By Steve Ranger, 4 July 2005 12:50
When risk management organisation Lloyd's Register started on an exercise to choose a single operating platform from Windows, Linux and Unix, it didn't expect to end up plumping for the Microsoft option.
The organisation, which operates from 234 offices in 120 countries, began life examining the sea-worthiness of merchant ships but has now expanded into other areas such as land-based industries and oil and gas.
Recently the organisation conducted a feasibility study to see if a consolidated environment should exist at the server level.
Stephen Hand, group IT director at Lloyd's Register told silicon.com that he inherited a very fragmented IT infrastructure when he joined.
"We were very unhappy with total cost of ownership and were looking for reductions across the board. We concluded that a consolidated operating platform was something that we should be doing," he explained.
In the feasibility study each operating system was given a score out of 140. And even though it was expected that Linux would come out top, this wasn't the case.
"Our preconceptions about where the exercise was going were wrong. We thought we'd be open source by now," Hand admits.
On its tests Windows 2003 scored 120 out of a possible 140, with RedHat Linux scoring 90 and HP Unix 86.
Hand said: "We decided Windows 2003 was the direction we should be going in."
In particular what swung Lloyd's Register towards the Microsoft option were issues such as interoperability and availability of staff - and the organisation's future plans to move to a hosted environment.
"We've been moving to the Windows platform for the last 18 months. We decided to take a pragmatic approach to migration," Hand said.
The organisation isn't planning an overnight migration but will move systems over when there is a good reason, he said: "We have a roadmap over the next two to three years. What we've decided to do is replace applications [when there is a] business reason for renewal."
As a result the organisation will slash the number of servers it has to manage from more than 70 to around 20. And the move to a Microsoft infrastructure has allowed it to cut a third of its IT spend over the last three years, with operational spend dropping from £24m to £16m.
Comments
There are 10 comments. Join the discussion
1. anonymous
Just a few questions:
1) What were the specific criteria
for the feasibility study?
2) How were the "operating systems"
evaluated based on the criteria?
Were they evaluated on a "theoretical academic thought experiment" or on
actual reproducible experimental tests?
3) Did Lloyd's receive a discount in
costs from Microsoft before or after
conducting its study? Did Lloyd's receive
a discount after mentioning the possibility of switching to Linux?
2. anonymous
Interesting, I would like to know how the OS's were scored. This seems a very suprising result (how come unix and linux got different scores when according to SCO linux is a rip off of unix (which we all know it is)!!!
Still just another victory for the MS lovers (yipee).
3. anonymous
Heh, I wonder if the author of this article gets any roaylties from Microsoft for posting this stuff.
I like it how the article starts with the dimension of "Lloyd's Register" and then moves to the "facts".
Lloyd's Register conclusion of migrating to Windows 2003 means absolutely nothing. It does not say which of the two OS's are better, yet the article tries to show Win2k3 in a better light.
(Ed note. No it doesn't, it's a case study - it's not a comprehensive evaluation, it's one company's opinion and coverage of the specific issues as they relate to that company. Which bit of the word 'case' - as in 'case by case' - was lost on you? Hence that is why the article began with a mention of Lloyds Register, because the case study was about... you guessed it, Lloyds Register.)
It is obvious that they found that their staff lacks the skills and the software they use does not work on Linux/Unix, and not because they found Windows more stabler or better.
Anyways, congratulation, I wish you all the best.. oh, and don't forget to restart those servers every month or two !
4. Paul Hubert
It would help to see their score sheet!
5. Paul Hubert
It would help to see their score sheet!
6. Steve
Pfft. No effing way. This is another "we needed Word because we lost the nerve to as our suppliers for another format," combined with "our MCSEs paniced and said they don't know anything about (anything?) linux."
FUD. Paid FUD, too, I'm guessing.
7. anonymous
This is not a comment on the article or reply, but on the Ed. note posted in the message.
I found the Ed. note's sarcasm "Which bit of the word 'case' - as in 'case by case' - was lost on you?" to be unprofessional. I do not believe that type of comment has a place in a professional organization.
8. Benny
I really don't get the anti-Microsoft animus. A simple story about one company's evaluation and decision and it raises claims of journalistic bribes? AT&T was a monopoly for almost a hundred years, but I doubt they generated the visceral hatred that is directed at Microsoft. They are hard on competitors, but so was (current) Linux-friend IBM when it was sued for monopoly back in the day.
I guess I don't understand why people identify so closely with their operating system of choice. Very strange.
9. Gabe
Funny, a company that focuses on risk management couldn't even measure their own risks! How about the hidden costs of Windows like lack of stability, poor security model and more support and licensing costs? I guess the ancient insurer knows more about seafaring then modern things like IT!
10. anonymous
Knowing Linux and Windows, it's pretty safe to say that ANY report that shows lower TCO for Windows is not to be trusted. TCO was invented as a metric to point out the fact that PC's were far more expensive over time than the basic purchase price. Virtually all the factors that give the PC a high TCO are much worse under Windows than Linux. Namely, Windows PC's are harder to manage centrally, harder to secure generally, require more in the way of license fees, less reliable overall, (though Win2K and WinXP are trending in a better direction, finally), and generally just more of a hassle to deal with. It is harder to find people with Linux skills, but once found, the number of Linux PC's that can be supported by one Linux person is greater, so the trend is still in Linux's favor.
With Linux and Windows starting from an even plane, Linux will beat Windows in TCO every time.
The problem is that no company starts from scratch. They start from where they are. This particular case study gives the game away when they mention they were already in the process of migrating everyone to Windows. So what the study really suggests is that they thought it was easier to keep going that way rather than reverse course and move to something else.
But these studies that say Windows has lower TCO are generally lumping in migration costs with actual TCO estimates.
Bottom line: never believe the answer from a study unless they "show you their work" to derive the answer.
After all, no matter how much it may sound like "conspiracy theory" thinking, the fact remains that Microsoft has a vested financial interest in the results of these studies, and regards Linux as Enemy Number One.