httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Brian Behlendorf <br...@wired.com>
Subject feature request
Date Mon, 03 Apr 1995 20:31:48 GMT

If none of the hotshit programmers leap at this I'll give it a try myself,
but I think someone will be intrigued by this suggestion.  

It would be very useful for administrative purposes to be able to see what
request a given httpd process is running - i.e., if an httpd process is
running away with the CPU or sitting around for a long time, I'd like to know
if it's because it's a really big file and all is kosher, or there's a logjam
somewhere.  To this extent, how hard would it be to have httpd rewrite its
argv[0] like sendmail does (I think that's how sendmail does it) so one could
do a "ps" and see what request a particular httpd is answering? 

I ask because I just noticed 30 concurrent apaches running away with the CPU
on hyperreal (each using 3% of CPU time :) causing the load average to spike
up to 110 (on a box that very rarely goes over .5), and I'd like to know if
it was because of a bug in apache or because one of my users wrote a CGI
script that calls "find / -print" or something CPU intensive like that.  I'll
see if I can't get that situation to recreate itself.... 

	Brian

Mime
View raw message