Sophie

Sophie

distrib > Mandriva > 2010.0 > i586 > media > contrib-release > by-pkgid > d265f71d7fce441fd72dd0a77d0f8893 > files > 124

apache-ssl-1.3.41_1.59-1mdv2010.0.i586.rpm

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">
  <head>
    <meta name="generator" content="HTML Tidy, see www.w3.org" />

    <title>Running a High-Performance Web Server for BSD</title>
  </head>
  <!-- Background white, links blue (unvisited), navy (visited), red (active) -->

  <body bgcolor="#FFFFFF" text="#000000" link="#0000FF"
  vlink="#000080" alink="#FF0000">
    <a id="initial" name="initial">
        <div align="CENTER">
      <img src="../images/sub.gif" alt="[APACHE DOCUMENTATION]" />

      <h3>Apache HTTP Server Version 1.3</h3>
        <p><small><em>Is this the version you want?  For more recent
         versions, check our <a href="/docs/">documentation 
         index</a>.</em></small></p>

    </div>


    </a> 

    <h1 align="CENTER">Running a High-Performance Web Server for
    BSD</h1>

    <p>This document assumes that you have read the appropriate
    overview documentation for
    <a href="http://www.FreeBSD.org/docs.html">FreeBSD</a>,
    <a href="http://www.NetBSD.org/Documentation/">NetBSD</a>, or
    <a href="http://www.OpenBSD.org/docum.html">OpenBSD</a>.
    In addition, the FreeBSD
    <a href="http://www.FreeBSD.org/cgi/man.cgi?query=tuning">tuning</a>
    manual page contains lots of wisdom, especially regarding sysctl
    options.</p>

    <p>Like other OS's, the listen queue is often the <strong>first
    limit hit</strong>. The following are comments from "Aaron
    Gifford &lt;agifford@InfoWest.COM&gt;" on how to fix this on
    BSDI 1.x, 2.x, and FreeBSD 2.0 (and earlier):</p>

    <p>Edit the following two files:</p>

    <blockquote>
      <code>/usr/include/sys/socket.h<br />
       /usr/src/sys/sys/socket.h</code>
    </blockquote>
    In each file, look for the following: 
<pre>
    /*
     * Maximum queue length specifiable by listen.
     */
    #define SOMAXCONN       5
</pre>
    Just change the "5" to whatever appears to work. I bumped the
    two machines I was having problems with up to 32 and haven't
    noticed the problem since. 

    <p>After the edit, recompile the kernel and recompile the
    Apache server then reboot.</p>

    <p>FreeBSD 2.1 seems to be perfectly happy, with SOMAXCONN set
    to 32 already.</p>

    <p><a id="detail" name="detail"><strong>Addendum for
    <em>very</em> heavily loaded BSD servers</strong><br />
    </a> from Chuck Murcko &lt;chuck@telebase.com&gt;</p>

    <p>If you're running a really busy BSD Apache server, the
    following are useful things to do if the system is acting
    sluggish:</p>

    <ul>
      <li>Run vmstat to check memory usage, page/swap rates,
      <em>etc.</em></li>

      <li>Run netstat -m to check mbuf usage</li>

      <li>Run fstat to check file descriptor usage</li>
    </ul>
    These utilities give you an idea what you'll need to tune in
    your kernel, and whether it'll help to buy more RAM. Here are
    some BSD kernel config parameters (actually BSDI, but pertinent
    to FreeBSD and other 4.4-lite derivatives) from a system
    getting heavy usage. The tools mentioned above were used, and
    the system memory was increased to 48 MB before these tuneups.
    Other system parameters remained unchanged. 
<pre>
maxusers        256
</pre>
    Maxusers drives a <em>lot</em> of other kernel parameters: 

    <ul>
      <li>Maximum # of processes</li>

      <li>Maximum # of processes per user</li>

      <li>System wide open files limit</li>

      <li>Per-process open files limit</li>

      <li>Maximum # of mbuf clusters</li>

      <li>Proc/pgrp hash table size</li>
    </ul>
    The actual formulae for these derived parameters are in
    <em>/usr/src/sys/conf/param.c</em>. These calculated parameters
    can also be overridden (in part) by specifying your own values
    in the kernel configuration file: 
<pre>
# Network options. NMBCLUSTERS defines the number of mbuf clusters and
# defaults to 256. This machine is a server that handles lots of traffic,
# so we crank that value.
options         NMBCLUSTERS=4096        # mbuf clusters at 4096

#
# Misc. options
#
options         CHILD_MAX=512           # maximum number of child processes
options         OPEN_MAX=512            # maximum fds (breaks RPC svcs)
</pre>

    <p>In many cases, NMBCLUSTERS must be set much larger than
    would appear necessary at first glance. The reason for this is
    that if the browser disconnects in mid-transfer, the socket fd
    associated with that particular connection ends up in the
    TIME_WAIT state for several minutes, during which time its
    mbufs are not yet freed. Another reason is that, on server
    timeouts, some connections end up in FIN_WAIT_2 state forever,
    because this state doesn't time out on the server, and the
    browser never sent a final FIN. For more details see the <a
    href="fin_wait_2.html">FIN_WAIT_2</a> page.</p>

    <p>Some more info on mbuf clusters (from sys/mbuf.h):</p>
<pre>
/*
 * Mbufs are of a single size, MSIZE (machine/machparam.h), which
 * includes overhead.  An mbuf may add a single "mbuf cluster" of size
 * MCLBYTES (also in machine/machparam.h), which has no additional overhead
 * and is used instead of the internal data area; this is done when
 * at least MINCLSIZE of data must be stored.
 */
</pre>

    <p>CHILD_MAX and OPEN_MAX are set to allow up to 512 child
    processes (different than the maximum value for processes per
    user ID) and file descriptors. These values may change for your
    particular configuration (a higher OPEN_MAX value if you've got
    modules or CGI scripts opening lots of connections or files).
    If you've got a lot of other activity besides httpd on the same
    machine, you'll have to set NPROC higher still. In this
    example, the NPROC value derived from maxusers proved
    sufficient for our load.</p>

    <p>To increase the size of the <code>listen()</code> queue, you
    need to adjust the value of SOMAXCONN. SOMAXCONN is not derived
    from maxusers, so you'll always need to increase that yourself.
    We use a value guaranteed to be larger than Apache's default
    for the listen() of 128, currently. The actual value for
    SOMAXCONN is set in <code>sys/socket.h</code>. The best way to
    adjust this parameter is run-time, rather than changing it in
    this header file and thus hardcoding a value in the kernel and
    elsewhere. To do this, edit <code>/etc/rc.local</code> and add
    the following line:</p>
<pre>
    /usr/sbin/sysctl -w kern.somaxconn=256
</pre>

    <p>We used <code>256</code> but you can tune it for your own
    setup. In many cases, however, even the default value of
    <code>128</code> (for later versions of FreeBSD) is OK.</p>

    <p><strong>Caveats</strong></p>

    <p>Be aware that your system may not boot with a kernel that is
    configured to use more resources than you have available system
    RAM. <strong>ALWAYS</strong> have a known bootable kernel
    available when tuning your system this way, and use the system
    tools beforehand to learn if you need to buy more memory before
    tuning.</p>

    <p>RPC services will fail when the value of OPEN_MAX is larger
    than 256. This is a function of the original implementations of
    the RPC library, which used a byte value for holding file
    descriptors. BSDI has partially addressed this limit in its 2.1
    release, but a real fix may well await the redesign of RPC
    itself.</p>

    <p>Finally, there's the hard limit of child processes
    configured in Apache.</p>

    <p>For versions of Apache later than 1.0.5 you'll need to
    change the definition for <strong>HARD_SERVER_LIMIT</strong> in
    <em>httpd.h</em> and recompile if you need to run more than the
    default 150 instances of httpd.</p>

    <p>From conf/httpd.conf-dist:</p>
<pre>
# Limit on total number of servers running, <em>i.e.</em>, limit on the number
# of clients who can simultaneously connect --- if this limit is ever
# reached, clients will be LOCKED OUT, so it should NOT BE SET TOO LOW.
# It is intended mainly as a brake to keep a runaway server from taking
# Unix with it as it spirals down...

MaxClients 150
</pre>
    Know what you're doing if you bump this value up, and make sure
    you've done your system monitoring, RAM expansion, and kernel
    tuning beforehand. Then you're ready to service some serious
    hits! 

    <p>Thanks to <em>Tony Sanders</em> and <em>Chris Torek</em> at
    BSDI for their helpful suggestions and information.</p>

    <p>"M. Teterin" &lt;mi@ALDAN.ziplink.net&gt; writes:</p>

    <blockquote>
      It really does help if your kernel and frequently used
      utilities are fully optimized. Rebuilding the FreeBSD kernel
      on an AMD-133 (486-class CPU) web-server with<br />
       <code>-m486 -fexpensive-optimizations -fomit-frame-pointer
      -O2</code><br />
       helped reduce the number of "unable" errors, because the CPU
      was often maxed out.
    </blockquote>

    <h2><a id="accf" name="accf">Accept filtering on
    FreeBSD</a></h2>

    <p>Versions of FreeBSD from August 2000 onwards include a
    feature called "accept filters" which delay the return from
    accept() until a condition has been met, e.g. an HTTP request
    has arrived. This postpones the requirement for a child process
    to handle the new connection which therefore increases the
    number of connections that a given number of child processes
    can handle. It also allows a child process to accomplish more
    immediately after accept() returns (because the request is
    already available to be read) so there is less context
    switching.</p>

    <p>Accept filters provide the most benefit on servers that are
    already so busy that they are configured with "<code>KeepAlive
    Off</code>". <a href="../keepalive.html">HTTP KeepAlive (aka
    persistent connections)</a> avoids the cost of setting up a new
    connection for every request, but connections that are being
    kept alive use up one of the available child processes. Since
    there is a limited number of child processes this can
    significantly reduce the capacity of the server. The viewers of
    a web site will still get a lot of the benefit of persistent
    connections even with a very small
    <code>KeepAliveTimeout</code> so you should try reducing it
    before turning it off altogether.</p>

    <p>To enable accept filtering, you must either load the
    appropriate accept filter module, e.g. with the command
    <code>kldload accf_http</code>, or compile a kernel with
    <code>options ACCEPT_FILTER_HTTP</code>. Apache will then
    enable filtering when it is restarted.</p>

    <p>Accept filters are compiled in if the symbol
    <code>SO_ACCEPTFILTER</code> is defined on the machine on which
    Apache is built. Additionally there is a directive <a
    href="../mod/core.html#acceptfilter">AcceptFilter</a> to switch
    the filters on or off. The default is on; except when apache is
    compiled with <code>-D AP_ACCEPTFILTER_ON</code>.</p>

    <p>See the manual page
    <a href="http://www.freebsd.org/cgi/man.cgi?query=accf_http">accf_http(9)</a>
    for more information.</p>

    <h3>More welcome!</h3>
    If you have tips to contribute, send mail to <a
    href="mailto:apache@apache.org">apache@apache.org</a> 
        <hr />

    <h3 align="CENTER">Apache HTTP Server Version 1.3</h3>
    <a href="./"><img src="../images/index.gif" alt="Index" /></a>
    <a href="../"><img src="../images/home.gif" alt="Home" /></a>

  </body>
</html>