So, I became curious about the performance of my site. This happened mainly because I hadn’t see the nice increase in performance from lighty I had hoped for. So I whipped up a matching apache config did a couple of quick ab tests.
You can see the results here:
Apache
Lighttpd
I know ab tests are not the end-all be-all of performance benchmarking, but I’m kinda disappointed.. I’ll go into more detail when it’s not one in the morning and I have to work. At that time I’ll also pose some alternative configurations. I’d like to see what I can get out of lighty because I was considering making a case for it at work…
Till then…
Rounnnd one....DING!
In preparation for the holiday from hell we're doing some performance tuning of our servers.
So, running valgrind on httpd with our current config yielded:
==7410== More than 100000 total errors detected. I'm not reporting any more.
==7410== Final error counts will be inaccurate. Go fix your program!
==7410== Rerun with --error-limit=no to disable this cutoff. Note
==7410== that errors may occur in your program without prior warning from
==7410== Valgrind, because errors are no longer being displayed.
So I ran it with --error-limit=no and still got the same result. Anyone have any ideas on this? Google wasn't terribly friendly on this subject.
Bullocks
So now I have trimmed up the apache config, and disabled some mod_python debugging. The results aren't pretty, but hey, I got results!
I'd love it if any readers or passer-by's could point me in the right direction on the error above.