andy's blog

Miko Matsumura, the Chief SOA Strategist for Software AG writes on his blog:

In my new role as Software AG’s Chief Strategist, I would like to define a broad technology trend that I refer to as “The Human Enterprise”.

Taking back the word “Enterprise”

My first order of business in establishing “The Human Enterprise” is to take back the word “Enterprise”.

"Virtualization is the mainframe for the 21st century," said Stephen Herrod, VMware's CTO.

When I read that quote, I did a double take. Throughout my entire career, the mainframe was something to laugh at, to denigrade, to replace. But now, it seems like everything old is new again. Everyone wants to be the new mainframe. Here is Larry Ellison talking about his plans post acquisition of Sun:

So, you signed up for Amazon EC2 Cloud services. Mazal Tov! Now what?

There are two interesting questions here.

1) How do you architect an application to take advantage of the grid?

2) How well does this Amazon EC2 grid scale?

One possible answer to the first question is to use the Open Source GridGain product. These guys have really been making good strides in proving an easy Java API to gridify your apps.

...
Most companies buy servers from the likes of Dell, Hewlett-Packard, IBM, or Sun Microsystems. But Google, which has hundreds of thousands of servers and considers running them part of its core expertise, designs and builds its own. Ben Jai, who designed many of Google's servers, unveiled a modern Google server before the hungry eyes of a technically sophisticated audience.

When I was a freshman at UIUC, I lived in a dorm with a guy who was also in the computer engineering program. One day he told me that he was working part time in Stephen Wolfram's lab. I nodded, because I had no clue who Stephen Wolfram was.

Well,

I gave the Amazon Clouds a whirl. I have to say it was quite enlightening.

First, I watched the video tutorial on YouTube. It is pretty good at walking you through the setup step by step.

I am constantly amazed by how the trading industry, especially the options and futures space, makes it so difficult to write systems that automate trading.

One thing computers really stink at is dealing with exceptions. Yet, the entire world of trading seems to be based on exceptions. And huge amounts of programming effort are spent dealing with handling these exceptions and writing convoluted programs to handle all the goofy cases.

It seems that cloud computing is the new SOA, but looking around at the current offerings, it feels like we are definitely in the 'kick the tires' stage.  I don't think the barriers are as much technological as they are logistical and organizational.  As mentioned in a recent InformationWeek article, Amazon can't even handle purchase orders in its EC2 business.  You can only pay for Amazon clouds with a credit card?  I guess Amazon still views buying cloud MIPS the same as buying a 'Marley and Me' DVD.

I've been trying to figure out how data mining is used in the trading space - it doesn't appear that the traditional vendors of ETL tools, such as Ab Initio, Data-Stage/Ascential, Informatica, etc.. are very dominant in this space.  Instead I am seeing companies such as GreenPlum and Vertica.  GreenPlum appears to be used by NASDAQ and NYSE (of course one of the board members of GreenPlum is also one of the founders of Archipelago...)  Netezza is another player that I heard mentioned while still in my traditional ETL days.