[concurrency-interest] ConcurrentSkipListMap performance

[Nathan Reynolds]

I turn on HotSpot's -XX:-PrintCompliation and run the warmup and test.  
If I see any messages during the actual test (after warmup), then I know 
I didn't warmup the JVM long enough.

The parameter -XX:CompileThreshold (i.e. optimizer threshold) defaults 
to 10,000 when using the "-server" flag.  I find that if I run the 
method under test and all of the test harness code 11,000 times then I 
don't see any compiler (optimizer) messages during my test.  I also put 
a Thread.sleep after warmup but before running the test.  This gives 
time and CPU resources to the compiler (optimizer) threads.

Getting this right was very tricky.  I am still not confident that I 
have mastered it yet.  I should really do some harness tests to gain 
more confidence.

Nathan Reynolds 
<http://psr.us.oracle.com/wiki/index.php/User:Nathan_Reynolds> | 
Consulting Member of Technical Staff | 602.333.9091
Oracle PSR Engineering <http://psr.us.oracle.com/> | Server Technology

On 10/25/2011 2:13 AM, Thorsten Möller wrote:
> Am 24.10.2011 um 22:20 schrieb Nathan Reynolds:
>
>> Note:  HotSpot does have an optimizer logging feature and can be used to determine if the optimizer is finished (i.e. by lack of further output).
> Interesting - wasn't aware of this yet. Could you detail how it can be used. Seems that it is a JVM parameter [1]?
>
> Thorsten
>
> [1] http://www.oracle.com/technetwork/java/javase/tech/vmoptions-jsp-140102.html
>
>
> _______________________________________________
> Concurrency-interest mailing list
> Concurrency-interest at cs.oswego.edu
> http://cs.oswego.edu/mailman/listinfo/concurrency-interest
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs.oswego.edu/pipermail/concurrency-interest/attachments/20111025/1cb174cb/attachment.html>