What’s a mainframe, Daddy?

After years of sliding my security card in the lock and entering the machine room/data centre and seeing the mainframes in there change from Sci-Fi-style boxes with flashing lights to more mundane-looking boxes. From seeing simple DASD with less capacity than the memory stick in this laptop be replaced with cache controllers and more sophisticated data storage devices. It always seemed that there were plenty of mainframes around and any normal person (me) was constantly being offered tours round installations. So it comes as a bit of a shock when a youngster clearly has no idea what a mainframe looks like or what it does!

OK, no-one may have actually said those words as such, but that was the message. Plus, I was with some friends on Saturday when the conversation turned to discussing what use a mainframe was in this day and age! As Arcati Director, Mark Lillycrop, so eloquently put it recently, mainframes are thought of as ‘your dad’s technology’. Most of the people I was chatting to felt that mainframes were relics of the past and anything they can do, a few servers could do just as well!

So for many of us mainframe verterans, our job is to get out there and spread the word. We need to tell people exactly what a mainframe is, what it can do, and how people are interacting with them all the time, but don’t realise it. That way, the new generation of youngsters that are beginning to get access to mainframe technology at universities and elsewhere will arrive with a knowledge of what mainframes can do, and why working with them can be so enjoyable.

So let’s just start with the absolute beginner’s guide to mainframes. They are computers – just like your laptop – except that over the years they faced and solved all the problems about back-ups and restores, security, and high-speed data access. They have been around for a long time – which is a good thing because lots of people have moved the technology forward. They allow millions of users controlled access to information – allowing them to create, modify, and save data from almost any data entry device you can think of, including browsers.

Mainframes have been virtualized since the 1980s, and some of the software first saw the light of day in the 1960s. Most Windows data centres have only been virtualizing for the past five years! It’s true that laptops etc are everywhere – in your home, at work, etc – but mainframes are working away in the background. Everytime you take money from an ATM (cash machine) your bank is running a transaction on a mainframe. And it is banks and large financial institutions that use mainframes. And they do it because of the reliability. They do it because, should there be an outage, they can recover back to almost the last second before they went down. Almost no transactions are lost. And as a bank customer, I like that. Lots of non-mainframe-using sites think they are doing quite well if they can recover data back to last night! You see the difference in scale here.

Mainframes run an operating system (z/OS, but could be z/VM or z/VSE) and on top of that are a number of subsystems – you might think of them as apps (but big ones!). These subsystems include CICS and IMS. Now, both of these have been being developed since the 1960s and provide ways of accessing data very quickly and securely. They allow users to fill in virtual forms. And they store data in a way that means it can be accessed very quickly.

Another ‘app’ you may have heard of is DB2. DB2 is a comparative youngster, having arrived in the 1980s. It stores data in a ‘relational’ way rather than the more traditional ‘hierarchical’ way. DB2 is a database that can exist on Windows machines as well as mainframes (and many devices in between).

Mainframes can also run Linux (z/Linux) and all the Linux applications. That makes them very cost-effective replacements for sites with numerous ageing Linux servers. Linux has been available on mainframes for just over 10 years.

And there’s plenty of software available to control all aspects of this mainframe behemoth. And you can link them together at different sites in different countries round the globe.

So if anyone asks you what’s a mainframe, you can tell them that it’s the most successful server architecture ever devised and it’s all around them doing important work.

IBM – C

IBM celebrating its 100th birthday makes you think you should write its age in Roman numerals – which is what I did in the title. How does a venerable old organization avoid being put out to grass and stay ahead of the business game? How does it become synonymous with cloud computing, smarter planet, and data analytics? I guess the answer is by completely re-inventing itself.

The truth is that IBM is not the controlling influence it once was, but that’s a good thing. No longer does it decide what customers want and try to sell it – it now listens to its customers. Another good thing.

In the 1970s, IBM was the king of computing with the combined revenues of the BUNCH significantly less its own. The BUNCH were Burroughs (still around as UNISYS), UNIVAC (still around as UNISYS), NCR (which was acquired by AT&T for a time and eventually sold its computer manfacturing business to Solectron), Control Data Corporation (now called Syntegra), and Honeywell (whose computer division became part of Groupe Bull).

During its lifetime, we’ve seen the birth of Hewlett-Packard (72-years-old), Intel (43-years-old), Apple (35-years-old), Google (12-years-old), Facebook (7-years-old), and other companies that have become kings of their not-so-small niches.

IBM itself was (obviously as we’re celebrating 100 years) founded in 1911, but was then called the  Computing Tabulating Recording Corporation. It originated with the merger of four companies: the Tabulating Machine Company, the International Time Recording Company, the Computing Scale Corporation, and the Bundy Manufacturing Company. The name International Business Machines didn’t arrive until 1924 (so more partying in 13 years time!). Bundy Manufacturing first appeared in 1889. The Tabulating Machine Company arrived in 1896.

The famous Thomas J Watson Sr joined the company in 1914 and soon became president. In 1956: Tom Watson Jr took over as CEO. Through the 1960 we get the System/360, VM, IMS, CICS, and many other things that we’re still familiar with today. In 1981 IBM invented the Personal Computer.

The early 1990s were a bad time as IBM announced its first loss. CEO John Akers talked about breaking up IBM into smaller companies that could compete better. In 1993, Lou Gerstner took over the reigns and things started to improve. In 1995 IBM acquired Lotus and its Notes software. 1995 also saw the launch of the much-admired ThinkPad laptop computer. And 1996 gave us DB2.

But over time, IBM has moved away from opportunities. It sold off its Lexmark printing business in 1991. It sold its networking business to Cisco in 1999. And it sold its PC business to Chinese-based Lenovo in 2005. But it’s still a big company, In 2011, Fortune ranked it as the 18th largest in the USA, and the 7th most profitable. Forbes ranked it as the 31st largest in the world. And it has thousands of patents under its belt.

The zEnterprise 196 hardware shows that IBM is planning for a long future. It allows all the good things about mainframes to be spread across other platforms. Plus, people are slowly grasping the amazing things they can achieve with zLinux on mainframes. Big Iron isn’t rusting away!

And IBM always had a fun side (honest!). In 1997 the Deep Blue computer defeated chess grandmaster Garry Kasparov at chess. And in 2011 a computer called Watson won the Jeopardy game show.

Happy birthday Big Blue.

Mainframes, cloud, and trans-derivational searches

Communication is a bit of a hit-and-miss business at the best of times. Suppose I was sitting in a large comfortable leather armchair and I wrote the word ‘chair’. I would be thinking about my armchair. You might be sitting at your computer on an office chair with wheels and no arms. When you read the word ‘chair’, you might think that your office chair was what I had in mind. Perhaps a trivial example, but you get the idea.

What happens if I write a complicated sentence and it’s not immediately clear what I’m saying. You would read it and perform a trans-derivational search in your mind to try to find meaning. What is a trans-derivational search? Well, it’s a bit like going to Amazon and typing the word ‘cloud’ into the search box. Amazon will search in ‘Books’, ‘Film, Music & Games’, ‘Kindle’, ‘Electronics’, etc and display all the answers. You can then make a choice. When it happens in your mind, you do a similar search and select the search result that seems to be the best fit for the unclear sentence.

The reason I mention this is because when one person says the words ‘cloud’ and ‘mainframe’, the image in their mind may be quite different to the image in anyone else’s mind. I like to think of mainframes as interesting and powerful bits of kit that keep the business world turning. Someone else may picture rusting hardware that should have disappeared in the 1990s! It’s much the same with the word ‘cloud’. One person may envisage an outsourcing opportunity, while another may view cloud as a security nightmare – a disaster waiting to happen. You can see now just how tricky this communication stuff really is!

There are people who argue that the mainframe is a cloud already. Suggesting that both are resource that can be dynamically allocated and de-allocated on demand, and can be made available within a company with the necessary security and management controls.

Typically the users of cloud computing have x86 platforms and need to be able to get what they want when they want it – ie on demand. They use primarily Amazon, Google, or Microsoft services. And, of course, everyone is now talking about cloud with announcement of Apple’s iCloud.

Going back to the mainframe, recent surveys have shown mixed interest in cloud computing from mainframers. It seems that the closer to the coal face (or whatever the mainframe metaphor for systems people might be) people are, the less interested they are in cloud. Whereas more high-level thinkers are interested to see what advantages cloud computing offers their organization.

Same words, different mental picture. It’s that communication conundrum again.

If you do have any experience of using cloud and mainframes, I’d be interested to hear about your experience over on IT Toolbox (http://datacenter.ittoolbox.com/groups/strategy-planning/data-center-infrastructure-sp/featured-cloud-computing-in-a-mainframe-environment-4234328).

And for those of you who are IMS professionals, don’t forget the Virtual IMS user group meeting on Tuesday. Gary Weinhold and Verna Bartlett from Data Kinetics will be talking about MSU reduction due to in-memory table management with (any) IMS applications. Full details including how to register are on the Web site at www.fundi.com/meetings.htm.

Cloudy but good

All this talk about 2011 being the year of the cloud rings true. I say that because marketing hype is usually a bit ahead of the curve. But I realize that I’m using cloud computing a lot these days – and I don’t mean (necessarily) for work.

Firstly, I’m using Google Cloud Connect for Microsoft Office. This is a small piece of code that’s free to download from http://tools.google.com/dlpage/cloudconnect. Once installed, you can save all your Word etc files on your laptop AND there’s a copy in Google Docs. Now, the reason I find that so useful is not simply because I know my documents are backed up (and to be honest that’s a really important reason and the one that I’d use as a selling point to most people), but it also means that I can access the documents from other computers. I have two laptops that I use all the time, and a third one, that’s usually away at users’ sites. I can now access my documents wherever I am and not have to say that I’ll send on the information after the meeting (or whatever).

Google also say on their site that this system allows users to simultaneously edit their files – which again is a great idea, although it’s not something I’ve tried.

Previously to using Google Cloud Connect, I used my Pogoplug for the same reason. I’ve mentioned Pogoplugs before in these blogs because they are so useful. It provides my own personal cloud. A Pogoplug is a little piece of kit that connects to your router. I then have four or five memory sticks plugged into it, but I could have a 2TB external hard drive. I can upload files to the Pogoplug from anywhere using a simple browser interface (https://my.pogoplug.com/) and I can share files with people, so they can see photos I’ve taken or short videos, or even Office documents. And, again, it provides me with a back-up copy of my important files.

The third piece of cloud computing I use is Wyse PocketCloud. I’ve downloaded the free app for my Android smartphone and I’ve downloaded the software for my laptop. Once they’re password synchronized, I can control my laptop from my phone. Which means, if I haven’t got a document with me, I can use my phone to launch Word on my PC and read the document. Last week I sent a document from someone else’s computer to my e-mail address. I used the PocketCloud software to log-on to my e-mail, download the file and print it. So when I returned back to base, there was the document ready for me to take to the next meeting (a school governors meeting – they don’t have wifi, hence I couldn’t be paperless). I’ve sat in McDonalds and called up files! It is a really impressive piece of software. Now, with the other cloud facilities mentioned above, I don’t know how much I’ll need to use it. But it just seems so amazing to be able to do it!

What I’m suggesting is that cloud computing is getting in under the radar and people are beginning to use it already. So, when IT departments suggest rolling out Cloud services in whatever form that takes for them, they will be pushing at an open door. There won’t be usual user resistance because the users will be perfectly happy with the concept. And, of course, mainframers will be saying that they’ve been using these cloud concepts since the 1960s – or whenever they first started working with mainframes.

Which all goes to show that 2011 really is the year of the cloud.