Hi

My patches were supposed to make this work a bit better but there are
still some issues.

It looks like the patches didn't work for multi-page documents, but at
the moment it appears as if they are working if you have multiple
page-sequences. So if possible, try breaking your document into multiple
page-sequences and see if that helps.

Also, if you have references (e.g. page-number-citations) to pages a
long way ahead, this will consume a lot of memory. E.g. if you are using
page counting ("Page 1 of 2134") then that's gonna break things in
memory.

These problems are all solvable however the new FOP design is probably
the best place to solve them :)

I'm hoping to look at a couple of the problems some time soon but I'm
very busy with some other XML stuff at the moment. However, once that's
done I'll have some very good "real world" data to test the performance
of FOP with.

Note that I speak only for FOP, I haven't played with XSLT performance
yet.

Regards
Mark

On Thu, 2001-09-13 at 02:52, [EMAIL PROTECTED] wrote:
> 
> We are seeing the same thing - roughly 60M is taken up for about 20-30
> pages.  We have increased the JVM max mem settings, but this is only a
> stopgap.
> -Lou
> 
> 
> 
> 
> 
> Darrel Riekhof <[EMAIL PROTECTED]> on 09/12/2001 12:41:36 PM
> 
> Please respond to [EMAIL PROTECTED]
> 
> To:   "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
> cc:
> 
> Subject:  fop/xalan memory usage
> 
> Has anyone found any effective techniques for keeping the memory
> requirements down for xalan/fop?
> 
> I was doing some testing yesterday with a 128m heap, and any report over 20
> something pages would produce out of memory exceptions in various places.
> I'm currently keeping the xml and fo in stream memory buffers, and
> considering writing those to disk instead to save some memory.  Any other
> ideas?
> 
> Darrel
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, email: [EMAIL PROTECTED]
> 
> 
> 
> 
> 
> 
> 
> 
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, email: [EMAIL PROTECTED]
> 

PGP signature

Reply via email to