The following message is a courtesy copy of an article that has been posted to bit.listserv.ibm-main,alt.folklore.computers as well.
[email protected] (Roy Hewitt) writes: > applications and systems strung together over 30 or 40 years. The z10s > are probably one of the most hi-tech bits of kit you'll have in your > machine room, and z/OS is pretty good too ;-).. But what gives the > mainframe such a bad name is usually the pile of 40 year apps stuck > together running on top of it and our resistance to change.. (oh, and > our morbid fascination with 3270!!) And why did we get this > way?.. well it's what I call the pair of "IBM's double-edged swords". re: http://www.garlic.com/~lynn/2009q.html#67 Now is time for banks to replace core system according to Accenture http://www.garlic.com/~lynn/2009q.html#68 Now is time for banks to replace core system according to Accenture we had been called in to consult with small client/server startup that wanted to payment transactions on their server; the startup also had this technology called "SSL" that they wanted to use. the result is now frequently called "electronic commerce". part of the deployment was something called a "payment gateway" ... some past posts http://www.garlic.com/~lynn/subnetwork.html#gateway ... which acts as intermediary between electronic commerce webservers on the internet and the payment infrastructure. The original "payment gateway" ... was an ha/cmp setup. at the time, although we had left ... we were still involved in various aspects of the ha/cmp product http://www.garlic.com/~lynn/subtopic.html#hacmp ... in fact, two people at the startup, responsible for the "commerce server" ... we had worked with earlier during our ha/cmp days ... this old post mentions a meeting in ellison's conference room that they were at http://www.garlic.com/~lynn/95.html#13 In any case, both for ha/cmp product ... some aspects of the payment gateway deployment ... and various of the large electronic commerce webserver farms ... that we were brought in to review ... we would comment that there had to be compensating procedures ... some of which would have been already there as part of a mainframe infrastructure. the description that we used was that many of the platforms in use had evolved up out of interactive environment and tended to default to presenting an error message to the user whenever anything went wrong ... and relied on the user to take corrective action. the platforms with a "batch" heritage ... tended to have a higher upfront learning curve and less user friendly ... but it was part of a paradigm ... that assumed that the person responsible for the program wasn't there ... and that there were a lot of heuristics which evolved over a 40yr period to automagically try and handle all possible things that might go wrong. there is a marine bumper sticker that is a take-off on "if it has to be positively, absolutely, delivered overnight" ... but "if it has to be positively, absolutely, destroyed overnight" ... this is things like large payrolls and other things ... "if it has to be positively, absolutely, run overnight" (and repeatedly, time-after-time). -- 40+yrs virtualization experience (since Jan68), online at home since Mar1970 ---------------------------------------------------------------------- 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
