I agree the MS case studies are BS, I linked to them to show that anyone can link to marketing BS.
The point I'm trying to make is that so is the IBM one and the devil is in the details. I've tried to highlight where they have failed with their analysis below.
When I found the detailed data on the IBM case study, and I damn near laughed my ass off.
Check page #27 on this report:
http://www.ngsi.com/company/Value_Prop_for_IBM_i_Full_Report_2010.pdf
A couple of notes:
1. They use the latest software of IBM I, and compare it to Windows 2003 in a report that came out in 2010!!! Why are they using such an OLD operating system and database platform (SQL 2005). This is like comparing IBM I Power 7 box to a 286 running DOS. Of course IBM I would win when you make these kinds of comparisons. At this point in time SQL 2008 R2 and Server 2008 R2 were out at the time, they could have virtualized everything without buying VMWare, and setup a cluster with built in software to the OS. But instead they decide to purposely make IBM i shine by making terrible comparisons.
2. Why did they choose CAL licensing at these user volumes when processor licensing on Microsoft platforms are cheaper in these 4 scenarios they identified. Looks like they wanted to purposely exaggerate the numbers.
3. They get their Microsoft SQL editions confused, they state they used two editions (standard edition or enterprise edition). Well which one is it, big price difference between the two. Sadly at this point I think they are just making crap up when I seen this line. You would only buy one edition, not both which leads me to believe they were just making stuff up at this point.
4. This line also cracks me up:
Core database servers in Windows and Linux/Oracle server scenarios for the three largest installations were configured in two-way clusters using a common third-party failover solution.
Why in the hell would you buy third party software for this when it's built in to windows if they used the latest OS (Server 2008 R2 at the time) instead of Windows Server 2003, see point # 1 above.
5. They are picking # of FTE's out of thin air to manage the systems: .35 for IBM I, .65 for Windows, .75 for Linux.
6. The personnel costs they lumped database administration and system administration into a SINGLE JOB ROLE for IBM i. While at the same time splitting that job role out for the linux and windows platforms. WTF, this is NOT apples to apples comparison here. I'm sorry but an employee of any worth can do both of those functions on ANY platform (linux, windows, or i) and if you are going to compare them that way then you have to make them the same.
7. They state this:
Windows and Linux/Oracle servers for all installations were also configured with third-party system management and security tools providing functionality corresponding to that incorporated in i for no additional charge.
Why aren't they using the management, reporting, and alerting tools built into the OS on Windows and Linux?!?!
They are buying all these additional pieces of software because of some "perceived gap" in functionality in windows and linux, but they don't elaborate on why in the hell they are purchasing this software in the first place. Instead they make the blanket statement of "it's incorporated in IBM i for no additional cost". Are they saying you are unable to manage linux and windows security without third party tools, because that is a bunch of hog wash especially when you are talking about just 3 windows / linux servers in their example (wholesale distributor example).
Matt
-----Original Message-----
From: Nathan Andelin [mailto:nandelin@xxxxxxxxx]
Sent: Thursday, May 16, 2013 11:36 AM
To: Midrange Systems Technical Discussion
Subject: Re: cloud services with new development on IBM i on Power
Matt,
That PDF was just the "Executive Summary". The claims are substantiated in the "Detailed Data Section" of the report, not the Executive Summary. I haven't tried to track down the full report, but your allegation about it being a typical marketing document is not true.
Contrast that with the "Microsoft Case Studies" you referenced earlier in this thread. Typical marketing BS. The one cited a migration of Baan ERP off an AS/400 to a pure Microsoft stack. Fact is, Baan ERP does not, nor ever did run under OS/400. The claim that it was hosted on an AS/400 was pure SPIN.
How does Microsoft get away with this stuff? Baan once supported a number of DBMS via ODBC, but after a point dropped support for the OS/400 database. Baan customers were essentially forced to migrate their databases.
I agreed with one point. Supporting disparate technologies is costly. One might save money by migrating to a pure MS technology stack. If you wan't to save even more money, then migrate off MS to a pure IBM i stack.
-Nathan
----- Original Message -----
From: Matt Olson <Matt.Olson@xxxxxxxx>
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
Cc:
Sent: Wednesday, May 15, 2013 2:20 PM
Subject: RE: cloud services with new development on IBM i on Power
Here is what I mentioned a few weeks ago about this IBM marketing slide deck:
There are so many things wrong with that document I stopped counting. It's a typical marketing document with pretty graphs and no real specifics as to how they came up with those values.
When you look at the meat behind their pretty graphs you find out it's based off of IBM funded research papers from ITG. See here for one of the footnote references they used in slide #11, and look at the last page in the footnotes in that PDF file and you can see its funded by IBM:
http://www.ngsi.com/company/pol03062usen.pdf
There is NOTHING in that document to substantiate the claim of "44% less than x86, Microsoft Windows Server and Microsoft SQL Server", and "57% less than x86, Linux, Oracle DB".
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list To post a message email: MIDRANGE-L@xxxxxxxxxxxx To subscribe, unsubscribe, or change list options,
visit:
http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx Before posting, please take a moment to review the archives at
http://archive.midrange.com/midrange-l.
midrange.com
