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: Fri, 19 Nov 2004 23:31:01 +0100
Message-Id: <200411192231.iAJMV1BA029238@ajax.apache.org>
From: bugzilla@apache.org
To: bugs@httpd.apache.org
Subject: DO NOT REPLY [Bug 25757]  -
    Too Many Files Open Error

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=25757>.
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=25757


ewhite@stsci.edu changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|INVALID                     |


------- Additional Comments From ewhite@stsci.edu  2004-11-19 23:31 -------
Hello, me again with more information, but perhaps not ALL information.

It happened again to me.  Just today 11/19 ... my check server script starting
randomly getting 403 errors, and the logfile showed 403's and the 'too many
files open' error message.

Unfortunately, I do not have access myself to run lsof on the servers, and
everyone who had access was already gone for the weekend.  Therefore, yet again,
I did not get the lsof output as people are wanting.

HOWEVER, I do in fact have more information now.

I mentioned last time that I had made 3 config changes, and not long after, the
problem started.

Well, I since then had also turned on mod_usertrack on our other 4 (rather
indentical-ish) servers.  I also turned on the same custom logging.

No other server is having an issue.

This means that the ONLY change to this server, unique to it, before the problem
started ... was in taking it from a multihost, 1 host per ip server ... to a
multihost, 1 host for 1 ip, 3 virtual hosts on another ip, type server.

It would seem that something about that change/setup in particular, is causing
the FD leak, which takes a few days to leak enough.

For the record, hopefully it will be a little longer before my next leak, as I
have upped the ulimit on the process to 1024 (was at 256), and I have disabled
.htaccess (AllowOverride None), since I wasn't using .htaccess files anyway, and
didn't need it on.   Yet all the errors were based around trying to get a
.htaccess file.

I'm marking this for reopening, so that someone hopefully takes a look at this,
and at the virtual host code for fd leaks (perhaps mixed in with .htaccess stuff?)

Please contact me for any more questions/information.

-- 
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