httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jim Jagielski <...@jaguNET.com>
Subject 1.1dev weirdness
Date Sun, 17 Mar 1996 03:57:43 GMT
I just decided to compile and run 1.1dev on one of our heavily
used servers, and saw some strange behavior. One thing is that
one of the kid processes was chewing up some _major_ CPU, which
has never been observed before (we were running 1.1ba or whatever
it is before that).

The other strange thing are some strange error messages:

    [Sat Mar 16 20:14:11 1996] access to  failed for 167.177.95.145, reason: File does not
exist
    [Sat Mar 16 20:17:39 1996] access to e/jpeg failed for 204.187.88.130, reason: File does
not exist
    [Sat Mar 16 21:10:28 1996] access to /staff/h failed for 204.162.98.47, reason: File does
not exist
    [Sat Mar 16 21:10:29 1996] access to  failed for 204.162.98.47, reason: File does not
exist
    [Sat Mar 16 21:51:09 1996] access to /staff/web/docs/robots.txt failed for 204.162.98.47,
reason: File does not exist
    [Sat Mar 16 21:51:09 1996] access to  failed for 204.162.98.47, reason: File does not
exist
    [Sat Mar 16 21:51:09 1996] access to 8 GMT failed for 204.162.98.47, reason: File does
not exist
    [Sat Mar 16 21:51:09 1996] access to /home/users1/ils/public_html/auto.htm failed for
204.162.98.47, reason: File does not exist
    [Sat Mar 16 22:33:33 1996] access to  failed for 205.161.186.155, reason: File does not
exist
    [Sat Mar 16 22:34:35 1996] access to Connecti failed for 204.162.98.47, reason: File does
not exist

Now some are from the some host, so maybe they are tring something weird,
or maybe something internal to Apache is getting mangled somehow...

I know that the 'csd dupped_csd' stuff has changed, and I originally
thought that that may be a factor (it was due to A/UX, which is what
we run here, that that code fragment was added) but POSTs seem to work
fine, which is what promoted the orginal fix. But it certainly does
seem as if the input and output streams are getting mixed somehow...
If the process chewing up time was the one connected to the above
host then maybe...

Any ideas? I'll try some obvious things but if anyone has any
clues, let me know. Thanks.
-- 
Jim Jagielski  << jim@jaguNET.com >>   |      "That's a Smith & Wesson,
  **  jaguNET Access Services  **      |       and you've had your six" 
      Email: info@jaguNET.com          |             - James Bond
++    http://www.jaguNET.com/         +++      Voice/Fax: 410-931-7060       ++

Mime
View raw message