March 2003 - Posts

I've just posted a new article on my site - In-depth Look at Caching: Part 1
This is the first part in the article series I'm working on, and it shows you how you can implement page caching on your site by using the @OutputCache directive.

Well anyway, go check it out! =)

Ok I've found the source of the problem - Network Load Balancing on my Windows Server 2003 box. It seemed to kill my entire home network. Once I got rid of it, everything went back to normal =)

Yeeeeehaaaaaw!!

Just ran a few more tests on a page that contains nothing more than

<%@ Page Language="C#" %>
<html>
<body>

</body>
</html>

The RPS peaks at 15 when running this on my machine with 1 connection. Other's I've talked with have told me their RPS usually lies at around 100-120.

Last night I started stress testing my component, CodeHighlight, just to make sure it performs well under heavy-load, but to my amazement it seemed to bring my beats of a machine to a stall. This happened when running the test in ACT, with 10 concurrent browser connections; everything worked fine for the first 30 seconds, the RPS peaked at around 90 or so and it didn't throw any errors at all, but then all of sudden it seemed to run out of memory or something. Because the entire thing suddenly choked completely.
At first I thought there problem was with my component, so I ended up working on it for about 6 hours trying to improve performance (lots of caching and so on.), but no matter what I did, it didn't help! Then a friend suggested that I should try to isolate the problem, by commenting out more and more sections of the component... To my amazement after commenting out the entire component of the test page, I still experienced the same behavior! This is in my mind very strange, because my machine is P4 2,4 GHz with 512 MB DDR 333 Mhz.
I've also tested it on my Windows Server 2003 box and the same happens there, even with just an empty page that does nothing at all.

Is this normal? In my mind it is not!

More Posts