Paul Murphy about Mainframe Linux
In Mainframe Linux Paul Murphy explains, why Linux on Mainframes isn´t a good idea (or as i tend to say: it´s utter bullshit to use something like that). It´s a must read for everyone, who considers such a move.
But furthermore, it´s a nice case study of IBM´s marketing at work. The succes of IBM is largely based on a central idea: Do not target the persons with knowledge. Or as Paul states:
These stories are intended to make Linux on zVM sound impressive - but both rely on a critical assumption: specifically that the reader knows little or nothing about relative system costs and performance.
Only under this circumstance the standard sales tactics of obfuscation and the mastery of the pseudo correlation can reach it´s target. They won´t go after the die-hard architect with 10 years experience in Unix. They go after the CIO, who has to fight more with budgeting than the newes technologies. Sometimes, they have luck with this tactics even with such obvious nonsense like Mainframe Linux. By the way … more often it´s more elaborated: A few weeks ago i saw a presentation of our loved blue competitor about Niagara and the pSeries. Without deep knowledge you hadn´t a chance to see that that this bluish presentation was a perfect case of FUD, full of overstreched facts, nonsensical comparisions, outright incorrect facts and so on. And this was only one presentation, i don´t want to know, how many other presentation are held by our beloved competition with the same amount of bullshitting. At least, i have an explanation for a lot of insane questions from customers who had a blue visit.
Back to the mainframes …. Paul gives a nice example: A company reduces the runtime of a job from 8 hours to an hour by moving from z/OS to Linux on a Mainframe. Looks as a business case for Linux. Well, not exacly. They´ve upgraded the mainframes, too:
IBM gave CALCULO two options moving to DB2 for z/OS or running DB2 under Linux on an Integrated Facility for Linux (IFL) engine in a new IBM z890 mainframe, replacing two older Multiprise 3000 H30 servers.
IBM doesn´t state, whats the effect of Linux and whats the effect of buying a new, vastly more powerful mainframe. Is this a business case for Linux or a business case for buying new iron? But it clearly seems to have the intention of putting Linux in to the good light of more performance. Pseudo correlation is a powerful marketing weapon …