They call this FUD
Stephen Forte’s Testimony to the New York City Council, April 29, 2003
Thank you all today for taking time to hear my testimony. My name is Stephen Forte, I was born and raised in New York City and am 31 years old. At the age of 23 I founded a high-tech consulting firm called The Aurora Development Group, which was sold 5 years later. I also served as the Chief Technology Officer of Zagat Survey here in New York from late 1999 until January 2002. Last April I co-founded Corzen, based up at Union Square where I currently serve as Chief Technology Officer.
I have had to do the economic and technical analysis of whether to use Open Source in my operations twice, once at Zagat where we had a 5 million dollar technology budget but not enough time and money to meet our deadlines for an IPO and second when I founded Corzen last year with only $300,000 of initial investment. Time to market and saving money was very important at Corzen epically since I did not get a paycheck until December 2002.
As a small business owner and resident of the City of New York for over 31 years, I appreciate the magnitude of the current budget shortfall. It may be tempting to make a blanket policy stating that the City must only use Open Source software to save money. On the surface Microsoft Windows, for a typical server machine configuration costs approximately $6,000. Linux costs nothing. Surely Linux is cheaper. Isn't it?
On the surface it appears that way. But once you dive into the details you will see that Open Source is not free and that while it may have a place in your organization as well as mine when the technology deems fit, there should be no blanket “Open Source” only policy. This is a policy I would strongly urge the council not to spend any taxpayer time considering. Here is why.
A benchmark recently performed by the non-profit TMC (www.tmc.org) compared a Linux machine running IBM’s middle tier software WebLogic/ WebSphere compared to a Windows Server machine running Microsoft middle tier software .NET. At this moment in time there is no viable open source “middle tier” component to compete with the IBM or Microsoft offerings. Just a technical note, the middle tier is what makes your custom applications work: the web server, the application server, the runtime environments and the programming languages.
The TMC broke down the cost of the server machine into three components: the hardware, the operating system (Windows or Linux), and the infrastructure (.NET, WebLogic, or WebSphere). Since the hardware was the same in all benchmarks, the differentiating costs were due to the operating system and the infrastructure.
It turns out that the cost of the operating system is relatively insignificant in the overall server costs. Of the total WebLogic server cost of $76,990, only $5,990 was attributable to Windows. Of the total WebSphere server cost of $84,990, again, only $5,990 was for Windows. In neither case was the cost of Windows more than 8% of the total server cost.
However the use of Linux does have one dramatic cost consequence. It eliminates the possibility of using .NET as the application server infrastructure. Since Linux does not have a comparable infrastructure, the use of Linux thus dictates the use of either WebSphere or WebLogic. While you are paying for Windows Server, the middle tier is built right into it so the costs of the middle tier are quite low when factored in as part of the cost of the Operating System.
While the cost of the operating system is relatively insignificant in the overall server cost equation, the cost of the infrastructure is not at all insignificant. WebLogic costs $40,000. WebSphere costs $48,000. This is as much as the server hardware! .NET, in contrast, cost nothing. It is part of the Windows operating system. The bottom line is that in order to save $5,990 of operating system costs, one must lock oneself into a "free" operating system that will require an additional $40-48,000 for server infrastructure!
TMC did not include the cost of the database in their benchmark cost analysis. This was a rather odd decision, given that the database is just as critical to this benchmark as is the operating system or the infrastructure. Had TMC included the cost of the database, the cost analysis would have further favored .NET, given that Microsoft SQLServer is much less expensive than either Oracle or DB2.
I am not advocating the use of Microsoft .NET over IBM or Windows over Linux, what I am saying is that the cost of open source software is not free, and for enterprise wide applications not even less expensive then a comparable non-open source platform.
When I was the CTO of Zagat Survey I had to do similar analysis. Preparing for an IPO, we were not scalable in our web site costs. Meaning that as the total amount of users went up, the cost per user was not going down. I was being pressured by the Board of Directors, one of which was the founder of Sun Microsystems to use the Linix/Java/J2EE approach on our Web Site. When I came in as CTO I inherited a site running on the open source platform using the Java platform. The site had cost us $650,000. The problem with the site was that for what we paid for we got only the “bare bones”, the vendor told us to increase traffic a factor of 10X, we would have to spend an additional 5 million dollars. My own analysis supported this view as did the Venture Capitalists who invested 34 million dollars in the company. In addition the total cost per user would go UP, not down!
I concluded after months of research that I can increase traffic by 20x and decrease the total cost per user by 1000% by not using open source technology. In a meeting on June 13th, 2000 with the board of directors I justified my approach based on my open source analysis:
“Everybody knows that we must have scalability if we are to build a web site that we can go public with. Surprisingly, very few people have any idea what scalability actually means. I consider a system to be "scalable" if we can add more workload to the system without increasing the cost of the system per unit of workload. The common unit of workload for a web site system is a transaction. If the commerce system is an on-line store, then the transaction equates to an actual sale. If the commerce system is a bank, then the transaction equates to a banking transaction. In the case of Zagat.com a transaction is a pure database read, which is almost each and every page.”
I went on to conclude that the June 2000 version of our web site which cost $650,000 broke down after 300 concurrent users (nobody denied this, this was the reason why I was before the BOD.) I proposed a “Microsoft” solution for 1.2 million dollars, not the 5 million. William Ford of General Atlantic Partners said to me, never have I offered a CTO 5 million dollars and he came back ask said he only wanted 1.2.
I got permission to build the site and as development took place from July to December 2000, Zagat’s needs grew, so we had to add capacity. We spent only $300,000 over our budget of 1.2 million (compared to 5 million) and achieved 20x user base and 1000% decrease in cost per customer. In December 2000 the site launched and has not been “down” since. That is 2.5 years of uptime for a real cheap price tag.
Exhibit 1. Total Cost Per User at Zagat.com
Total Cost Per Customer |
|
|
|
|
|
June 2000 (Java) |
Proposed |
December 2000 |
Future Plans |
Hardware |
$80,000.00 |
$210,000.00 |
$300,000.00 |
$420,000.00 |
Hosting |
$90,000.00 |
$168,000.00 |
$264,000.00 |
$528,000.00 |
Human Resources |
$480,000.00 |
$700,000.00 |
$800,000.00 |
$900,000.00 |
Software |
$0.00 |
$150,000.00 |
$200,000.00 |
$250,000.00 |
Total System Cost (Annual) |
$650,000.00 |
$1,228,000.00 |
$1,564,000.00 |
$2,098,000.00 |
Max Simultaneous Customers |
300 |
6,000 |
10,000 |
20,000 |
Total Cost Per Customer |
$2,167 |
$205 |
$156 |
$105 |
The Open Source community is dynamic and exciting. There are pieces of my architecture both at Zagat and at Corzen where I use Open Source products (SendMail for instance). I would like to conclude with a warning. Unix was plagued by different “flavors” and “distributions” that dramatically increased TCO. This fragmentation is now occurring in the Linux space. Take a look at this quote:
"’Enterprises now realize that they are writing to a distribution, not to Linux in general. What works on Red Hat Advanced Server will not work on SuSE Linux,’ Schwartz [Sun's executive vice president of software] said. ... There is little doubt that the notion of ‘Linux and free have gone away. Red Hat's pricing model now makes that clear,’ he said.”
- Sun Drops Its Linux Distribution, in eWeek, March 28, 2003By Peter Galli http://www.eweek.com/article2/0,3959,981455,00.asp
I would like to thank the council for taking the time to hear my testimony today.