Mailing-List: contact bugs-help@httpd.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Apache HTTPD Bugs Notification List" <bugs@httpd.apache.org>
Date: Mon, 3 Jan 2005 16:08:55 +0100
Message-Id: <200501031508.j03F8tXA018426@ajax.apache.org>
From: bugzilla@apache.org
To: bugs@httpd.apache.org
Subject: DO NOT REPLY [Bug 32479]  -
    Apache 2.0.42 hangs on AIX 5.2

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG�
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://issues.apache.org/bugzilla/show_bug.cgi?id=32479>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND�
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=32479


------- Additional Comments From trawick@apache.org  2005-01-03 16:08 -------
>(74)No buffer space available:
What is happening is that there is not enough of a certain type of kernel memory
to perform common TCP/IP-related operations, so various applications can fail.

The most likely cause of this is:

a) your network buffer cache maximum size is too large for the amount of memory
on your system
and
b) you are sending a wide variety of static files with Apache such that a large
amount of memory gets used by the network buffer cache

The quick fix is to do both of these steps:

a) get Apache to stop using network buffer cache by adding "EnableSendfile Off"
in httpd.conf, then restart Apache
b) empty the network buffer cache to relieve the memory problem, which will
continue even after restarting Apache since the network buffer cache won't be
cleared automatically; empty the cache by running "no -o nbc_limit=0"
After clearing the cache, if you still have that type of problem ("No buffer
space available") you may have to reboot the system to recover.  (I'd like to
hear if clearing the cache didn't resolve the problem.)  The current cache usage
can be displayed via "netstat -c".

A more permanent solution:

After you've determined that the above change resolves the problem permanently,
there is still a potential problem on your system: the default network buffer
cache size for your system is too large, so that if any application (whether or
not Apache) fully utilizes the cache you'll hit the same problem again.  You can
permanently avoid such issues by adding "no -o nbc_limit=0" to a startup script,
or you can specify a relatively small cache size (e.g., 100MB) and get Apache to
use sendfile again by commenting out your "EnableSendfile Off" directive and
restarting Apache, then see if that small cache size is safe.

-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org