count as meeting the SLA. If you&#39;d like to invert this logic, so that the Owner and AdminCcs are the only worker email addresses, and all others are external, see the <a href="#AssumeOutsideActor">&quot;AssumeOutsideActor&quot;</a> configuration.</p>

<p>The owner is never treated as an outside actor; if they are also the requestor of the ticket, it will have no SLA.</p>

<p>If an outside actor replies multiple times, their later replies are ignored; the deadline is always calculated from the oldest correspondence from the outside actor.</p>

</dd>
<dt id="Using-both-Resolve-and-Response-in-the-same-level">Using both Resolve and Response in the same level</dt>
<dd>

<p>Resolve and Response can be combined. In such case due date is set according to the earliest of two deadlines and never is dropped to &#39;not set&#39;.</p>

<p>If a ticket met its Resolve deadline then due date stops &quot;flipping&quot;, is freezed and the ticket becomes overdue. Before that moment when an inside actor replies to a ticket, due date is changed to Resolve deadline instead of &#39;Not Set&#39;, as well this happens when a ticket is closed. So all the time due date is defined.</p>

<p>Example:</p>

<pre><code>    &#39;standard delivery&#39; =&gt; {
        Response =&gt; { RealMinutes =&gt; 60*1  }, # one hour
        Resolve  =&gt; { RealMinutes =&gt; 60*24 }, # 24 real hours
    },</code></pre>

<p>A client orders goods and due date of the order is set to the next one hour, you have this hour to process the order and write a reply. As soon as goods are delivered you resolve tickets and usually meet Resolve deadline, but if you don&#39;t resolve or user replies then most probably there are problems with delivery of the goods. And if after a week you keep replying to the client and always meeting one hour response deadline that doesn&#39;t mean the ticket is not over due. Due date was frozen 24 hours after creation of the order.</p>

</dd>
<dt id="Using-business-and-real-time-in-one-option">Using business and real time in one option</dt>
<dd>

<p>It&#39;s quite rare situation when people need it, but we&#39;ve decided that business is applied first and then real time when deadline described using both types of time. For example:</p>

<pre><code>    &#39;delivery&#39; =&gt; {
        Resolve =&gt; { BusinessMinutes =&gt; 0, RealMinutes =&gt; 60*8 },
    },
    &#39;fast delivery&#39; {
        StartImmediately =&gt; 1,
        Resolve =&gt; { RealMinutes =&gt; 60*8 },
    },</code></pre>

<p>For delivery requests which come into the system during business hours these levels define the same deadlines, otherwise the first level set deadline to 8 real hours starting from the next business day, when tickets with the second level should be resolved in the next 8 hours after creation.</p>

</dd>
</dl>

</dd>
<dt id="Keep-in-loop-interval-no-defaults">Keep in loop (interval, no defaults)</dt>
<dd>

<p>If response deadline is used then Due date is changed to repsonse deadline or to &quot;Not Set&quot; when staff replies to a ticket. In some cases you want to keep requestors in loop and keed them up to date every few hours. KeepInLoop option can be used to achieve this.</p>

<pre><code>    &#39;incident&#39; =&gt; {
        Response   =&gt; { RealMinutes =&gt; 60*1  }, # one hour
        KeepInLoop =&gt; { RealMinutes =&gt; 60*2 }, # two hours
        Resolve    =&gt; { RealMinutes =&gt; 60*24 }, # 24 real hours
    },</code></pre>

<p>In the above example Due is set to one hour after creation, reply of a inside actor moves Due date two hours forward, outside actors&#39; replies move Due date to one hour and resolve deadine is 24 hours.</p>

</dd>
<dt id="Modifying-Agreements">Modifying Agreements</dt>
<dd>

<dl>

<dt id="OutOfHours-struct-no-default">OutOfHours (struct, no default)</dt>
<dd>

<p>Out of hours modifier. Adds more real or business minutes to resolve and/or reply options if event happens out of business hours, read also &lt;/&quot;Configuring business hours&quot;&gt; below.</p>

<p>Example:</p>

<pre><code>    &#39;level x&#39; =&gt; {
        OutOfHours =&gt; { Resolve =&gt; { RealMinutes =&gt; +60*24 } },
        Resolve    =&gt; { RealMinutes =&gt; 60*24 },
    },</code></pre>

<p>If a request comes into the system during night then supporters have two hours, otherwise only one.</p>

<pre><code>    &#39;level x&#39; =&gt; {
        OutOfHours =&gt; { Response =&gt; { BusinessMinutes =&gt; +60*2 } },
        Resolve    =&gt; { BusinessMinutes =&gt; 60 },
    },</code></pre>

<p>Supporters have two additional hours in the morning to deal with bunch of requests that came into the system during the last night.</p>

</dd>
<dt id="IgnoreOnStatuses-array-no-default">IgnoreOnStatuses (array, no default)</dt>
<dd>

<p>Allows you to ignore a deadline when ticket has certain status. Example:</p>

<pre><code>    &#39;level x&#39; =&gt; {
        KeepInLoop =&gt; { BusinessMinutes =&gt; 60, IgnoreOnStatuses =&gt; [&#39;stalled&#39;] },
    },</code></pre>

<p>In above example KeepInLoop deadline is ignored if ticket is stalled.</p>

<p><b>NOTE</b>: When a ticket goes from an ignored status to a normal status, the new Due date is calculated from the last action (reply, SLA change, etc) which fits the SLA type (Response, Starts, KeepInLoop, etc). This means if a ticket in the above example flips from stalled to open without a reply, the ticket will probably be overdue. In most cases this shouldn&#39;t be a problem since moving out of stalled-like statuses is often the result of RT&#39;s auto-open on reply scrip, therefore ensuring there&#39;s a new reply to calculate Due from. The overall effect is that ignored statuses don&#39;t let the Due date drift arbitrarily, which could wreak havoc on your SLA performance. <code>ExcludeTimeOnIgnoredStatuses</code> option could get around the &quot;probably be overdue&quot; issue by excluding the time spent on ignored statuses, e.g.</p>

<pre><code>    &#39;level x&#39; =&gt; {
        KeepInLoop =&gt; {
            BusinessMinutes =&gt; 60,
            ExcludeTimeOnIgnoredStatuses =&gt; 1,
            IgnoreOnStatuses =&gt; [&#39;stalled&#39;],
        },
    },</code></pre>

</dd>
</dl>

</dd>
<dt id="Defining-service-levels-per-queue">Defining service levels per queue</dt>
<dd>

<p>In the config you can set per queue defaults, using:</p>

<pre><code>    Set( %ServiceAgreements, (
        Default =&gt; &#39;global default level of service&#39;,
        QueueDefault =&gt; {
            &#39;queue name&#39; =&gt; &#39;default value for this queue&#39;,
            ...
        },
        ...
    ));</code></pre>

</dd>
<dt id="AssumeOutsideActor">AssumeOutsideActor</dt>
<dd>

<p>When using a <a href="#Resolve-and-Response-interval-no-defaults">Response</a> configuration, the due date is unset when anyone who is not a requestor replies. If it is common for non-requestors to reply to tickets, and this should <i>not</i> satisfy the SLA, you may wish to set <code>AssumeOutsideActor</code>. This causes the extension to assume that the Response SLA has only been met when the owner or AdminCc reply.</p>

<pre><code>    Set ( %ServiceAgreements = (
        AssumeOutsideActor =&gt; 1,
        ...
    ));</code></pre>

</dd>
</dl>

</dd>
<dt id="ServiceBusinessHours"><code>%ServiceBusinessHours</code></dt>
<dd>

<p>In the config you can set one or more work schedules, e.g.</p>

<pre><code>    Set( %ServiceBusinessHours, (
        &#39;Default&#39; =&gt; {
            ... description ...
        },
        &#39;Support&#39; =&gt; {
            ... description ...
        },
        &#39;Sales&#39; =&gt; {
            ... description ...
        },
    ));</code></pre>

<p>Read more about how to describe a schedule in <a href="http://metacpan.org/module/Business::Hours">Business::Hours</a>.</p>

<dl>

<dt id="Configuring-business-hours">Configuring business hours</dt>
<dd>

<p>Each level supports BusinessHours option to specify your own business hours.</p>

<pre><code>    &#39;level x&#39; =&gt; {
        BusinessHours =&gt; &#39;work just in Monday&#39;,
        Resolve    =&gt; { BusinessMinutes =&gt; 60 },
    },</code></pre>

<p>then <a href="http://metacpan.org/module/%ServiceBusinessHours">%ServiceBusinessHours</a> should have the corresponding definition:</p>

<pre><code>    Set( %ServiceBusinessHours, (
        &#39;work just in Monday&#39; =&gt; {
            1 =&gt; { Name =&gt; &#39;Monday&#39;, Start =&gt; &#39;9:00&#39;, End =&gt; &#39;18:00&#39; },
        },
    ));</code></pre>

<p>Default Business Hours setting is in $ServiceBusinessHours{&#39;Default&#39;}.</p>

</dd>
</dl>

</dd>
</dl>

<h1 id="Administrative-interface"><a href="#___top">Administrative interface</a></h1>

<dl>

<dt id="ShowRTPortal"><code>$ShowRTPortal</code></dt>
<dd>

<p>RT can show administrators a feed of recent RT releases and other related announcements and information from Best Practical on the top level Admin page. This feature helps you stay up to date on RT security announcements and version updates.</p>

<p>RT provides this feature using an &quot;iframe&quot; on <code>/Admin/index.html</code> which asks the administrator&#39;s browser to show an inline page from Best Practical&#39;s website.</p>

<p>If you&#39;d rather not make this feature available to your administrators, set <code>$ShowRTPortal</code> to 0.</p>

</dd>
<dt id="AdminSearchResultFormat"><code>%AdminSearchResultFormat</code></dt>
<dd>

<p>In the admin interface, format strings similar to tickets result formats are used. Use <code>%AdminSearchResultFormat</code> to define the format strings used in the admin interface on a per-RT-class basis.</p>

</dd>
<dt id="AdminSearchResultRows"><code>%AdminSearchResultRows</code></dt>
<dd>

<p>Use <code>%AdminSearchResultRows</code> to define the search result rows in the admin interface on a per-RT-class basis.</p>

</dd>
</dl>

<h1 id="Development-options"><a href="#___top">Development options</a></h1>

<dl>

<dt id="DevelMode"><code>$DevelMode</code></dt>
<dd>

<p>RT comes with a &quot;Development mode&quot; setting. This setting, as a convenience for developers, turns on several of development options that you most likely don&#39;t want in production:</p>

<ul>

<li><p>Disables CSS and JS minification and concatenation. Both CSS and JS will be instead be served as a number of individual smaller files, unchanged from how they are stored on disk.</p>

</li>
<li><p>Uses <a href="http://metacpan.org/module/Module::Refresh">Module::Refresh</a> to reload changed Perl modules on each request.</p>

</li>
<li><p>Turns off Mason&#39;s <code>static_source</code> directive; this causes Mason to reload template files which have been modified on disk.</p>

</li>
<li><p>Turns on Mason&#39;s HTML <code>error_format</code>; this renders compilation errors to the browser, along with a full stack trace. It is possible for stack traces to reveal sensitive information such as passwords or ticket content.</p>

</li>
<li><p>Turns off caching of callbacks; this enables additional callbacks to be added while the server is running.</p>

</li>
</ul>

</dd>
<dt id="RecordBaseClass"><code>$RecordBaseClass</code></dt>
<dd>

<p>What abstract base class should RT use for its records. You should probably never change this.</p>

<p>Valid values are <code>DBIx::SearchBuilder::Record</code> or <code>DBIx::SearchBuilder::Record::Cachable</code></p>

</dd>
<dt id="MasonParameters"><code>@MasonParameters</code></dt>
<dd>

<p><code>@MasonParameters</code> is the list of parameters for the constructor of HTML::Mason&#39;s Apache or CGI Handler. This is normally only useful for debugging, e.g. profiling individual components with:</p>

<pre><code>    use MasonX::Profiler; # available on CPAN
    Set(@MasonParameters, (preamble =&gt; &#39;my $p = MasonX::Profiler-&gt;new($m, $r);&#39;));</code></pre>

</dd>
<dt id="StatementLog"><code>$StatementLog</code></dt>
<dd>

<p>RT has rudimentary SQL statement logging support; simply set <code>$StatementLog</code> to be the level that you wish SQL statements to be logged at.</p>

<p>Enabling this option will also expose the SQL Queries page in the Admin -&gt; Tools menu for SuperUsers.</p>

</dd>
<dt id="HideOneTimeSuggestions"><code>$HideOneTimeSuggestions</code></dt>
<dd>

<p>On ticket comment and correspond there are &quot;One-time Cc&quot; and &quot;One-time Bcc&quot; fields. As part of this section, RT includes a list of suggested email addresses based on the correspondence history for that ticket. This list may grow quite large over time.</p>

<p>Enabling this option will hide the list behind a &quot;(show suggestions)&quot; link to cut down on page clutter. Once this option is clicked the link will change to &quot;(hide suggestions)&quot; and the full list of email addresses will be shown.</p>

</dd>
</dl>

<a href="./">&larr; Back to index</a>