Slashdot Mirror
The Mainframe Is Dead! Long Live the Mainframe!
HughPickens.com writes The death of the mainframe has been predicted many times over the years but it has prevailed because it has been overhauled time and again. Now Steve Lohr reports that IBM has just released the z13, a new mainframe engineered to cope with the huge volume of data and transactions generated by people using smartphones and tablets. "This is a mainframe for the mobile digital economy," says Tom Rosamilia. "It's a computer for the bow wave of mobile transactions coming our way." IBM claims the z13 mainframe is the first system able to process 2.5 billion transactions a day and has a host of technical improvements over its predecessor, including three times the memory, faster processing and greater data-handling capability. IBM spent $1 billion to develop the z13, and that research generated 500 new patents, including some for encryption intended to improve the security of mobile computing. Much of the new technology is designed for real-time analysis in business. For example, the mainframe system can allow automated fraud prevention while a purchase is being made on a smartphone. Another example would be providing shoppers with personalized offers while they are in a store, by tracking their locations and tapping data on their preferences, mainly from their previous buying patterns at that retailer.
IBM brings out a new mainframe about every three years, and the success of this one is critical to the company's business. Mainframes alone account for only about 3 percent of IBM's sales. But when mainframe-related software, services and storage are included, the business as a whole contributes 25 percent of IBM's revenue and 35 percent of its operating profit. Ronald J. Peri, chief executive of Radixx International was an early advocate in the 1980s of moving off mainframes and onto networks of personal computers. Today Peri is shifting the back-end computing engine in the Radixx data center from a cluster of industry-standard servers to a new IBM mainframe and estimates the total cost of ownership including hardware, software and labor will be 50 percent less with a mainframe. "We kind of rediscovered the mainframe," says Peri.
IBM brings out a new mainframe about every three years, and the success of this one is critical to the company's business. Mainframes alone account for only about 3 percent of IBM's sales. But when mainframe-related software, services and storage are included, the business as a whole contributes 25 percent of IBM's revenue and 35 percent of its operating profit. Ronald J. Peri, chief executive of Radixx International was an early advocate in the 1980s of moving off mainframes and onto networks of personal computers. Today Peri is shifting the back-end computing engine in the Radixx data center from a cluster of industry-standard servers to a new IBM mainframe and estimates the total cost of ownership including hardware, software and labor will be 50 percent less with a mainframe. "We kind of rediscovered the mainframe," says Peri.
There was once a programmer who wrote software for personal computers. "Look at how well off I am here," he said to a mainframe programmer who came to visit. "I have my own operating system and file storage device. I do not have to share my resources with anyone. The software is self-consistent and easy-to-use. Why do you not quit your present job and join me here?"
The mainframe programmer then began to describe his system to his friend, saying, "The mainframe sits like an ancient Sage meditating in the midst of the Data Center. Its disk drives lie end-to- end like a great ocean of machinery. The software is as multifaceted as a diamond, and as convoluted as a primeval jungle. The programs, each unique, move through the system like a swift-flowing river. That is why I am happy where I am."
The personal computer programmer, upon hearing this, fell silent. But the two programmers remained friends until the end of their days.
"First they came for the slanderers and i said nothing."
The IBM pricing really is quite high (there are a ton of licensing fees for the hardware, maintenance, and software). But the systems work reliably. You get a giant system that can run a whole lot of VMs, with fast and reliable interconnects, transparent hardware failover (e.g. CPUs inside most mainframes come in redundant pairs), etc. To get a similar setup on commodity hardware you need some kind of "cloud" orchestration environment, like OpenStack, which can deal with VM management and migration, network storage, communication topology, etc. The advantage of an x86-64/OpenStack cluster solution is that the hardware+licensing costs are loads cheaper, and you don't have IBM levels of vendor lockin. The disadvantage is that it doesn't really work reliably; you're not going to get 5 9s of uptime on any significantly sized OpenStack deployment, and it will require an army of devops people to babysit it. The application complexity also tends to be higher, because failures are handled at the application level rather than at the system level: all your services need to be able to deal with non-transparent failover, split-brain scenarios, etc. Also the I/O interconnects between parts of the system (even if you're on 10GigE) are much worse than mainframe interconnects.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
For example, the mainframe system can allow automated fraud prevention while a purchase is being made on a smartphone.
Because that's so much different than preventing fraud on a purchase being made from a desktop PC.
From a business point of view they can be similar.
From the perspective of the mainframe guys, the whole point of a mainframe is that it is a single machine handling all of your transactions. Basically, it is simpler to deal with all kinds of transaction problems when you are not using a vastly distributed system with thousands of nodes. Typically PaaS/SaaS are large distributed systems.
To reliably and consistantly handle a very large stream of very important transactions where you basically need 100% reliability, they are a real option. The business case for a mainframe is something like, it would cost 200mln per year for some bank to make a failure proof distributed system, and 100mln to do it with a a mainframe. Outside of this type of systems, it is hard to think of any use for a mainframe, given the cost and complexity.
.... the more they stay the same. :)
I keep telling my friends that "cloud computing" is not a new concept. We used to call them "dumb terminals." Not a precise analogy of course but close enough for our purposes. You just know that's going to come full circle in another decade or so.
I want peace on earth and goodwill toward man.
We are the United States Government! We don't do that sort of thing.
No. PaaS is scale-out. while a mainframe is scale-up. A scale-out architecture is good at processing a lot of different requests, but does not offer very good results for high-frequency complex operations because by nature the distribution of workloads over a large network is costly. Anything similar to Newton's method would be a good example of a workload that doesn't translate well on a scale-out architecture.
I'm not saying that many mainframe applications couldn't be replaced by a cloud computing solution, but there are situations where latency and expensive orchestration are not acceptable.
lucm, indeed.
Thanks, that's an interesting comment. Especially with x86 servers getting fairly big these days (the 80-core, 4TB-ram monsters you mention), I can see that being plausible for some scenarios. Are all the services you ran previously each able to fit in a single x86 server now? If so that sounds like it'd greatly ease migration. One of the big pain-points of migration from mainframes to x86 clusters has traditionally been that it's hugely expensive to re-architect complex software so that it will run (and run reliably) on a distributed system, if it was originally written to run on a single system. But if the biggest single service you run is small enough to fit in your biggest x86 box, then you don't have to do the distributed-system rewrite.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Having to rewrite 4 decades worth of COBOL is also a prohibitive factor.
What did you pay to replace the software and test the new version?
Oh, and what is the name of the place? I don't want to have an account there for at least 10 years.
"Cock Up Your Beaver" does not mean what you think. This sig is intended to clog filters and annoy do-gooders
This is starting to sound more and more like bullshit. Were all of your cobol programs completely self-contained (highly unlikely). You didn't use any CICS, IMS, database, or any other middleware? You didn't use any VSAM datasets or any record IO? You didn't have any dependancies on JCL associating 'files' to 'datasets' and specifying how files should be opened and what should happen when they are closed?