You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@directory.apache.org by Trustin Lee <tr...@gmail.com> on 2005/04/24 17:59:08 UTC

[mina] Interesting bench mark result: One Queue vs Three Queues

Hi,

I ran some benchmark related with MINA thread pool.  I attached the
code, so you could review it.

MINA thread pool uses three queues to buffer three parameters not
creating any event object JavaBeans.  I did because it takes longer to
allocate any extra objects, but here is a new result:

One Queue is 25% faster in sipte of extra object creation.

I ran this benchmark at Sun VM, Windows XP, AMD64 3200+.

Any other guys who can run this benchmark at other environments?

Thanks,
Trustin
-- 
what we call human nature is actually human habit
--
http://gleamynode.net/

Re: [mina] Interesting bench mark result: One Queue vs Three Queues

Posted by Trustin Lee <tr...@gmail.com>.
Hi,

Thanks for running the benchmark.  I optimized BaseThreadPool
according to this result.  I also ran with -verbose:gc option.  I see
a lot of incremental GC logs, but it seems to outperform two extra
queue operations.

Trustin

2005/4/25, Alex Karasulu <ao...@bellsouth.net>:
> Trustin Lee wrote:
> 
> >Hi,
> >
> >I ran some benchmark related with MINA thread pool.  I attached the
> >code, so you could review it.
> >
> >MINA thread pool uses three queues to buffer three parameters not
> >creating any event object JavaBeans.  I did because it takes longer to
> >allocate any extra objects, but here is a new result:
> >
> >One Queue is 25% faster in sipte of extra object creation.
> >
> >
> >
> Hi I ran the test as a black box and here were my results on a Linux
> machine running on a vanilla AMD 2600+ with 1 GB mem (333 Mhz).  Here
> are the results and more specific info on the test machine.
> 
> Results:
> 
> [akarasulu@newton target]$ java -classpath '.:mina-0.7-SNAPSHOT.jar' Main
> Warming up.
> Done.
> 8613
> 12940
> 
> Machine info ...
> 
> [akarasulu@newton target]$ uname -a
> Linux newton 2.6.11-1.14_FC3 #1 Thu Apr 7 19:23:49 EDT 2005 i686 athlon
> i386 GNU/Linux
> 
> [akarasulu@newton target]$ cat /proc/cpuinfo
> processor       : 0
> vendor_id       : AuthenticAMD
> cpu family      : 6
> model           : 8
> model name      : AMD Athlon(TM) XP 2600+
> stepping        : 1
> cpu MHz         : 2083.203
> cache size      : 256 KB
> fdiv_bug        : no
> hlt_bug         : no
> f00f_bug        : no
> coma_bug        : no
> fpu             : yes
> fpu_exception   : yes
> cpuid level     : 1
> wp              : yes
> flags           : fpu vme de pse tsc msr pae mce cx8 apic mtrr pge mca
> cmov patpse36 mmx fxsr sse pni syscall mmxext 3dnowext 3dnow
> bogomips        : 4128.76
> 
> HTH,
> Alex
> 
> 


-- 
what we call human nature is actually human habit
--
http://gleamynode.net/

Re: [mina] Interesting bench mark result: One Queue vs Three Queues

Posted by Alex Karasulu <ao...@bellsouth.net>.
Trustin Lee wrote:

>Hi,
>
>I ran some benchmark related with MINA thread pool.  I attached the
>code, so you could review it.
>
>MINA thread pool uses three queues to buffer three parameters not
>creating any event object JavaBeans.  I did because it takes longer to
>allocate any extra objects, but here is a new result:
>
>One Queue is 25% faster in sipte of extra object creation.
>
>  
>
Hi I ran the test as a black box and here were my results on a Linux 
machine running on a vanilla AMD 2600+ with 1 GB mem (333 Mhz).  Here 
are the results and more specific info on the test machine.

Results:

[akarasulu@newton target]$ java -classpath '.:mina-0.7-SNAPSHOT.jar' Main
Warming up.
Done.
8613
12940

Machine info ...

[akarasulu@newton target]$ uname -a
Linux newton 2.6.11-1.14_FC3 #1 Thu Apr 7 19:23:49 EDT 2005 i686 athlon 
i386 GNU/Linux

[akarasulu@newton target]$ cat /proc/cpuinfo
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 6
model           : 8
model name      : AMD Athlon(TM) XP 2600+
stepping        : 1
cpu MHz         : 2083.203
cache size      : 256 KB
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic mtrr pge mca 
cmov patpse36 mmx fxsr sse pni syscall mmxext 3dnowext 3dnow
bogomips        : 4128.76

HTH,
Alex

Re: [mina] Interesting bench mark result: One Queue vs Three Queues

Posted by Julien Vermillard <jv...@archean.fr>.
Linux pcjvermillard 2.6.10-9-amd64-k8 #1 Tue Mar 15 01:45:16 CET 2005
x86_64 GNU/Linux
CPU: AMD Athlon(tm) 64 Processor 3200+

java version "1.5.0_01"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_01-b08)
Java HotSpot(TM) 64-Bit Server VM (build 1.5.0_01-b08, mixed mode

Result :

Warming up.
Done.
4198
4624

Le lundi 25 avril 2005 à 00:59 +0900, Trustin Lee a écrit :
> Hi,
> 
> I ran some benchmark related with MINA thread pool.  I attached the
> code, so you could review it.
> 
> MINA thread pool uses three queues to buffer three parameters not
> creating any event object JavaBeans.  I did because it takes longer to
> allocate any extra objects, but here is a new result:
> 
> One Queue is 25% faster in sipte of extra object creation.
> 
> I ran this benchmark at Sun VM, Windows XP, AMD64 3200+.
> 
> Any other guys who can run this benchmark at other environments?
> 
> Thanks,
> Trustin