OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   Re: [xml-dev] Validation vs performance - was Re: [xml-dev] Fast text ou

[ Lists Home | Date Index | Thread Index ]

I've created one online application used by many millions of people that 
processed thousands of messages per second (probably about 2500-6000), 
with database retrievals for many of those at up to 1700 per second, and 
simultaneous control of over 300,000 remote clients, with a single 
process on a single processor of a 96Mhz HPUX server.

A couple years later I architected and helped build an application that, 
on a 4 processor high-end Sun Sparc system was able to do a transaction 
every 2 seconds at full speed.  I knew the latter would be slow for a 
variety of reasons and began working on better architectures that led to 
esXML, i.e supported the elegance and sophistication of the latter 
application with the efficiency approaching the former.  These two 
extremes represent, to a certain extent, the difference between 
different extremes in overhead, abstraction, communication model, etc.

Processing overhead, including the major components of parsing / object 
creation / data copies / serialization, is not a 'future problem'.  It 
has always been a problem.  Network bandwidth is not much of an issue, 
just like disk storage isn't much of an issue anymore.  Certainly it 
would be helpfull, and for certain corners of the market like mobile 
phones, more usefull, although less than it used to be.  The scarce 
resource is time.  Anything that eats time is bad.  This could be 
bandwidth usage, CPU, memory, or suboptimal communication and semantic 


David Megginson wrote:

> My memory gets hazy sometimes, but I think I remember the networking 
> people
> complaining just as much a decade and a half ago about the 
> inefficiencies of
> IP, TCP, etc.  I had to fight hard as late as 1993-94 to get a university
> lab to use TCP/IP because the administration had been convinced by their
> vendor (Novell) that TCP/IP was too slow for serious work.
> I think that we might be at the same place right now with XML.  The
> self-annointed analysts are telling the networking people that they'll 
> have
> to handle enormous volumes of XML network traffic in a few years, and the
> networking people are freaking out over hypothetical future problems like
> verbosity and parsing time.  I don't think that we have much of a clue 
> yet
> whether (a) there actually will be much XML network traffic over the next
> few years, or (b) what it might look like, so any optimization is waaaay
> premature.
> All the best,
> David
> -----------------------------------------------------------------
> The xml-dev list is sponsored by XML.org <http://www.xml.org>, an
> initiative of OASIS <http://www.oasis-open.org>
> The list archives are at http://lists.xml.org/archives/xml-dev/
> To subscribe or unsubscribe from this list use the subscription
> manager: <http://www.oasis-open.org/mlmanage/index.php>

swilliams@hpti.com http://www.hpti.com Per: sdw@lig.net http://sdw.st
Stephen D. Williams 703-724-0118W 703-995-0407Fax 20147-4622 AIM: sdw

fn:Stephen Williams


News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS