blob: d169d2288be38c3e979748a7540c2db9598bca89 [file] [log] [blame]
<?xml version='1.0' encoding='UTF-8' ?>
<!DOCTYPE manualpage SYSTEM "./style/manualpage.dtd">
<?xml-stylesheet type="text/xsl" href="./style/manual.en.xsl"?>
<!-- $LastChangedRevision$ -->
<!--
Licensed to the Apache Software Foundation (ASF) under one or more
contributor license agreements. See the NOTICE file distributed with
this work for additional information regarding copyright ownership.
The ASF licenses this file to You under the Apache License, Version 2.0
(the "License"); you may not use this file except in compliance with
the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
-->
<manualpage metafile="stopping.xml.meta">
<title>Stopping and Restarting Apache HTTP Server</title>
<summary>
<p>This document covers stopping and restarting Apache HTTP Server on
Unix-like systems. Windows NT, 2000 and XP users should see
<a href="platform/windows.html#winsvc">Running httpd as a
Service</a> and Windows 9x and ME users should see <a
href="platform/windows.html#wincons">Running httpd as a
Console Application</a> for information on how to control
httpd on those platforms.</p>
</summary>
<seealso><program>httpd</program></seealso>
<seealso><program>apachectl</program></seealso>
<seealso><a href="invoking.html">Starting</a></seealso>
<section id="introduction"><title>Introduction</title>
<p>In order to stop or restart the Apache HTTP Server, you must send a signal to
the running <program>httpd</program> processes. There are two ways to
send the signals. First, you can use the unix <code>kill</code>
command to directly send signals to the processes. You will
notice many <program>httpd</program> executables running on your system,
but you should not send signals to any of them except the parent,
whose pid is in the <directive
module="mpm_common">PidFile</directive>. That is to say you
shouldn't ever need to send signals to any process except the
parent. There are four signals that you can send the parent:
<code><a href="#term">TERM</a></code>,
<code><a href="#graceful">USR1</a></code>,
<code><a href="#hup">HUP</a></code>, and
<code><a href="#gracefulstop">WINCH</a></code>, which
will be described in a moment.</p>
<p>To send a signal to the parent you should issue a command
such as:</p>
<example>kill -TERM `cat /usr/local/apache2/logs/httpd.pid`</example>
<p>The second method of signaling the <program>httpd</program> processes
is to use the <code>-k</code> command line options: <code>stop</code>,
<code>restart</code>, <code>graceful</code> and <code>graceful-stop</code>,
as described below. These are arguments to the <program>
httpd</program> binary, but we recommend that
you send them using the <program>apachectl</program> control script, which
will pass them through to <program>httpd</program>.</p>
<p>After you have signaled <program>httpd</program>, you can read about
its progress by issuing:</p>
<example>tail -f /usr/local/apache2/logs/error_log</example>
<p>Modify those examples to match your <directive
module="core">ServerRoot</directive> and <directive
module="mpm_common">PidFile</directive> settings.</p>
</section>
<section id="term"><title>Stop Now</title>
<dl><dt>Signal: TERM</dt>
<dd><code>apachectl -k stop</code></dd>
</dl>
<p>Sending the <code>TERM</code> or <code>stop</code> signal to
the parent causes it to immediately attempt to kill off all of its
children. It may take it several seconds to complete killing off
its children. Then the parent itself exits. Any requests in
progress are terminated, and no further requests are served.</p>
</section>
<section id="graceful"><title>Graceful Restart</title>
<dl><dt>Signal: USR1</dt>
<dd><code>apachectl -k graceful</code></dd>
</dl>
<p>The <code>USR1</code> or <code>graceful</code> signal causes
the parent process to <em>advise</em> the children to exit after
their current request (or to exit immediately if they're not
serving anything). The parent re-reads its configuration files and
re-opens its log files. As each child dies off the parent replaces
it with a child from the new <em>generation</em> of the
configuration, which begins serving new requests immediately.</p>
<p>This code is designed to always respect the process control
directive of the MPMs, so the number of processes and threads
available to serve clients will be maintained at the appropriate
values throughout the restart process. Furthermore, it respects
<directive module="mpm_common">StartServers</directive> in the
following manner: if after one second at least <directive
module="mpm_common">StartServers</directive> new children have not
been created, then create enough to pick up the slack. Hence the
code tries to maintain both the number of children appropriate for
the current load on the server, and respect your wishes with the
<directive module="mpm_common">StartServers</directive>
parameter.</p>
<p>Users of <module>mod_status</module>
will notice that the server statistics are <strong>not</strong>
set to zero when a <code>USR1</code> is sent. The code was
written to both minimize the time in which the server is unable
to serve new requests (they will be queued up by the operating
system, so they're not lost in any event) and to respect your
tuning parameters. In order to do this it has to keep the
<em>scoreboard</em> used to keep track of all children across
generations.</p>
<p>The status module will also use a <code>G</code> to indicate
those children which are still serving requests started before
the graceful restart was given.</p>
<p>At present there is no way for a log rotation script using
<code>USR1</code> to know for certain that all children writing
the pre-restart log have finished. We suggest that you use a
suitable delay after sending the <code>USR1</code> signal
before you do anything with the old log. For example if most of
your hits take less than 10 minutes to complete for users on
low bandwidth links then you could wait 15 minutes before doing
anything with the old log.</p>
<note>
<p>When you issue a restart, a syntax check is first run, to
ensure that there are no errors in the configuration files.
If your configuration file has errors in it, you will get an
error message about that syntax error, and the server will refuse to
restart. This avoids the situation where the server halts and then
cannot restart, leaving you with a non-functioning server.</p>
<p>This still will not
guarantee that the server will restart correctly. To check the
semantics of the configuration files as well as the syntax, you
can try starting <program>httpd</program> as a non-root user. If there
are no errors it will attempt to open its sockets and logs and fail
because it's not root (or because the currently running
<program>httpd</program> already has those ports bound). If it fails
for any other reason then it's probably a config file error and the error
should be fixed before issuing the graceful restart.</p></note>
</section>
<section id="hup"><title>Restart Now</title>
<dl><dt>Signal: HUP</dt>
<dd><code>apachectl -k restart</code></dd>
</dl>
<p>Sending the <code>HUP</code> or <code>restart</code> signal to
the parent causes it to kill off its children like in
<code>TERM</code>, but the parent doesn't exit. It re-reads its
configuration files, and re-opens any log files. Then it spawns a
new set of children and continues serving hits.</p>
<p>Users of <module>mod_status</module>
will notice that the server statistics are set to zero when a
<code>HUP</code> is sent.</p>
<note>As with a graceful restart, a syntax check is run before the
restart is attempted. If your configuration file has errors in it, the
restart will not be attempted, and you will receive notification of the
syntax error(s).</note>
</section>
<section id="gracefulstop"><title>Graceful Stop</title>
<dl><dt>Signal: WINCH</dt>
<dd><code>apachectl -k graceful-stop</code></dd>
</dl>
<p>The <code>WINCH</code> or <code>graceful-stop</code> signal causes
the parent process to <em>advise</em> the children to exit after
their current request (or to exit immediately if they're not
serving anything). The parent will then remove its <directive
module="mpm_common">PidFile</directive> and cease listening on
all ports. The parent will continue to run, and monitor children
which are handling requests. Once all children have finalised
and exited or the timeout specified by the <directive
module="mpm_common">GracefulShutdownTimeout</directive> has been
reached, the parent will also exit. If the timeout is reached,
any remaining children will be sent the <code>TERM</code> signal
to force them to exit.</p>
<p>A <code>TERM</code> signal will immediately terminate the
parent process and all children when in the "graceful" state. However
as the <directive module="mpm_common">PidFile</directive> will
have been removed, you will not be able to use
<code>apachectl</code> or <code>httpd</code> to send this signal.</p>
<note><p>The <code>graceful-stop</code> signal allows you to run multiple
identically configured instances of <program>httpd</program> at the
same time. This is a powerful feature when performing graceful
upgrades of httpd, however it can also cause deadlocks and race
conditions with some configurations.</p>
<p>Care has been taken to ensure that on-disk files such as lock files
(<directive module="core">Mutex</directive>) and Unix socket files
(<directive module="mod_cgid">ScriptSock</directive>) contain the server
PID, and should coexist without problem. However, if a configuration
directive, third-party module or persistent CGI utilises any other on-disk
lock or state files, care should be taken to ensure that multiple running
instances of <program>httpd</program> do not clobber each other's files.</p>
<p>You should also be wary of other potential race conditions, such as
using <program>rotatelogs</program> style piped logging. Multiple running
instances of <program>rotatelogs</program> attempting to rotate the same
logfiles at the same time may destroy each other's logfiles.</p></note>
</section>
</manualpage>