On Wednesday, July 11, 2001, at 04:35 AM, Jon Stevens wrote:
> There are still a few issues to resolve, but let me just say, it
> kicks some serious ass. Jason is doing a really good job on it. It is
> much
> faster than before with the new pipeline methodology.
I agree. I ported the TDK and sample app to CVS head last night and
it's a difference you can actually see. I just hit it with a quick
ApacheBench test vs. the the 2.1 TDK to see what would happen and the
results are already very encouraging. Results from my quick (and
basically meaningless) test are below. Gives a good idea of how much
faster the 2.2 already seems though.
Both tests were done with the same settings (-n 500 -c 20) on my local
machine on the standard TDK sample app running off of the same MySql DB.
TDK 2.1 Sample App
-----------------------------------------------------
Time taken for tests: 17.500 seconds
Requests per second: 28.57
Transfer rate: 9.26 kb/s received
Connnection Times (ms)
min avg max
Connect: 0 0 32
Processing: 205 685 1378
Total: 205 685 1410
TDK CVS Head Sample App
------------------------------------------------------
Time taken for tests: 4.579 seconds
Requests per second: 109.19
Transfer rate: 35.52 kb/s received
Connnection Times (ms)
min avg max
Connect: 0 23 2658
Processing: 4 155 174
Total: 4 178 2832
Awesome job so far Jason.
-Kurt
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
