You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tomee.apache.org by "Jacek Laskowski (JIRA)" <ji...@apache.org> on 2006/10/26 17:07:19 UTC
[jira] Resolved: (OPENEJB-295) o.a.o.client.ServerMetaData
optimization wrt InetAddress.getLocalHost
[ http://issues.apache.org/jira/browse/OPENEJB-295?page=all ]
Jacek Laskowski resolved OPENEJB-295.
-------------------------------------
Fix Version/s: 2.2
Resolution: Fixed
Assignee: Jacek Laskowski
Sending ServerMetaData.java
Transmitting file data .
Committed revision 468031.
Thanks Mark!
A (friendly) note: please run svn diff at the top-level directory of OpenEJB sources.
> o.a.o.client.ServerMetaData optimization wrt InetAddress.getLocalHost
> ---------------------------------------------------------------------
>
> Key: OPENEJB-295
> URL: http://issues.apache.org/jira/browse/OPENEJB-295
> Project: OpenEJB
> Issue Type: Improvement
> Components: general
> Affects Versions: 2.2
> Environment: Windows XP
> Reporter: Mark DeLaFranier
> Assigned To: Jacek Laskowski
> Priority: Minor
> Fix For: 2.2
>
> Attachments: OpenEJB-295.patch
>
>
> While running OptimizeIt on a small test client, I stumbled across the call to "InetAddress.getLocalHost();" which was taking 33% in a client test. In this test the client was creating several copies of an InitialContext object rather than sharing one.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
Re: [jira] Resolved: (OPENEJB-295) o.a.o.client.ServerMetaData optimization
wrt InetAddress.getLocalHost
Posted by Mark DeLaFranier <de...@hotmail.com>.
Hey David,
I am not really working on anything specific, I was just curious as to
how many calls per second a client could make to a stateless session
bean. I could never get more than ~4000 ejb calls. I was sure it
wasn't a configuration issue since the first 4000 calls worked. I
discovered that there was a tonne of sockets in a "TIME_WAIT" state on
the server side, which didn't seem right. Eventually, I found that the
client was doing the open-write-read-close. I wrote a client/server
program that did exactly the same and sure enough after about ~4000
calls my sample demonstrated the same behaviour. After some more
research, I learned that windows reserves the range 1024 > n < 5000 for
the "dynamic" ports (not sure what the correct term is). This is
co-incided with the connection limits I was seeing.
Anyway, I just thought that could be better so I made some changes. :-)
I'd like to slowly complete the code and allow user configuration and
then I'd had it over to the OpenEJB team for review.
Thanks
Mark
David Blevins wrote:
> On Oct 26, 2006, at 8:31 PM, Mark DeLaFranier wrote:
>
>> Jacek,
>>
>> I was going to do a top level diff, but it had lots of other diffs
>> due to having other modified files. To reduce the complexity of the
>> diff, I only diff'd the ServerMetaData.java file.
>>
>> I am certainly not a svn experty :-), but a current diff shows me:
>>
>> > svn diff | wc
>> 1189 5247 49331
>>
>> Thats quite a lot to filter out and make your job easier of assessing
>> my patch.
>>
>> What are your thoughts?
>
> That's totally fine, Mark. It only takes a second to figure out where
> they go. Thanks for all the time you put into tracking this
> performance issue down!
>
> You've got me curious about what you're working on that has you
> running OptimizeIt on the client :) No sweat of you're not at liberty
> to share too many details -- we're always happy to be on the receiving
> end of any kind of work people do with OpenEJB.
>
> Thanks,
> David
>
>
>> Thanks
>> Mark
>>
>> Jacek Laskowski (JIRA) wrote:
>>> [ http://issues.apache.org/jira/browse/OPENEJB-295?page=all ]
>>>
>>> Jacek Laskowski resolved OPENEJB-295.
>>> -------------------------------------
>>>
>>> Fix Version/s: 2.2
>>> Resolution: Fixed
>>> Assignee: Jacek Laskowski
>>>
>>> Sending ServerMetaData.java
>>> Transmitting file data .
>>> Committed revision 468031.
>>>
>>> Thanks Mark!
>>>
>>> A (friendly) note: please run svn diff at the top-level directory of
>>> OpenEJB sources.
>>>
>>>
>>>> o.a.o.client.ServerMetaData optimization wrt InetAddress.getLocalHost
>>>> ---------------------------------------------------------------------
>>>>
>>>> Key: OPENEJB-295
>>>> URL: http://issues.apache.org/jira/browse/OPENEJB-295
>>>> Project: OpenEJB
>>>> Issue Type: Improvement
>>>> Components: general
>>>> Affects Versions: 2.2
>>>> Environment: Windows XP
>>>> Reporter: Mark DeLaFranier
>>>> Assigned To: Jacek Laskowski
>>>> Priority: Minor
>>>> Fix For: 2.2
>>>>
>>>> Attachments: OpenEJB-295.patch
>>>>
>>>>
>>>> While running OptimizeIt on a small test client, I stumbled across
>>>> the call to "InetAddress.getLocalHost();" which was taking 33% in a
>>>> client test. In this test the client was creating several copies
>>>> of an InitialContext object rather than sharing one.
>>>>
>>>
>>>
>>
>>
>
>
>
Re: [jira] Resolved: (OPENEJB-295) o.a.o.client.ServerMetaData optimization wrt InetAddress.getLocalHost
Posted by David Blevins <da...@visi.com>.
On Oct 26, 2006, at 8:31 PM, Mark DeLaFranier wrote:
> Jacek,
>
> I was going to do a top level diff, but it had lots of other diffs
> due to having other modified files. To reduce the complexity of
> the diff, I only diff'd the ServerMetaData.java file.
>
> I am certainly not a svn experty :-), but a current diff shows me:
>
> > svn diff | wc
> 1189 5247 49331
>
> Thats quite a lot to filter out and make your job easier of
> assessing my patch.
>
> What are your thoughts?
That's totally fine, Mark. It only takes a second to figure out
where they go. Thanks for all the time you put into tracking this
performance issue down!
You've got me curious about what you're working on that has you
running OptimizeIt on the client :) No sweat of you're not at
liberty to share too many details -- we're always happy to be on the
receiving end of any kind of work people do with OpenEJB.
Thanks,
David
> Thanks
> Mark
>
> Jacek Laskowski (JIRA) wrote:
>> [ http://issues.apache.org/jira/browse/OPENEJB-295?page=all ]
>>
>> Jacek Laskowski resolved OPENEJB-295.
>> -------------------------------------
>>
>> Fix Version/s: 2.2
>> Resolution: Fixed
>> Assignee: Jacek Laskowski
>>
>> Sending ServerMetaData.java
>> Transmitting file data .
>> Committed revision 468031.
>>
>> Thanks Mark!
>>
>> A (friendly) note: please run svn diff at the top-level directory
>> of OpenEJB sources.
>>
>>
>>> o.a.o.client.ServerMetaData optimization wrt
>>> InetAddress.getLocalHost
>>> --------------------------------------------------------------------
>>> -
>>>
>>> Key: OPENEJB-295
>>> URL: http://issues.apache.org/jira/browse/
>>> OPENEJB-295
>>> Project: OpenEJB
>>> Issue Type: Improvement
>>> Components: general
>>> Affects Versions: 2.2
>>> Environment: Windows XP
>>> Reporter: Mark DeLaFranier
>>> Assigned To: Jacek Laskowski
>>> Priority: Minor
>>> Fix For: 2.2
>>>
>>> Attachments: OpenEJB-295.patch
>>>
>>>
>>> While running OptimizeIt on a small test client, I stumbled
>>> across the call to "InetAddress.getLocalHost();" which was taking
>>> 33% in a client test. In this test the client was creating
>>> several copies of an InitialContext object rather than sharing one.
>>>
>>
>>
>
>
Re: [jira] Resolved: (OPENEJB-295) o.a.o.client.ServerMetaData optimization
wrt InetAddress.getLocalHost
Posted by Mark DeLaFranier <de...@hotmail.com>.
Jacek,
I was going to do a top level diff, but it had lots of other diffs due
to having other modified files. To reduce the complexity of the diff, I
only diff'd the ServerMetaData.java file.
I am certainly not a svn experty :-), but a current diff shows me:
> svn diff | wc
1189 5247 49331
Thats quite a lot to filter out and make your job easier of assessing my
patch.
What are your thoughts?
Thanks
Mark
Jacek Laskowski (JIRA) wrote:
> [ http://issues.apache.org/jira/browse/OPENEJB-295?page=all ]
>
> Jacek Laskowski resolved OPENEJB-295.
> -------------------------------------
>
> Fix Version/s: 2.2
> Resolution: Fixed
> Assignee: Jacek Laskowski
>
> Sending ServerMetaData.java
> Transmitting file data .
> Committed revision 468031.
>
> Thanks Mark!
>
> A (friendly) note: please run svn diff at the top-level directory of OpenEJB sources.
>
>
>> o.a.o.client.ServerMetaData optimization wrt InetAddress.getLocalHost
>> ---------------------------------------------------------------------
>>
>> Key: OPENEJB-295
>> URL: http://issues.apache.org/jira/browse/OPENEJB-295
>> Project: OpenEJB
>> Issue Type: Improvement
>> Components: general
>> Affects Versions: 2.2
>> Environment: Windows XP
>> Reporter: Mark DeLaFranier
>> Assigned To: Jacek Laskowski
>> Priority: Minor
>> Fix For: 2.2
>>
>> Attachments: OpenEJB-295.patch
>>
>>
>> While running OptimizeIt on a small test client, I stumbled across the call to "InetAddress.getLocalHost();" which was taking 33% in a client test. In this test the client was creating several copies of an InitialContext object rather than sharing one.
>>
>
>