I read the article and it did not contain any information concerning the size of the mainframe that is being compared to the grid. I have seen detailed analysis of grid computing versus the use of a parallel sysplex in z/OS and the "z" environment compared favorably when using its capabilities. This article talks about the mainframe using VSAM files and some IMS. It seems to me like they needed to be using less VSAM and more IMS, perhaps FP to get the throughput that was required. The mainframe has plenty of options for processing large amounts of data in parallel -- similar to what this customer achieved through the use of a grid -- that provide for faster response time.
Also, this article mentioned that the application changes were difficult and time consuming and that the newly written application could handle changes more quickly. This too could be handled with good application design on the mainframe. So -- what is your point -- this appears to be a project where everyone was comfortable with the programming environment of a grid and developed an application that maximized it strengths to produce a good result for the client. That is great for the client. My caution is to take this article and imply that anyone else would achieve the same result by using the same technology and that the same result could not be achieved using a mainframe. How about some basic math on the numbers presented -- one mainframe versus the grid of "more than 120 Dell servers" to produce an application that finished in 65% less time. Did they compare the cost and performance of three mainframes loosely coupled as a sysplex to determine what the performance would be? Let's see, three versus 120 and the three could probably outperform the 120? Don't get me wrong, I know that it is easy to second guess a solution. However, this company spent a lot of money on the completed solution. The article even says that money was in third place on their priority list. So, cost of the solution did not rule out the mainframe. It appears from the article that a single mainframe does not scale as well as 120 Dell servers. I think I could have made that leap of faith without an elaborate test. I walked away from the article with the impression that good technicians on this project did not want to use the mainframe and developed a good solution without it. However, this does not tell me that the mainframe could not have done this job equally as well had the project been composed of mainframe technicians that knew how to make the environment scale and perform equally as well as a grid. Just my bucks worth. So many words could not be worth two cents, could they? Tom Moulder <snip> Slashdot had this article today: http://linux.slashdot.org/linux/07/01/05/0538224.shtml "IBM touted 2006 as a resurgence year for the mainframe, but not so fast. At R.L. Polk and Co., one of the oldest automobile analytics firms in the U.S., an aging mainframe couldn't cut it, so the IT staff looked elsewhere. Their search led to a grid computing environment - more specifically, a grid computing environment running Linux on more than 120 Dell servers. The mainframe's still there, apparently, but after an internal comparison showed the Linux grid outperforming the mainframe by 70% with a 65% reduction in hardware costs, Polk seemed content banishing the big box to a dark, lonely corner for more medial tasks." With a link to: http://searchopensource.stage.techtarget.com/originalContent/0,289142,sid39_ gci1237399,00.html <snip> ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html