commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sandy McArthur <sandy...@gmail.com>
Subject Re: Pool: comments requested for a new implementation
Date Mon, 14 Nov 2005 22:01:19 GMT
I made a small benchmark that tests single threaded throughput
comparing similarly configured CompositeObjectPool (COP) with
GenericObjectPool (GOP).

I've posted the results from a rather beefy HyperThreaded Pentium 4 3
GHz idle machine using Sun's 1.5 server JVM at:
http://sandy.mcarthur.org/pool/bench/single-thread-compare-to-generic.html

For that configuration COP  is faster than GOP by between 40% to 55%.

On my much less powerful old mac laptop with the client jvm the
difference was only about 20%.

The test I used can be found in the source download at:
http://sandy.mcarthur.org/pool/ Feel free to explain how my benchmark
is broken. :-)

I'll make a multi-threaded benchmark in the near future. Before I was
nervous COP wouldn't be as fast as GOP but now I'm rather confident
COP will be an even larger percentage faster in a multi-threaded
benchmark.

I'm still hoping for feedback.

--
Sandy McArthur

"He who dares not offend cannot be honest."
- Thomas Paine

---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org


Mime
View raw message