You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by Brian Behlendorf <br...@organic.com> on 1995/06/06 20:30:53 UTC

0.7.1 test results

on www.links.net:

  many errors of "too many files open in the system"
  hits that looked like the following:

(null) - - [06/Jun/1995:07:23:52 -0700] "GET /pix/lilhot.gif HTTP/1.0" 500 1340
(null) - - [06/Jun/1995:07:23:53 -0700] "GET /pix/lilhot.gif HTTP/1.0" 500 1340
(null) - - [06/Jun/1995:07:23:56 -0700] "GET /pix/lilhot.gif HTTP/1.0" 500 1340
...
(null) - - [06/Jun/1995:07:20:39 -0700] "GET /latest.html HTTP/1.0" 500 6835
(null) - - [06/Jun/1995:07:20:39 -0700] "GET /latest.html HTTP/1.0" 500 6835
(null) - - [06/Jun/1995:07:20:39 -0700] "GET /latest.html HTTP/1.0" 500 6835
...

Eventually a reboot was called for.

on hyperreal.com:

  no system errors, but hits that looked like the following:

204.160.88.172 - - [06/Jun/1995:05:34:03 -0700] "GET /~dhchung/pspace.gif HTTP/1.0" 403 -
193.73.213.9 - - [06/Jun/1995:05:34:14 -0700] "GET / HTTP/1.0" 403 -
studpws167.unisg.ch - - [06/Jun/1995:05:34:22 -0700] "GET /drugs/pictures/ HTTP/1.0" 200 4187

  (success on directory indexes, 403 errors on everything else).
  No reboot was called for.

I'm watching the two machines now - response times seem to climb in 
direct correlation with size of process and process time, indicating it 
might not just be a lot of requests in the listen queue that's slowing 
some things down.  A kill -HUP to the parent seems to cause response 
times to jump back up - I'm going to see if doing that by cron every half 
hour causes things to stay in check.

Rob (H), if there's anything you want to set up on hyperreal to load test 
any patches let me know.

	Brian


--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
brian@organic.com  brian@hyperreal.com  http://www.[hyperreal,organic].com/