james-server-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Niklas Therning <nik...@trillian.se>
Subject Re: [Mime4j] Benchmark of 0.3 vs 0.4
Date Mon, 04 Aug 2008 14:17:20 GMT
I'm using Ubuntu 8.04 64-bit on a 2.4 GHz core duo Dell laptop.

niklas@niklas-laptop:~$ uname -a
Linux niklas-laptop 2.6.24-19-generic #1 SMP Fri Jul 11 21:01:46 UTC 
2008 x86_64 GNU/Linux
niklas@niklas-laptop:~$ java -version
java version "1.6.0_06"
Java(TM) SE Runtime Environment (build 1.6.0_06-b02)
Java HotSpot(TM) 64-Bit Server VM (build 10.0-b22, mixed mode)

/Niklas

Stefano Bagnara wrote:
> Hi Niklas,
>
> thank you for sharing this interesing results!
>
> What JVM (java version / platform) are you using?
>
> Stefano
>
> Niklas Therning ha scritto:
>> Hi,
>>
>> I've done some benchmarking of the recent performance improvements in 
>> trunk vs Mime4j 0.3. I must say that the result is very promising! :)
>>
>> I did my tests on about 130000 messages, most of them less than 10 kB 
>> in size. All message files had been cached by the Linux file cache 
>> when I conducted the tests so it doesn't take the disk I/O in 
>> consideration. I guess I should also run the tests without the cache 
>> but I don't know how to force it to flush. The test uses the SAX-like 
>> parser interface and a ContentHandler which does nothing (only 
>> consumes the data).
>>
>> Here are the 0.3 results:
>>  Statistics for all messages (130606)
>>  Total duration: 23830 ms
>>  Average duration: 0.182457 ms
>>
>>  Statistics for all messages (124480) more than 0 kB and less than 10 
>> kB in size
>>  Total duration: 15264 ms
>>  Average duration: 0.122622 ms
>>
>>  Statistics for all messages (5934) more than 10 kB and less than 100 
>> kB in size
>>  Total duration: 5173 ms
>>  Average duration: 0.871756 ms
>>
>>  Statistics for all messages (161) more than 100 kB and less than 
>> 1000 kB in size
>>  Total duration: 1033 ms
>>  Average duration: 6.416149 ms
>>
>>  Statistics for all messages (31) more than 1000 kB in size
>>  Total duration: 2360 ms
>>  Average duration: 76.129032 ms
>>
>>
>> And here are the 0.4 results:
>>  Statistics for all messages (130606)
>>  Total duration: 10069 ms
>>  Average duration: 0.077094 ms
>>
>>  Statistics for all messages (124480) more than 0 kB and less than 10 
>> kB in size
>>  Total duration: 8583 ms
>>  Average duration: 0.068951 ms
>>
>>  Statistics for all messages (5934) more than 10 kB and less than 100 
>> kB in size
>>  Total duration: 1106 ms
>>  Average duration: 0.186384 ms
>>
>>  Statistics for all messages (161) more than 100 kB and less than 
>> 1000 kB in size
>>  Total duration: 115 ms
>>  Average duration: 0.714286 ms
>>
>>  Statistics for all messages (31) more than 1000 kB in size
>>  Total duration: 265 ms
>>  Average duration: 8.548387 ms
>>
>>
>> As you can see 0.4 is about twice as fast on messages less than 10 kB 
>> in size and about 10 times faster for messages larger than 100 kB. 
>> That is quite amazing!
>>
>> Please let me know if my benchmarking method is flawed in any way. 
>> I'm attaching the source code of the test tool.
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: server-dev-unsubscribe@james.apache.org
> For additional commands, e-mail: server-dev-help@james.apache.org
>



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


Mime
View raw message