ws-xmlrpc-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From nick.ebb...@uk.bnpparibas.com
Subject Re: Fw: [PATCH] suggested patches for memory issues
Date Thu, 06 Oct 2005 09:30:53 GMT
Thanks Jochen,
(and John for your previous comments)

I take your point that this is application specific, and if older JVMs do
not release interned Strings this is a strong reason not to intern by
default.
I will put in an alternative factory, leaving the default behaviour as it
is currently (no interning) and send you the diffs when this is done.

I think it might be a good idea for the implementation to write something
in the logs/standard out when the factory is created, to indicate whether
String intern is being used or not
 - since it seems an important choice for the developer to make and this
might help to bring the issue into focus.

Nick




Extranet
jochen.wiedmann@gmail.com - 05/10/2005 21:09


Please respond to xmlrpc-dev@ws.apache.org



To:    xmlrpc-dev

cc:


Subject:    Re: Fw: [PATCH] suggested patches for memory issues


nick.ebbutt@uk.bnpparibas.com wrote:

> Our application will run with approximately 1/5th of the previous memory
> with the interns in (which is the difference between total success of the
> project and total failure)

This indicates, that your application has a lot of hits in the map of
intern'd strings. In other words, intern() is good for you.  (Which I
wasn't discussing.) Other applications will not have such a lot of hits.
In other words, they will increase the internal map without any reason:
The achieved effect may be quite the converse of the desired.

Interning is an application specific optimization. In general, it pays
to leave the JVM doing its work without enforcing certain optimizations.
I do not believe, that a general purpose library like XML-RPC should do
things like that. That is my personal experience, which may not be
yours, of course. However, note that Johns previous posting has already
suggested that at least some JVM's may have a totally different
behaviour than what you would expect. Besides, if interning would always
help, then it could be made a JVM option.

I am still convinced that the defaults should be left as they are.
However, your numbers are clearly indicating that there are good reasons
for creating a new option.


 Jochen



This message and any attachments (the "message") is 
intended solely for the addressees and is confidential. 
If you receive this message in error, please delete it and 
immediately notify the sender. Any use not in accord with
its purpose, any dissemination or disclosure, either whole 
or partial, is prohibited except formal approval. The internet 
can not guarantee the integrity of this message. 
BNP PARIBAS (and its subsidiaries) shall (will) not 
therefore be liable for the message if modified. 

**********************************************************************************************

BNP Paribas Private Bank London Branch is authorised 
by CECEI & AMF and is regulated by the Financial Services
Authority for the conduct of its investment business in the
United Kingdom.

BNP Paribas Securities Services London Branch is authorised
by CECEI & AMF and is regulated by the Financial Services
Authority for the conduct of its investment business in the 
United Kingdom.
  
BNP Paribas Fund Services UK Limited is authorised and 
regulated by the Financial Services Authority.

Mime
View raw message