Naman_65674
Feb 25, 2009Nimbostratus
Perfomance gain of using HTTP Classes over iRules
Hi,
I raised this issue in the general iRule forum (http://devcentral.f5.com/Default.aspx?tabid=53&forumid=5&postid=33291&view=topic), but I thought this would be the best place to focus the attention on the actual performance testing.
As mentionted before, we currently have a block of code in an iRule which sends traffic to different pools based on the uri using dataclasses. We want to move to using HTTP Classes but first would like to understand what kind of performance gain (if any) we should expect.
I tried running a few jMeter tests with ~1000 req/sec in both scenarios, comparing the throughput and using the following guide http://devcentral.f5.com/Default.aspx?tabid=63&articleType=ArticleView&articleId=123 but i couldnt see any notable differences.
Does anyone have any ideas of any other tools i could use to compare my two options? Maybe i need to generate more load?
Thanks,
Naman