You are viewing a plain text version of this content. The canonical link for it is here.
Posted to cvs@httpd.apache.org by el...@apache.org on 2018/02/10 12:15:37 UTC
svn commit: r1823749 - /httpd/httpd/trunk/docs/manual/mod/mod_proxy_fcgi.xml
Author: elukey
Date: Sat Feb 10 12:15:36 2018
New Revision: 1823749
URL: http://svn.apache.org/viewvc?rev=1823749&view=rev
Log:
mod_proxy_fcgi.xml: add a note about connection reuse
Modified:
httpd/httpd/trunk/docs/manual/mod/mod_proxy_fcgi.xml
Modified: httpd/httpd/trunk/docs/manual/mod/mod_proxy_fcgi.xml
URL: http://svn.apache.org/viewvc/httpd/httpd/trunk/docs/manual/mod/mod_proxy_fcgi.xml?rev=1823749&r1=1823748&r2=1823749&view=diff
==============================================================================
--- httpd/httpd/trunk/docs/manual/mod/mod_proxy_fcgi.xml (original)
+++ httpd/httpd/trunk/docs/manual/mod/mod_proxy_fcgi.xml Sat Feb 10 12:15:36 2018
@@ -82,6 +82,30 @@ ProxyPass "/myapp/" "fcgi://localhost:40
</highlight>
</example>
+ <note><title>Enable connection reuse to a FCGI backend like PHP-FPM</title>
+ <p>Please keep in mind that PHP-FPM (at the time of writing, February 2018)
+ uses a prefork model, namely each of its worker processes can handle one
+ connection at the time.<br />
+ By default mod_proxy (configured with <code>enablereuse=on</code>)
+ allows a connection pool of
+ <directive module="mpm_common">ThreadsPerChild</directive> connections to the
+ backend for each httpd process, so the following use cases should be taken
+ into account:</p>
+ <ul>
+ <li>Under HTTP/1.1 load it will likely cause the creation of up to
+ <directive module="mpm_common">MaxRequestWorkers</directive>
+ connections to the FCGI backend.</li>
+ <li>Under HTTP/2 load, due to how <module>mod_http2</module> is implemented,
+ there are additional h2 worker threads that may force the creation of other
+ backend connections. The overall count of connections in the pools may raise
+ to more than <directive module="mpm_common">MaxRequestWorkers</directive>.</li>
+ </ul>
+ <p>The maximum number of PHP-FPM worker processes needs to be configured wisely,
+ since there is the chance that they will all end up "busy" handling idle
+ persistent connections, without any room for new ones to be established,
+ and the end user experience will be a pile of HTTP request timeouts.</p>
+ </note>
+
<p> The following example passes the request URI as a filesystem
path for the PHP-FPM daemon to run. The request URL is implicitly added
to the 2nd parameter. The hostname and port following fcgi:// are where