Feeds:
Posts
Comments

Archive for the ‘Job Scheduling 2.0’ Category

It’s not even 3 weeks ago that the SOA thought-leader community announced The SOA Manifesto as part of the closing keynote of the 2nd international SOA Symposium in Rotterdam.

It seems a good time for Manifestos, that’s taken for granted, but do we need to wrap the SOA approach into a Manifesto too? Now that the technological benefits of SOA applications over existing legacy applications have been well documented? Now that it is well known that SOA offers business benefits across applications and platforms, including location independent services, a loosely coupled approach providing agility, the dynamic search for other services, and that services are reusable?

According to Joe McKendrick, “SOA is not a thing that you do, and it definitely isn’t a thing that you buy … but something tangible, a style if you will, just as Roman, Greek or Modern are styles of architecture.” This perfectly paraphrases the very first statement of the SOA Manifesto, which philosophically states that service orientation is more an attitude, “a paradigm that frames what you do.”

The good thing about the Manifesto is that it reminds us of the cultural implications of a SOA approach. And that it plays nicely with the principle that states that “SOA is not made out of products”. But on the other hand we should not neglect the technical challenges, of course. Because the situation we are focussing on is the fact that over 50% of the processing in the current application landscape is background processing.

That is why we need interfaces that guarantee the inclusion of background processes into the SOA business process. That is why we need intelligent Workload Automation tools based on Web Services to initiate, monitor, and manage background processes. And that is why UC4 customers prefer an Automation Broker (here is the Whitepaper for Download) to bridge between SOA vision and batch reality than being part of a platonic style symposium.

I am looking forward to your comments!

Here is the video of the announcement:

You can find a complete list of the authors at the bottom of the manifesto, or view a picture of everybody on stage at the signing of the SOA Manifesto.

Read Full Post »

Are we talking about real things when we talk about “just in time processing”? I mean, things that we are confronted with in daily business. Or is everything just about wishes and visions? And how things might be better?

Michael L. Moser gives kind of a disillusioning answer in his blog claiming that “the typical IT organization has a hodgepodge of disparate job and workload solutions that only rarely talk to each other and often aren’t integrated. Each solution requires its own specialized staff to manage and maintain it. Worst of all, this outdated job scheduling approach is costing companies a lot of money — something they can ill afford in light of the current economic crisis.”

No question, in order to reduce costs, mitigate risks, and increase business agility, enterprises of all sizes have to get rid of these automation islands.

The recent economic gloom has created a troubling paradox in IT management across a broad swath of industries. An EMA research report found that 74% of CIOs must maintain or reduce IT budgets over the next two years – emphasizing both the importance AND the danger of always having a red pencil within reach.

Naturally, businesses do not relax service expectations despite shrinking budgets. As the recession continues, the options dwindle for satisfying these apparently incompatible objectives. With each “red pencil” passing through the budget spreadsheets yielding fewer cost reduction candidates, layoffs loom as the final alternative. With each layoff, the organization pushes increasingly valuable IT expertise out the door, reducing the ability to compete.

Don’t throw your visions to catch reality! Better discover how real-world customers are succeeding: 6 Ways to Cut Process Costs.

Read Full Post »

When people are talking about process automation it often happens that they disregard the power of change. They talk about dynamic markets and simultaneously think about processes as engraved in stone. But this would imply that they are perfect and timeless.

In fact, any automation effort must be accompanied by an ongoing optimization effort to ensure that the processes are up-to-date and the engine is not automating errors and detours. For this you will need some tools between the automation engine and the user, helping you to see what’s going on inside the process, to recognize and visualize patterns, to correlate events, to simulate workflows, to follow critical paths, and to report key performance indicators (KPIs) to the management.

In dynamic business environments monitoring is just half the battle. That’s why Complex Event Processing (CEP) (>see Wikipedia) is the technology of choice in the highly competitive automation market – “extracting business value from event data”, as Gartner Analyst Roy Schulte emphasizes.

A Bloor Research Whitepaper puts it in a nutshell: “Reacting to events to minimize risk´, predicting events to capitalize on opportunities and analyzing events to improve future performance; and doing all of this now rather than next week when it may be too late, is what event processing is all about. It is increasingly being recognized as a crucial capability for optimizing the business.”

Read Full Post »

There was a time when everybody talked about ‘real-time’. You couldn’t speak highly of something without stressing that it was almost or close or already ‘real-time’. Do you remember the 70s in the car industry? I do. And I remember the similarities as well. It was a time when no kid could help looking at the speedometer when passing a car. It wasn’t just my childhood, but also the childhood of car industry. A time of immaturity – when technology was praising itself. Going back to my starting point: Looking at the speedometer is a bit like talking about ‘real-time’. What comes off badly, is the user (the driver).

Because the mature driver doesn’t care about 280 miles per hour on the speedometer as long as he is travelling on streets with a speed limit of 80 or 100 miles. But he does care about security, travelling comfort, fuel consumption etc. He is like the business user, who doesn’t care about real-time. What he wants is ‘just-in-time’: that he gets the information he needs at the time he needs it. No more, no less.

That’s also your company’s challenge, when it comes to automating across different people, platforms, systems and applications. You don’t want to over or under provision your IT infrastructure as there is a related cost to doing this. What you want is to effectively deliver the right business information at the right time at the lowest possible cost. Why deliver at real-time, when there is no need and nobody cares?

Take Amazon or other online selling companies: A consumer ordering a product finds it more than acceptable to receive an email order confirmation a few minutes after having placed their order. The consumer will get no added value if the email arrives within 1 millisecond, but will certainly be disappointed if the email arrives 24 hours later. Hence online buying and selling is a classic example of using ‘Just-In-Time’ processing to satisfy consumer needs and keep automation costs to a minimum.

In fact, there is a parallel to the manufacturing industry with the familiar technique of ‘Just-in-time’ manufacturing that revolutionized this industry. Why shouldn’t the same approach be able to revolutionize a business’ automation strategy – receiving maximum value from automating its processes at the most effective cost point?

Read Full Post »

Technically speaking an event is a change in status. That’s not really inspiring, I know. But on the other hand, hitherto you got only half the truth. Because status changes are innumerable, and therefore you call an event only something which is changing a status AND (simultaneously) causing an action. That’s more tricky, I know, even in daily life. Because our digesting ability is enormous and sitting in front of the tv-set the procedure of absorbing huge piles of information without reaction becomes an important cultural skill. The weird thing is that in the media business, almost everything is an event and nearly nothing produces a status change within the audience.

No wonder, that in real life I sometimes feel like one of these old-fashioned schedulers, who can not respond dynamically to changing external data coming from business. So I remain sitting and eating junk, ignoring the data coming from the application (which is supposed to be the tv-program in my case).

The question behind is: when do I start thinking of a process not just as stupid external routine I cannot escape, but as something dynamic making me part of it? You wanna know the source of these thoughts? It was a white paper by UC4, which made me look Beyond Job Scheduling and the edge of my plate. Feel free to decide whether this event was user or system driven.

Read Full Post »

In today’s IT environment, it is still common for enterprises to use two or more job schedulers – one for the mainframe and another for the remainder of the enterprise. This practice stems from the belief that mainframe-specific schedulers are superior to those that offer centralized control and automation for both the mainframe and the entire enterprise.

The following are some key misconceptions regarding legacy mainframe schedulers that often lead IT managers to believe they need to maintain two schedulers at once.

  • Myth 1: A mainframe-specific scheduler is the most reliable solution for managing mainframe jobs. But actually a mainframe scheduler is limited, whereas an enterprise scheduler can be relied on to process any job throughout the entire enterprise. And: It can be implemented for non-stop operations while improving scalability across any number of CPUs and physical servers.
  • Myth 2: A mainframe-specific scheduler is the most stable solution for managing mainframe jobs. In fact, an enterprise job scheduler allows the most stable scheduling by providing visibility and control – for the entire enterprise – in one location. Processes can constantly be monitored and the manager can be assured that everything is running properly through a centralized GUI.
  • Myth 3: Using a mainframe-specific scheduler keeps business-critical jobs under control. But what about the business-critical jobs which are not kept on the mainframe? There is no assurance that the mainframe scheduler can see the job through to its end if the job is integrated across the entire enterprise. An enterprise-wide scheduler can monitor all business-critical jobs across an enterprise, ensuring completion.

What about you? Wanna add some more?

Read Full Post »

Can you imagine people talking about “aligning” accounting and finance to the business? Or “aligning” the CEO to the business? That would be pretty pathetic, wouldn’t it? So why do we constantly fret about “business-IT alignment”? I like these provoking questions Joe McKendrick raises in his ZDnet post.

Because the more we think about aligning IT and business, the more we feel the chasm between the two, or the wall. Joe McKendrick would say that there is none. And – culture wise – he is right, of course. And it’s also true what Andy Kyte, vice president and Gartner fellow, claimed at a Gartner IT conference: “None of you are in IT; all of you are in business.”

But on the other hand it still makes sense to talk about system logic and business logic as two different things. Otherwise it would be difficult to tell the difference between job scheduling and workload automation, as Mark Carkeek, UC4 Vice President of Product Development, points out: “Job scheduling was always just an IT thing (to run X-process at Y-time), whereas workload automation actually helps running the business. Because it hooks into many different business products, like files, reports, databases, and applications, UC4 is able to see into these disparate IT resources, and can act on them. We have become workload automation because our software is not just for IT, and it’s not just for business. Because systems cross business and IT boundaries and need to interact reliably for the company’s good.”

Business/IT Fusion Even if Peter Hinssen (Business/IT Fusion: How to move beyond Alignment …) believes that “alignment is simply a dead-end street” … and that it’s time “for a complete overhaul of IT, and a complete rethinking of the relationship between business and IT” – the discussion – whether ALIGNMENT or FUSION is the right term to describe the challenge of this journey – remains a bit academic. But we agree with him about what really counts: “The elimination of the wall separating IT and business around/ through/over/under which alignment is supposed to happen.”

Read Full Post »

Older Posts »