To some, it might be surprising to learn where to look to catch a glimpse of what’s on the “horizon” for networking. According to Gartner VP/Distinguished Analyst Jay Pultz, over the next three years, “It will be ‘back to the future,’ as we can expect to see a resurgence of policy-based ideas in networking management.”

Speaking at today’s Symposium session, “Communications and Computing: Everything Converges,” Pultz said that some folks in networking might recall that Gartner published research about policy-based management when it emerged on the scene in 1999. Policy-based management is a systematic approach to address different (and, sometimes conflicting) business needs and priorities. It’s a framework to help networking organizations allocate resources to the infrastructure, and is key to improving agility, Pultz explained.

So what happened to policy-based management in networking? “It didn’t see light of day because bandwidth became cheap,” said Pultz.

What's different about policy-based networking this time? According to Pultz, first, the focus is on newer application models and delivery mechanisms. And second, there is less of a focus on bandwidth management. “Policy-driven infrastructure is fundamental to the real-time enterprise, so its time is now,” said Pultz.

As service director of the Gartner for IT Leaders IT Infrastructure & Operations role, Pultz’s agenda for this presentation examined how networking will (and must) be more closely integrated with the business, applications and computing. Other ideas that he offered for network managers and I&O Leaders during today’s presentation included:

1. To address longer-term application demands, get involved early in application developments that involve networking. Some entry points: Communications-enabled business process technologies and, possibly, simulation tools.

2. To manage the network in a more agile fashion, automate network management processes now and integrate with IT operations.

3. To more tightly couple networking to business needs, translate technical performance metrics into meaningful business ones via business service management tools.

Submit a Comment | Permalink

Why do enterprises view real-time infrastructure (RTI) as an imperative? How can enterprises lay the groundwork for achieving RTI and what have visionary enterprises achieved? Which technology suppliers are critical to achieve RTI, and how are they positioned? Gartner analysts Donna Scott and Tom Bittman will address these issues Thursday at 1:45 p.m. in a session titled, “Real-Time Infrastructure: The Technology Vision of Infrastructure and Operations.”

In a survey from Gartner’s December 2005 data center conference, 73 percent of attendees said real-time infrastructure (RTI) was an imperative for their organization. “Yet, since 2004 and into 2005, many large vendors have de-emphasized their messaging on RTI – including the likes of Sun, HP and IBM, in favor of other messages focused on business innovation.”

“This is partially due to the fact that success with RTI architectures will have the effect of decreased investment in hardware and software infrastructure (due to shared pooling efficiencies), and no vendor wants to encourage less revenues.” said Scott. But despite this de-emphasis by vendors, she added, “The user momentum to build a strong, agile and efficient IT platform to meet ever-changing business needs will continue to drive RTI.”

Submit a Comment | Permalink

On Sunday and Monday, I had sessions on IT infrastructure consolidation and key issues that I&O leaders face. Completing my I&O trilogy on Thursday at 9:30 a.m., "Communications and Computing: Everything Converges!" is the culmination of research that has been two years in the making. But completing it in time for Symposium meant sequestering myself in a hotel room in Phoenix.

IMHO, this is way cool stuff - a different perspective on the whole critical issue of IT/business alignment. This issue will be dealt with from a networking perspective: How do we better align the network with business needs and new application delivery approaches like Web-based services and SOA -- plus help enable real-time infrastructure and for networking to be "real-time" itself?

Hint: The network isn't the problem, but how we manage it is.

Another hint: How technology/tools -- and your buddies in the rest of IT -- can really help here. A lot of what I have to say applies to all of I&O and to IT and the business, for that matter.

Please join me! Go beyond the hints!!

Submit a Comment | Permalink

Seats went like the proverbial hotcakes for Tuesday's IT Infrastructure & Operations Workshop – and for good reason, according to Jay Pultz, VP Distinguished Analyst. “One of the problems with infrastructure and operations is that it’s such a broad, expansive area,” explained Pultz. “Gartner has over 100 analysts covering this area, and we have published more than 2,200 pieces of research focused on I&O in the past year.”

To reflect the broad scope of I&O Leaders’ responsibilities and concerns, workshop attendees were organized into three roundtable-style forums, led by Pultz, as well as Gartner analysts John Enck and Ed Holub. During these interactive discussions, participants discussed three key issues:

1. The challenge of communicating the business value of IT I&O to the business, using the key metrics that business understands. “We identified availability, speed, capacity and cost,” explained Pultz. “You want to relate these metrics to the business’ objectives.”

2. Enterprises’ need to improve service delivery via a “Quality Service Desk.” “Enterprises can look at the service desk in a routine fashion,” said Holub. “But these are the people who primarily interface with your customers, so there’s an opportunity to get better information and develop better metrics.”

3. The struggle to control mobile devices (or “control devices that we can’t control,” as Enck stated). Summarizing some of the points brought to the table by workshop participants, Enck pointed out, “With many of these devices, there are associated hidden infrastructure costs – the tool ecosystem is immature, vendor support is poor at best, and there’s a high risk of information exposure. It’s cool if the CEO is running around with a BlackBerry, but not so cool if he or she doesn’t recognize the risk.”

The workshop was designed as an outgrowth of the Gartner for IT Leaders Infrastructure & Operations (I&O) role program, for which Pultz serves as director. “Since I&O Leaders are short on time, we’ve tried to slice and dice our research into nuggets that make it easier to quickly find what you need,” he said. “Getting I&O Leaders together in these roundtable situations is just another one of the new ways that we’re working to help them identify ways to address their daily and long-term challenges.”

2 Comments | Submit a Comment | Permalink

What single IT element will have the greatest effect on IT Infrastructure and Operations during the next few years? “Virtualization,” according to Gartner VP and Distinguished Analyst Jay Pultz.

During a Gartner Symposium session today titled “The IT Infrastructure and Operations Leader: Balancing Strategies and Operations,” Pultz explained that now that server virtualization is well on its way toward maturity, I&O Leaders can expect virtualization to extend to most I&O subsystems during the next seven years.

What all this means for I&O Leaders? “We’ll see this trend with PCs and networks first, with storage taking longer because of the complications of all the stacks involved,” said Pultz. “I&O folks are accustomed to working with things that are more tangible, that they can kick around. As virtualization becomes increasingly prevalent, the skill sets required of I&O Leaders will change.”

Symposium attendees who are interested in interacting with Gartner analysts and their peers to discuss these and other pressing I&O issues can sign up for tomorrow’s Infrastructure & Operations workshop, a Gartner for IT Leaders exclusive. The workshop will be held at 1:30 p.m. in the back of the Gartner for IT Leaders Demo Center, located in the Northern Hemisphere of the Dolphin.

Submit a Comment | Permalink

Enabling mobile access for those end users that need it (or just want it and have the clout to get it) is too strategic to be treated with tactical, point solutions. We will make the case that IT Infrastructure & Operations managers should stop treating their mobile architecture(s) as tactical plans and make them a core IT skill. There are some fundamental avenues to building a center of excellence for mobility.

We will provide Gartner’s six mobile architecture styles in a session on Monday that planners can use to assess every type of vendor offering: point solutions, suites, platforms, toolkits, and services. For example, if you need three or more highly-custom applications, use an architected approach using mobile application gateways. Ultimately, we think an enterprise-wide multi-channel portal strategy will be viable - but not yet. One of the key challenges is that mobility costs money and the money to fund it will likely be in your business units – not within IT. Having a clear story and the rationale to back it up will help pry those funds loose.

Submit a Comment | Permalink

Linux and Windows are the operating systems that are both significant in size and growing in stature in the industry. While they are both still trailing Unix and other environments in functionality and capability, users wrestle with which one to adopt and how rapidly to proceed with a switch over. For some workloads it is easy to choose either one, yet other workloads demand performance, availability or specific functions that are difficult to attain with Windows and Linux.

Many users wrestle with whether they should wait or switch now. Additionally, they generally have to pick one or the other with each choice bringing both advantages and disadvantages. We project that the gap separating these system types will narrow sufficiently in just a few years to permit the choice to be made without compromise. We will give attendees guidance in thinking about the choices based on where they are starting from—“greenfield” deployments, migrations and those that come with hereditary traits. Additionally, we will take into account the hot topic of virtualization and help clients build it into their decision frameworks. Our session Windows vs Linux vs Everything Else on Wednesday, October 11 at 3:15 p.m. will illustrate this shift, identify the timing of this trend, and yield solid advice on how to make choices in the interim.

Submit a Comment | Permalink

We use "Scenarios" to paint the big picture in a given area of users' responsibilities. This week we'll turn our attention to the networking side of Infrastructure & Operations and highlight a Gartner specialty – myth-busting.

The six dangerous myths of enterprise networking are:

• Networks Are Just Dumb Bandwidth


• My Vendor Is My Architecture


• Bandwidth Costs Are Going to Go UP


• Big is Good, So Biggest Must Be Best


• IT Must Own Everything


• Centralized IT Is Better

Our analysts will make the case that now is the time to reverse the trend of the last several years: Add intelligence to the network – not just bandwidth. Add just the services you need -don't build a monolithic, single vendor architecture. Accept consumerization - but control its impact within the enterprise network. Maintain multiple suppliers and keep contract terms SHORT. Large organizations should adopt a shared services model.

The challenge is that in some shops vendor allegiance is very strong. Microsoft, Cisco and IBM will be battling for control over the management layer – and you'll likely have factions in your own company. It will take a rational approach to architecting and procuring network components to counter human nature and the Six Myths.

Submit a Comment | Permalink

A sea change in processor design is leading to the use of more cores and threads in microprocessors as the primary vehicle for increasing performance in each generation. The Intel Core 2 Duo, IBM Power5 quad core, and Azul Systems Vega 24-core are examples of this sea change. These changes do more than simply deliver on the performance improvement expectations of “Moore’s Law”; they bring consequences – consequences that are poorly understood or very subtle today. Users of these new systems may experience some disruptions that will require deliberate action to adjust to these new realities.

Some of these consequences are already clear to the market, such as the impact on software licensing budgets of this explosion when vendors price their software licenses by the chip, socket, core, or thread. Others, among them the requirement to update or redesign software to turn latent capacity into real performance, and subtleties in how virtual machines will operate in a multi-threaded and multi-core world, are the hidden icebergs ahead. The sessions “How Will Multi-core and Multi-thread Affect You?” and “The Future of Servers and Storage” will help attendees to navigate around these obstacles.

Submit a Comment | Permalink