via Riverbed evolves its Steelhead product family to better address performance, consolidation needs of global organisations | ITWeb.

The other big takeaway from this report is that 3/4 of the budget will remain in the data center; all these organizations will need help consolidating their data centers and making them more efficient. This represents a tremendous opportunity to help agencies determine which workloads are appropriate candidates for cloud/SaaS, staying in-house, or a combination of the two (think IRS during tax season). Vendors that can provide expertise in helping to both optimize internal assets and facilitate the migration of others to cloud/SaaS environments should have an advantage.

Read more of Bob’s blog entries at Data Center Continuum.

If the government is inefficient, all citizens pay–both as customers, paying high prices for bad service, and as shareholders. The bad news? Governments are far from efficient.

Many government agencies have similar processes and could share systems, but even e-mail systems are not shared. Some agencies  run on Microsoft Exchange, some on Lotus Notes, and some even use internally developed systems. And all agencies  need to handle anti-virus, anti-spam, backups and disaster recovery…

It is expensive, risky and irrational.

Governments are the largest IT buyers globally and do not benefit from scale.

How to solve this problem?  The first and most important measure is to have someone responsible and accountable for all government IT systems–a government CIO. The US has one, but many countries, like Brazil, don’t.

The government CIO is responsible for designing shared systems and infrastructure for cost reduction.

However, he (or she) needs enough power to enforce usage of the shared systems. Otherwise, many agencies will stay with current systems or even implement costly alternatives–don’t forget about corruption.

Shared infrastructure, or the Government Cloud, would bring huge savings for everyone, while reducing risks and increasing security. It is easier to have NSA auditing security policies in the Government Cloud, than ensuring that all state agencies follow a set of  guidelines.

More important than sharing infrastructure is ensuring that agencies share systems and automate processes. Automated processes reduce human errors and corruption–there is no need to bribe a computer system.

Today, there is enough technology to enable every government request to be completed using the web.

By automating processes, government employees may spend more time thinking about how to better serve the population, listening to citizens about how to improve health care or delivering new learning methodologies to classes. It could become the largest, most powerful, and most interesting non-profit organization ever.

As Steve Todd pointed out, it should be easier to innovate inside a powerful and rich organization than in a cash strapped startup–it is all about processes and culture.

The US has a unique opportunity to build a new type of government–more efficient, less expensive, and with better services. The financial meltdown created a favorable political environment for change and it shouldn’t be missed.

For Brazil, I am not optimistic.  There is no pressure to change, and, as a result of what seems like eternal economy growth,  the government is only becoming larger and more corrupt.

You can read Ronaldo’s other blog entries at IT Matters.

Overview

QLogic recently announced availability of its third generation converged networking technology, aimed to further drive adoption of converged networking infrastructures. The announcement highlighted:

• The release of three new products based on QLogic’s 3G CNA technology:

o    8200 series 10GbE Converged Network Adaptors

o    3200 series 10GbE Intelligent Ethernet Adaptors

o    8200 series Converged LAN on Motherboard (cLOM )

• QLogic’s assertion of a number of industry firsts with the release of these products including:

o   Concurrently running Fibre Channel over Ethernet (FCoE), offloaded iSCSI and TCP/IP traffic

o   Performing switch-agnostic VM-to-VM communication within physical machines

o   Deploying quad 1GbE and dual 10GbE ports from a single chip using advanced FlexLOM technology

• The first implementation of QLogic’s Adaptive Convergence strategy. Building upon its Flex Ports technology, it includes:

o   VMflex, which provides advanced virtualization services like creating multiple partitions and flexible bandwidth provisioning via a cleverly embedded Layer 2 Ethernet switch in the adapters, VMflex enables direct VM-to-VM communication within a physical machine and supports SR-IOV, NPAR, NIV (VNtag), and VEPA/VEB.

o   ConvergeFlex, which delivers concurrent protocol processing services that allow an organization to run FCoE, iSCSI, and Ethernet at the same time and provision as needed without any disruption to the server or any of the services running on it.

o   FlexOffload, which enables full offload of FCoE, iSCSI, and TCP/IP, which QLogic claims results in an average CPU utilization of only 9%, in contrast to competitive offerings which can choke CPUs by running utilization up to 90% or greater, according to QLogic.

o   SecureFlex, which secures data in flight leveraging encryption services, but allows organizations to leverage their own encryption at rest strategies.

The entire portfolio is managed by the new QConvergeConsole, which provides multi-platform, single-pane-of-glass management of FCoE, iSCSI, and TCP/IP functions for ease-of-administration and converged network deployment.

These new releases are available to OEMs now and will be generally available in Q4 2010. The company expects ASPs on parity with the previous generation of its CNAs. Working products are currently in a number of customer environments and in OEM testing labs.

Analysis

QLogic is stepping up to the plate with a host of new products and capabilities. This new functionality should strengthen its position in the market and be attractive to users because:

• Consolidation and virtualization initiatives are still top of mind. ESG research shows that organizations’ top IT initiatives for 2010 include increasing the use of server virtualization and consolidating data centers (see Figure 1).[1] As organizations look to consolidate and drive greater efficiency in the data center, leveraging a converged network makes sense. The ability to reduce the number of cards, cables, and switching infrastructure not only lowers costs, but also increases service levels. Solutions that leverage offload capabilities in the card to maximize CPU utilization for VMs should be preferred. Security is also top of mind—organizations should welcome the added protection of encryption in flight to minimize risk.

• Users haven’t decided which protocol they will use and don’t want to be locked in. Taking a similar approach as those storage vendors deploying unified storage platforms, QLogic now offers a unified converged network adaptor that provides the flexibility to dynamically allocate and concurrently run multiple different transport protocols. ESG research indicates that users currently employ a wide range of protocols to connect servers to storage and environments will probably remain mixed depending on the workloads that are running.[2] Because VMs are highly dynamic and can easily move from one physical machine to another, dynamically reallocating data center connectivity will be very important. The ability to manage disparate data center connectivity protocols via a single interface will ease operational concerns.

• The timing is right. This has been the year to test and validate converged networking. Double digit quarterly growth indicates organizations are testing the converged networking waters and buying products to validate in their environments. Virtually every major storage and server vendor has tested, qualified, and now offers second generation converged technology in one form or another; the introduction of the third generation technology should be fast-tracked through the same qualification process. As enterprises and service providers wrap up the qualification tests and begin to deploy it into production in earnest in 2011, QLogic’s latest offering should be well positioned to capitalize on the trend towards consolidated and converged environments.

What’s Next

While the market is still emerging, it will be important for QLogic to get a few things done. QLogic should:

• Continue to educate the market and highlight how converged networking adaptors are a strategic platform to enable virtualized, dynamic, and fluid data centers. QLogic should emphasize its ability to run multiple protocols concurrently and dynamically shift protocol support on the fly. QLogic, with its array of InfiniBand technology, may also be uniquely positioned to offer to drive convergence even further as 40 and 100 GbE are developed.
• Work with its existing partners and forge new partnerships to ensure the third generation technology is placed on vendors’ qualification and support matrices as soon as possible to take advantage of an anticipated push for converged networking solutions in 2011.

The Bigger Truth

The converged networking space, led initially to drive Fibre Channel traffic over Ethernet, is now expanding and several players are now offering a multi-protocol platform to truly enable a wire once, connect with (almost) any protocol. QLogic is addressing this need with its third generation technology that provides support for FCoE, iSCSI, and Ethernet. This technology will enable all those protocols to run concurrently and be dynamically allocated without causing disruption—an important feature for those organizations running or planning on running mixed workloads that will take advantage of VM mobility features.

It should be noted that QLogic is first to market with third generation CNA technology and first to market with the ability to run multiple diverse, protocols concurrently. While QLogic has established a competitive lead for third generation technology, a key focus for the company should be taking advantage of this time to market advantage before the competition catches up. QLogic has alluded to a number of design wins for the new products already, so getting those solutions deployed and in production environments will be critical.

As organizations continue to expand their use of virtualization technologies and optimize their infrastructure based on criticality of applications and types of workloads, deploying a universal connectivity platform makes sense. As further proof of the validity of these platforms, vendors are also certifying this technology for use on the motherboard as Converged LOMs, freeing up adaptor slots for additional converged connectivity. 2011 is a highly anticipated year for the adoption of converged networking and QLogic has timed the release of its third generation technology to be well in line with mainstream adoption.

[2] Source: ESG Research Brief, IT Organizations Look to IP Storage Networks to Increase SAN-Attached Server Rates, October 2009.

For those of you who think the moving business is boring, think again.  There’s lots of money to be made schlepping servers, pumping data across town, and otherwise keeping the lights on while massive IT infrastructures move into new homes.

A recent ESG study on Data Center Construction and Consolidation trends shows that an increasing number of companies are committed to building one or more data centers.  Thirty six percent of respondents cited the simple fact of “running out of space” as the primary driver for this trend.  Enterprise class organizations (1000 employees or more) put data center consolidation in their top 3 IT priorities for the next 12-18 months.

The Healthcare industry is leading this charge with 42% of companies stating the intent to build at least 1 new data center.  Financial services firms are not far behind at 37%.  While not quite as many Telecommunications & Media firms are building a new data center, when they do, they are doubling down with 19% stating they are planning on building multiple new facilities.

This trend should continue to drive services opportunities for both specialty infrastructure providers and vendor services organizations such as HP, IBM, Dell, EMC, and HDS.  Storage vendors in particular have been zeroing in on these services opportunities since their devices and the customers’ data are at the center of the story.  HDS is making a big play here with services to assist customers in transitioning to the data center of the future as they pick up and move into new digs.

A data center move is one of the most disruptive events an IT organization can go through.  Going it alone is for the incredibly well resourced and skilled, or the foolhardy.  The anywhere, anytime, tundra-cooled, bigger and better data center will continue to drive opportunities for service providers who can plan and execute high- risk projects like data center moves.

The full Data Center Consolidation and Construction Trends report is available to ESG subscribers at our web site.

You can read Jeff’s other blog entries at Speed & Change — The Business of IT Services.

The federal government is now catching on as well and recently (June 10, 2010) President Obama wrote a memo titled “Disposing of Unneeded Federal Real Estate” which called for a moratorium on new data centers and urged that consolidation plans commence almost immediately. The memo reads,

In addition, in order to address the growth of data centers across the Federal Government, agencies shall immediately adopt a policy against expanding data centers beyond current levels, and shall develop plans to consolidate and significantly reduce data centers within 5 years. Agencies shall submit their plans to OMB for review by August 30, 2010.

The complete memo can be found here.

The Obama administration has also been a vocal supporter of “cloud” computing as well. While many of the large, private enterprises that have embarked on consolidation strategies have retained control, others are looking to outsource their data center or leverage software-as-a-service (SaaS) models. Indeed, the EDS consolidation effort is specifically targeted at enabling more efficient and flexible computing options and ESG research indicates that those organizations that are focused on consolidating data centers are almost twice as likely to be exploring outsourced IT or SaaS models. This should be good news for those selling to the federal market and services organizations with substantial IT outsource businesses.

Read more of Bob’s blog entries at Data Center Continuum.

Overview

In 2010, 52% of mid-size (i.e., less than 1,000 employees) and enterprise (i.e., more than 1,000 employees) organizations will increase IT spending.  This is a significant increase from 2009, when only 43% of organizations planned to increase IT spending during the global recession.  While these organizations have a long list of priorities, ESG’s research indicates that many will increase their use of server virtualization, upgrade their network infrastructure, continue to consolidate data centers, invest in BC/DR programs, and implement desktop virtualization (see Figure 1).[1]

While these initiatives may seem independent, ESG sees a pattern. Many organizations are:

• Centralizing services and applications. This long-standing trend is only gaining momentum. Business applications, collaboration systems, and IT services continue to migrate from secondary data centers and remote offices to large central data centers.  Why?  Many organizations have realized tremendous cost savings, uptime improvements, greater hardware optimization, more automated administration, and reduced power/cooling/space requirements.
• Doubling down on virtualization technology. IT centralization and data center consolidation efforts continue to accelerate as server and desktop technologies from vendors such as Citrix, Microsoft, Oracle, and VMware mature.  Many organizations now run production and mission-critical applications as virtual machines, physical appliances are transforming to virtual appliances, and virtual desktop initiatives are expanding beyond the test and pilot stage; according to ESG research, 21% of large organizations have already deployed a VDI solution. 
• Relying on networks to weave everything together.  While network connectivity isn’t new, enterprises have never been more reliant of high performance/low latency connections for several reasons.  N-tier applications now rely on a tremendous amount of server-to-server communications across physical systems and virtual machines.  Remote office workers who used to connect to local servers now depend upon central services and WAN links.  The growing numbers of mobile workers face a more difficult situation: they have to depend upon VPN connections across public networks.

In total, ESG’s data indicates a steady “race to the center” with a strong dependency on data centers, server virtualization, and networking.  Organizations clearly believe that this can help them get their arms around spiraling IT cost and complexity.  These investments will reverberate on both sides of the WAN, in data centers and branch offices.

The Potential Hazards of Ubiquitous Centralization

When organizations pursue simultaneous data center consolidation, application/services centralization, and virtualization initiatives, they tend to buy a lot of IT equipment and services.  Centralizing and virtualizing applications and services requires new servers and storage systems.  All of these new servers need to be connected together through Ethernet switches.  Finally, providing central services means more bandwidth between data centers and remote users.

Certainly, IT investments provide a strong foundation, but technology upgrades are not always enough.  In spite of new equipment and services, many “data center transformation” projects can lead to application performance problems that can make the user and IT experiences into nightmares.  When productivity plummets and the help desk is flooded with angry calls from the east coast branch, the Boston-based line-of-business manager won’t care a bit about all the money the company is saving on power and cooling costs.

With investments in new data center equipment and increased bandwidth, what causes these performance problems?  They are often related to a lack of:

• Understanding on IT dependencies.  In the rush to consolidate, data center operations staff find it difficult to determine which applications run on specific servers or to map out server to server connectivity dependencies.  Knowing these dependencies is complicated by server virtualization, where workloads move from server to server.  Typically, IT relies on spreadsheets, a slew of configuration databases, or direct user feedback for this information.  Manual tasks and haphazard efforts are a good start, but not the depth of information and visibility necessary for effective enterprise planning and documentation.
• Business impact analysis. Servers and applications are simply tools for employee productivity—and all applications and services are not created equal.  Without analysis, IT can mistakenly make changes to IT services that impact user productivity and employee morale.  When the CRM application consolidation project is scheduled just before the qualification deadline for a sales contest, IT heads will roll.
• Baseline metrics about network traffic and application performance. Oftentimes, IT services and business applications are consolidated sequentially.  This can help avoid problems with each consolidation project, but it ignores the impact that new consolidation project can have on aggregate network traffic.  To get this right, IT needs visibility into network traffic—from source to destination—at all times in order to understand the impact of each change on everything else.

These issues can affect all applications, IT services, and users, but the impact is usually compounded by the WAN.  Why?  WAN connections support a variety of critical and discretionary traffic flows—through a much smaller pipe than the corporate LAN. If corporate employees suffer when it takes 45 seconds to open a file, branch office workers will be rendered completely unproductive.

The Riverbed Alternative

Does IT simply neglect these issues?  No.  Most IT managers understand the importance of IT dependencies, business impact analysis, and network traffic/application performance baselining, but they depend upon either a multitude of management tools or manual processes to adequately address these requirements.

To overcome these obstacles, IT managers need a baseline of IT knowledge and visibility tools that provide real-time granular data about the impact that any IT change has on everything else.  With this knowledge in hand, IT also needs the ability to identify and fix problems immediately—especially as they relate to the WAN.

Many technology solutions are available, but a number are sold as point tools without the integration needed to respond rapidly and reduce costs.  Riverbed Technology provides a suite of products that may be one of the only exceptions to this rule.  With its 2009 acquisition of Mazu Networks, Riverbed added advanced network and application performance and visibility to its portfolio of leading WAN optimization products.  With its Cascade, Steelhead appliances, and Riverbed Services Platform (RSP), Riverbed provides the tools for a consolidation cycle with three phases:

• Discover. Before throwing money at IT products and services, large organizations should discover all assets and map out connectivity, normal behavior, and application performance.  These are the exact capabilities provided by Cascade.  Armed with this information, CIOs can improve consolidation project planning through automation, reducing the costs associated with manual processes or the risks related to the unanticipated problems described above.
• Enable. By mapping out typical application and network performance, IT can choose an appropriate timeframe for consolidation projects without unnecessary business interruption.  This is true of the consolidation project as well as additional phases like bringing a new branch online.  Network and performance administrators can monitor the impact of each step of consolidation projects and adjust project plans accordingly.  When IT staff is asked to evaluate a disaster recovery plan, they can assess traffic on the WAN links to ensure that there is enough bandwidth for backup operations.
• Consolidate. These discovery, planning, and enablement processes can help accelerate overall IT consolidation because changes are scheduled and monitored continuously.  In addition, problems that do arise are quickly discovered and addressed.  When e-mail consolidation impacts the performance of latency-sensitive IP-telephony, IT can pinpoint the problem and take action immediately.  Riverbed can help in two ways here:  Steelhead appliances can help ensure adequate performance when branch office applications are moved to central data centers and the Riverbed Services Platform (RSP) can help consolidate remaining branch office services onto a single platform.

In this way, Cascade anchors IT consolidation and data center transformation with rich contextual information about application performance and network utilization.  Once problems are discovered, however, it is also important to fix them as soon as possible.  This is where Steelhead appliances come into play; minor LAN problems can truly interrupt business processes for remote workers and mobile users connecting over WAN links or the public Internet.  When Cascade discovers a performance problem or network anomaly, Steelhead appliances can be fine-tuned to accelerate business-critical traffic and throttle back less important packets.

The Bigger Truth

ESG’s research data on IT spending is good news.  Large organizations realize the benefits of data center transformation and IT consolidation—and are investing accordingly.  Unfortunately, IT consolidation benefits don’t come for free; without the right knowledge, planning, implementation, and visibility, consolidation can cause more problems than it solves, especially for remote and mobile users.

To overcome these problems, IT managers can’t rely on guesswork or an assortment of point tools.  It is important to proceed with knowledge, data, metrics, and visibility throughout EVERY IT consolidation project.  This is exactly what Riverbed’s Cascade advanced network and application performance and visibility provides.  With Cascade, IT can baseline the existing environment in order to streamline projects, lower costs, and minimize the risk associated with making a consolidation faux pas.  Cascade information can also work hand-in-hand with Steelhead appliances to fine-tune WAN performance, accelerate critical traffic, and throttle back everything else.  As such, CIOs would be prudent to put Riverbed, with its Cascade, Steelhead appliances, and RSP solutions, on their list of potential IT consolidation and WAN optimization partners.

Click here for the PowerPoint slide.

Click here for the full ESG Research Report: Data Center Consolidation and Construction Trends. ]]> http://www.enterprisestrategygroup.com/2010/06/cheaper-real-estate-costs-as-a-driver-of-new-data-center-facility-construction-by-region-and-it-spending-mode/feed/ 0 http://www.enterprisestrategygroup.com/2010/06/factors-that-influence-plans-for-new-data-center-facility-construction/ http://www.enterprisestrategygroup.com/2010/06/factors-that-influence-plans-for-new-data-center-facility-construction/#comments Tue, 29 Jun 2010 20:03:02 +0000 kevin http://www.enterprisestrategygroup.com/?p=19193

Click here for the PowerPoint slide.

Click here for the ESG Research Report: Data Center Consolidation and Construction Trends. ]]> http://www.enterprisestrategygroup.com/2010/06/factors-that-influence-plans-for-new-data-center-facility-construction/feed/ 0