“Cloud computing is not just one more way to deploy information systems. It represents a total shift in how IT resources are delivered and ultimately will replace most if not all internally-maintained IT infrastructure.” This is how Franc Scavo starts his latest blogpost on “the inexorable dominance of cloud computing” and the raise of utility computing – inspired by a speech by Nicholas Carr at a Cloud Computing conference organized last week in London by Google.
This is, by the way, the same Nicholas Carr who shook the world of many IT managers and CIOs when he published his article “IT doesn’t matter” in the Harvard Business Review in May 2003 – trying to make the case that, from a strategic standpoint, infrastructural technologies will commoditize and become more and more invisible.
Nowadays, and some years later, we´re over this initially offending perspective. We are able to see that the value propositions of SaaS and Cloud computing strategies are significantly better than those of on-premise software. And that both SaaS and the Cloud rely on integrated technologies from an automated backbone. The more invisible, the more mature. The more mature, the better.
It was Nicholas Carr who coined the matching Cloud koan in his blog: “Not everything will move into the cloud, but the cloud will move into everything.”
The Gartner Symposium/ITxpo 2009 we have been attending in Orlando not only endorsed the big hypes aroung virtualization and cloud computing, but also our ongoing investments in service-aware process automation – offering real-time intelligence for just-in-time execution. It matched perfectly that Gartner analyst Roy Schulte and K. Mani Chandy, Professor at the California Institute of Technology in Pasadena, used this event to introduce their brand new book called “Event Processing: Designing IT Systems for Agile Companies” about the business drivers, costs and benefits of event-processing applications.
According to Mr. Schulte and Mr. Chandy the new aspirations in situation awareness and reaction accuracy can`t be achieved by simply speeding up traditional business processes or exhorting people to work harder and smarter with conventional applications. Instead they urge companies to make fundamental changes in the architecture of business processes and the application systems that support them by using more of the event-processing discipline. “While a typical business process has time-driven, request-driven and event-driven aspects, event-driven architecture (EDA) is underutilized in system design resulting in slow and inflexible systems,” said Mr. Chandy. “Event-driven systems are intrinsically smart because they are context-aware and run when they detect changes in the business world rather than occurring on a simple schedule or requiring someone to tell them when to run.”
“Event-driven CEP is a kind of near real-time business intelligence (BI), a way of `connecting the dots` to detect threats and opportunities,” explained Mr. Schulte. “By contrast, conventional BI is time-driven or request-driven. Complex events may be reactive, summarizing past events or predictive, identifying things that are likely to happen based on what has happened recently compared with historical patterns.”
Sometimes industry hypes can reciprocally enforce each other and sometimes they even coexist so closely that the question: “Who was first” is verging on the “chicken or the egg causality dilemma”. With “Virtualization” and “Cloud Computing” it’s different. Of course, they do “hype” each other but the concept of cloud computing is not even thinkable without having virtualization technologies in mind. A concept which is defined by the U.S. National Institute of Standards and Technology (NIST) as “a model for enabling convenient, in-demand network access to a shared pool of configurable computing resources … that can be rapidly provisioned and released with minimal management effort or service provider interaction.”
I found this definition whilst going through a brand new EMA White Paper titled “Achieving Virtualization Control with Intelligent Service Automation”. In this study EMA researcher Andi Mann develops the supporting argument that an efficient use and dynamic provisioning of resources depends on service-aware automation technologies.
Workloads don’t care whether they are happening on physical or virtual systems. That’s why – according to Andi Mann – an automated virtual service management is the basis of any serious cloud approach. “It sets the stage for well-managed cloud computing services, with two essential components. Firstly, the virtual infrastructure provides a turnkey approach to flexibility, agility, and scalability, and the essential convenient, on-demand configuration of shared computing resources. This is difficult, and perhaps impossible, with a tightly-coupled physical environment. Secondly, intelligent and sophisticated automation is essential to ensuring minimal management effort or service provider interaction, which would be impossible with a manual management approach.”
And that’s why the UC4 Intelligent Service Automation platform is an important building block for cloud computing: 1) It initiates the dynamic distribution of workloads out of the process. 2) It immediately integrates newly-provisioned systems into your daily recurring housekeeping routines for backup and maintenance. 3) It introduces real-time intelligence to Cloud-Computing by acting on not reacting to events inside the applications. And finally 4) it provides an end-to-end view for predictive process management, integrating real and virtual worlds.
“We recently had to replace the server in my office. It was seven years old and one of the hard drives failed. It was not an expenditure I expected to have this year. My IT guy said that my desktop is seven years old. He also informed me that half the machines in the office are between five and eight years old and that I should budget to replace all of them next year. If we did not have a weak economy, I would normally have replaced these machines after five years of service. I think my business is typical of many businesses around the world“. This is how Ronald Roge, chairman of R. W. Roge & Company, one of a highly regarded wealth management firm describes the situation.
The related article from forbes.com is about the IT market and the pent-up demand in many firms due to the crisis in the last year or two. The good news: Concerning 2010 forecasts, the industry’s experts expect the cork to pop this year.
What strikes me is that in the article they talk mainly about IT lifecycles and replacement procedures and not about how technology itself has changed in the last two years. Take virtualization technologies and the role they can play – even in a tense economic situation – as the key to doing more with less; to reducing hardware, space and energy demand and to making your business more available, more agile and more productive.
Of course, not before you have done your management lessons on integrating real and virtual environments in a consistent way. Using VMware it becomes clear that the dynamic provisioning of new systems is not enough. Unless they become part of your automation strategy they remain outside your business processes waiting for costly manual integration.
Talking about the cork which is supposed to pop next year we should also talk about the deadlocks threatening the virtualization issue. We should talk about virtual machine sprawl and costly process interruption at the intersections between virtual instances and physically deployed systems.
And we should underline that it is not enough to find out the status of your server hardware the moment you want to dynamically provision workloads; you have to go deeper – into the application layer that correlates and acts on events to bring real-time intelligence and real-time dynamic to virtual and cloud computing environments. It’s obviously much more than fulfilling IT lifecycles. It’s about considering IT as a strategic asset and not as a cost center.
Have you ever tried rafting? And did you ever feel the adrenaline rush of a wild ride through white water? Processes are like rivers. Floating. That’s common, of course. But let’s dig a bit deeper. And see how and why white water happens.
For a loose definition I would note that white water is not just a matter of speed but the result of different floating velocities. When passing a chute, water speeds up but just for a short time, because it cannot escape, is always part of the river and surrounded by water at a lower speed. This is the reason why maelstroms suddenly occur, marking the interfacing line between two velocities. These maelstroms are not only dangerous, but malicious – forcing you to seriously “read” the water.
Against this background and thinking about virtualization issues, the analogy that processes are like rivers is not that commonplace anymore. This explains perfectly why the creation of virtualization islands – in our new terms: the creation of “virtualization chutes” – doesn’t lead us anywhere from a process point of view and just produces maelstroms which are difficult to handle and threaten operational efficiency.
They work as short chutes but don’t speed up the river as a whole. And so they don’t speed up the end-to-end process from back-office to the customer’s interface. But having customers in mind and SLAs can you seriously talk about processes without thinking end-to-end? Certainly not. And that’s why we need virtualization automation more than ever. To assure that your automation strategy covers all applications and processes in real AND virtual world – and to increase productivity alongside the river.
Process or Business Rule? This is the question Gartner Analyst Jim Sinur raises in one of his latest blogposts. Although this approach appealed to me from the beginning, I could not get rid of the feeling that this kind of comparison was slightly academic – playing with narrow concepts. Because a process is no more a stupid external routine than a business rule can be extracted out of books. And neither is a “process a bunch of rules strung together in a static or dynamic sequence”, nor “act a business rule as simple constraint/boundary to keep processes on track.”
Following the argumentation for both sides, it seems that the answer will not be found by a diplomatic balancing of these points, as suggested. Because the right answer must reflect the dynamic nature of business operations and process composition, I would rather subscribe to James Taylor’s comment: the crown belongs to “neither rules nor process but decisions … You MUST have the process to fulfill the decision – to take the action you decided on – but it is the decision that rules.”
Mark Norton’s comment is along the same lines: “Decisions are the policy-driven specification of how value is created and recognized by each business. For value to be created, a change of state of “something” must be explicitly acknowledged by the organization, and this acknowledgement is a bespoke, decision-centric activity.”
A change of state is what we call an event. And events drive both businesses and operations. Maybe it is events that fit the crown best. The king is dead. Long live the king.
It wasn’t really a surprise that one of the big topics at the VMworld virtualization fest in San Francisco early this month was the virtual machine sprawl and the long desiderated management answers by the IT departments.
This is not just the case in VMware environments as Jamie Erbes, Chief Technology Officer for HP, points out in a ZDNet interview: “On the need to manage virtualization efforts, the problem of “virtualization sprawl” is becoming real. Developers have adopted virtualization and IT managers don’t use the rigor needed for its use. For instance, if some department wants a new server a virtual machine can be created in minutes. Multiply those departments in a multinational corporation and you have a lot of virtual machines” – and a big threat to end-to-end management.
In daily IT business the problem is not the quantity of the virtual machines, as the title borrowed from Timothy Pricket Morgan’s blog suggests. The problem is that these VMs are just automated and orchestrated inside the VMware virtual environment and isolated in relation to the IT processes which are still running separately in real and virtual environments. From this process viewpoint, managing the virtual environment is a far cry from managing the entire IT environment – as long as the dependencies between the virtual and the real world are unsolved. That’s the catch. With this in mind, serious virtualization efforts just make sense together with a serious automation strategy. And a serious automation strategy is cross-platform or not – as I stated in this blog months ago.
This is also in tune with Jamie Erbe’s point of view. She emphasizes in the interview mentioned above that “companies need to focus on their policies and how rules for things like storage, applications, cloud computing and storage all work together. If these policies aren’t well thought out the effort to automate infrastructure can be wasted and wind up costing you more.”
Don’t forget. You can do great things with automation. But first you have to make sure that the automation train goes in the right direction. This train is fueled by “integration”. Integration without restraints.
To begin with, I really love to read feature stories. The more meandering they are the better. But the closer I am to making a decision, the more I search for facts and figures. And reading them in a spreadsheet manner. No wonder that during these periods checklists like the one I just hit upon have a special appeal with me.
It’s a checklist for UC4 Insight which is mainly built out of questions. I picked out my favorites and, of course, would love to know about yours.
□ Do you like to have a tool to detect similarities and patterns that resulted in errors?
□ Do you sometimes have incidents where you cannot say why they occur?
□ Do you sometimes have problems to proof that the origin of an incident is not in your department?
□ Would you like to have a tool that gives a general overview of at which times and days delays in processing occur (jobs take longer to finish than normally)?
□ Would you like to have a tool that shows on which servers and operating systems most of the processing-incidents and/or processing-delays occur?
□ Would you like to have a tool that gives you a quick overview of user behavior?(access violations, cancellation behavior, etc.)?
You think these questions are suggestive and don’t allow a NO? You are right. They are! But did you also ask yourself if this may be something to do with the way businesses are run today? Maybe you ask yourself whether you can be successful in a dynamic market without these valuable insights into your production environment?
It was in April 2009 when Venture Research CEO Mark Smith proclaimed in a blogpost that Business Activity Monitoring (BAM) had finally reached its dead end. This was mostly due to the fact, as he argues, that BAM “was not designed and focused to handle the dynamic need of business” and kind of powerless toward “the need to bring the analytic and smart components of business intelligence (BI) and design it to operate on an event-based architecture where the modeling can operate at a very fast instantaneous basis or operate within the relevance of an event to the needs of the business or policies.”
When we talk about events we have to consider that they can happen at very different levels in the enterprise with the ability to shape and change processes at any time. No wonder that Operational Intelligence which establishes a relation between business data and IT data has become a key business priority for organizations.
In reality, Complex Event Processing (CEP) was the gravedigger for BAM going one decisive step further than watching or monitoring what is happening along the business process. CEP is key to automatically responding to events coming from different sources and different levels. And it’s considered complex not just because of the number of event sources or levels. It is complex because it sets up a relation between the event and his impact to the process.
That is truly a revolution, as Mark Smith points out in his newest blogpost: “Unfortunately many in the industry have confused the discussion where they believe CEP is the technology market for a new generation of capabilities which actually already occurred decades ago. The new advancement is actually in the tools that can be used or applications that are assembled to help business or consumers understand the relevant importance of specific events and only provide the information when it is needed.
This is the essential divide between monitoring and intelligence where monitoring is based on thresholds and does not provide the context and relevance of the actually impact of events occurrence to something that matters to an individual, business activity or process.”
We, at UC4, are going even one step further, because we are able to not just analyze current data but also to integrate historical data. And, of course, we do much more with these data than notifying and presenting them in a dashboard. For CEP is no less than intelligence AND action.
Automatically generated messages can help us to survive in information overloaded environments. Think about the Out-of-Office-Notifications and how they help us to curb the amount of stuff we have to get through when returning from vacations.
There are probably several prosaic or poetic ways to circumscribe the only relevant thing you want to communicate in these Out-of-Office-Replies, namely the date when you are In-the-Office again. Considering this, why don’t we take the chance to have a laugh with it, as Zack Whittakersuggests in his ZDNet blog?
My favourite one is the following:
“You are receiving this automatic notification because I am out of the office. If I was in, you wouldn’t have received anything at all.”
Wolfgang Tonninger
Ralf Paschen
Christian Plaichner
Paul Rosien
