Re: [docbook-apps] Dynamic docbook processing

["Bob Stayton" <bobs@sagehill.net>]

The memory usage scales to some degree with the content size.  Also, how much memory does AxKit take?  What XSL processor is AxKit using?  If Java based, then that has a lot of overhead.

 

If you are able to examine the size of the memory image, you should be able to do some experiments.  I would try:

 

a. Running a very small book and a large book, and see how much is overhead and how much scales with the content size.

 

b.  Turn off olink data collection and see if that makes a difference.

 

c.  Set the olink database name to null so it doesn't get loaded and see if that makes a difference (your olinks won't resolve, of course).

 

Bob Stayton
Sagehill Enterprises
DocBook Consulting
bobs@sagehill.net

 

 

----- Original Message -----

From: tom.kirkpatrick@virusbtn.com

To: docbook-apps

Sent: Friday, October 14, 2005 3:05 AM

Subject: [docbook-apps] Dynamic docbook processing

I am using the Apache::AxKit framework to dynamically transform my docbook articles on a per request basis. When I process a docbook document (by requesting it in my web browser) the apache process grows some 70mb in size. Is this normal?

I am currently running the sheets with the generate olink database parameter set to true, although I do intend on changing this as there really is no need to regenerate the database each time the article is viewed. Could it be the olink database generation code that is taking up so much of my ram?

Is there anything I can do to a) conserve memory, and b) speed things up a little?

thanks in advance

--
Tom Kirkpatrick
Web Developer - Virus Bulletin