Desktop-as-a-Service: Now, Near or Never?

Presenter: Brian Gammage

This is a change in approach, not a change in technology.

• What does Desktop-as-a-Service mean?
  
• What issues get solved or introduced with desktop-as-a-service?
  
• How do you deliver Desktop-as-a-Service?
  

"Desktop-as-a-Service is an emerging model for provisioning a user environment."

Four perspectives:

• Contractual View: Negotiated SLAs, cost constraint, outsourcing
  
• Technology View: agility and lower management effort, virtualized delivery
  
• Financial View: centralized procurement, leasing
  
• User View: give me freedom, cloud computing
  

It is a pay-per-use approach to delivering client computing in a scalable fashion across shared infrastructure. Provides capabilities on an as-needed basis. Leverages shared services. Not just applications as a service; not just outsourcing; not just leasing; not just a delivery model. Treating IT assets just as we treat other business assets.

The desktop is the user personality, the software infrastructure and physical infrastructure. It is not the applications. A service approach is about redefining the responsibility and risk.

Costs as a driver, maturing technology as enabler as well as sophisticated, mature management tools.

Support business agility: faster reactive and proactive time to market; rapid redesign of user environment; leverage broad external skills; refocus internal skills.

Improve financial position: share common services across larger user base and lower costs; free up committed capital; shift costs and gain flexibility with financial obligations.

Internal to external provisioning can be time-consuming and costly. Co-ordinating multiple players takes work. Users often don't know exactly what they want. Users may not like the experience. The cloud doesn't solve all the problems. Does the "good enough" baseline change?

How might Desktop-as-a-Service be delivered?

Look at whole lifecycle: acquisition; deployment; operations; support; disposal. Split into categories: user environment, operational platform and devices. It is not a single decision—there are many aspects of the "desktop" that could be delivered as a service. We can look at each of those separate decision points.

Virtual Worlds: Beyond Second Life

Presenter: Steve Prentice

Virtual worlds represent an emerging environment which will be incredibly valuable in the future. There is much more to virtual worlds than second life. Second life is still the primary focus of attention. But what are the alternatives? Rule of thumb: 9 out of 10 Virtual World projects within organisations fail. Why have they failed?

Synthetic environments, not games (no goals or mandatory objectives), online platforms, immersive environments—allowing people to meet, interact etc. Best not to fixate on the technology.

What can virtual worlds be used for? Education…

Non-verbal communication; teaching social skills (medical and training applications); one-to-one, one-to-many, several-to-several, educational courses, political canvassing. Particularly useful for teaching social skills—learning through interaction with peers.

Application for Virtual Worlds: Role-based, scenario-driven training exercises; complex situational simulation; presentation of complex data sets through 3D visualization. When training in real-life can be very expensive, consider virtual worlds—simulating an emergency in a city centre, military simulations.

Avatar-enabled Enterprise Collaboration: keep the scope small and tight, allowing people to get together in an immersive environment.

www.qwaq.com : provides a secure, persistent, interactive virtual workspace.

https://lg3d-wonderland.dev.java.net : Open source toolkit to support the creation of 3D virtual worlds using Project Darkstar, jVoiceBridge, Java 3D and Project Looking Glass.

"If you want to see what business leadership may look like in three to five years, look at what's happening in online games."Byron Reeves, PhD. Professor of Communication.

Generating valuable skills for running global teams.

Social networking: understanding demographics and demands.

Questions:

• What are you trying to achieve?
  
• How will you define success and what metrics will you use?
  
• What feature set do you need?
  
• Can you define your intended audience?
  
• Why use Virtual Worlds rather than WebEx or other Web-based tools?
  
• How does this technology relate to your infrastructure now and in 2-3 years time?
  

  
   

We are heading into the trough of disillusionment so exercise extreme caution with new investments and projects (but don't abandon hope).

 

I came here wondering if there was a virtual world solution to recruitment and admissions processes, but come away thinking that there is not a good application for those technologies right now.

Greening the Datacenter

Presenter: Rakesh Kumar

Considering a data-centre as a living breathing dynamically changing organism.

Which critical forces will drive enterprise data-centre strategies during the next five years?

Tier one cities in on Europe and US are maxed-out—there is no capacity left for new data-centres.

Most immediate concern is server efficiency, cooling technologies, processor design, facilities design. In the next 2-5 years: assessing the green credentials of suppliers; PC power management; energy management tools. Longer term still—using alternative energy sources, recycling, data-centre heat recycling, carbon offsetting.

The green data-centre can't look only at the efficiency of the servers themselves—they take little of the energy (6%)—only 1% to the processor itself. Look at the whole data-centre, including cooling and so on. More and more processors are being packed into the same space that contained mainframes 30 years ago. New design envelop needs to see flexibility of cooling. Design in a modular way, with each module's design with a lifecycle of 3-5 years so that you can continue to take advantage of new technologies as you move forward.

What are the best practices and processes that users should follow when designing a green data centre?

Mixing chilling of different kinds: liquid, air. Looking at alternative energy sources for data-centres. Green data-centre can cost about 20% more. Recycling the air is worth considering. Green data-centre is highly monitored—thermal images, thermometers, power management tools.

Considering the data-centre as an organism makes us think about how the data-centre can respond to changing requirements—consider adding more server resources to Agresso in the approach to year-end.

There is research from Gartner that allows the comparison of servers vendor to vendor. Some vendors are more forthcoming than others. Look also at the software tools provided for power management. Consider waste management for the end-of-life for the assets.

Which key green technologies will help us design a green data centre?

Software for power and energy management, maybe just showing what energy consumption looks like (this is consistent with what Simon Mingay has been saying earlier in the week).

Virtualization, consolidation and utilization (again, making sure that every service is still being used).

Green Grid looks at data-centre efficiency: http://www.thegreengrid.org

Proposes KPIs for data-centre efficiency. Maybe we can look at their metrics (which are likely to become standards) to examine energy efficiency in our data centres.

Look also at Gartner "Green ICT Scorecard Model".

Projects and Programs: The Future Must be Smaller, Faster and Easier

Presenter: Donna Fitzgerald

"Current PPM best practices are optimized for a world gone by."

• What business trends will force us to radically transform our project, program and portfolio management practices in the future?
  
• What technology trends will enable the business to move faster?
  
• What should PPM leaders do to prepare for the upcoming change?
  

Moving mail and packages as an example of process change. Sporadic and expensive to start with (1000 BC); government involved, governance increases reliability (400 BC); infrastructure built makes things more efficient (200 AD); competition and classes of mail (1400 AD); tracking of progress via the web (1998 AD); real-time tracking via RFID, mash-ups, wireless.

Time-based competition: focuses on the customer; embraces quality improvements; strategically manages costs—speed as the most versatile competitive weapon.

Projects are no longer change activities between one stable point and another. Instead there is constant change with no points of stability. "Toyota doesn't have corporate convulsions, and it never has. It restructures a little bit every work shift." No Satisfaction at Toyota, fastcompany.com January 2006.

Stakeholders judge a project's success by: product meeting client requirements; gets used by clients; effectiveness for the client organization. NOT: on time, on budget, compliance with process.

WebKinz – research…

Business process goals were: consistency. Now, quick and agile.

Business information/intelligence evolution. From: local, limited. To: end-to-end.

Need to get slides for this presentation: slide showing quadrant diagram with business process goals mapped to business intelligence goals. Relates lower risk, local impact to lower control, more agility.

Looking at "change operations" rather than "project management." Governance focuses on developing the rules for making decisions and monitoring results; management by exception; push decision making down; unit of interest is the process. Contrast with PPM today…

"I get more questions about governance that I do about how work's going to get done. How are you going to govern that project? Who's going to be involved? What's the governance model? Who cares?" Jeff Smith, CIO SunCorp, October 2007. Focus instead on what the real measures of project success should be.

Use a "just enough" approach to project process. Stop watching something, meeting about something; start providing intervention policies, managing only what matters. Mix business and IT staff; hire/use experimenters. Narrow the focus on what needs the executive touch. Adjust project processes to optimize business trade-off and flexibility, not just process consistency.

 

Great Debate: Does Green Information Technology Matter

Chair: John Mahoney

Debaters: Simon Mingay, Mark Raskino (for) David W. Cearley, Rakesh Kumar (against)

The format of this session is quite different from the others I have attended so far. There are four debates and the audience are armed with devices that allow us to vote (like sophisticated, wireless "ask the audience" devices). Two debaters are "for" the proposition that Green IT Matters, and two against.

The audience:

• 91% Believe human activity is driving dangerous climate change
  
• 79% Would spend money to reduce their business' impact on the environment
  

Initial vote: Does Green IT really matter? 77% yes, 0% No, 23% undecided

550 million cell phones retired in 2007. End-to-end, they'd circle the earth and then some. Consumers, regulators, investors and employees are all pushing business leaders to green-up.

Research from the university of Tokyo: For every 1000 tons of carbon expended on IT (efficiently and focused in some key areas), 3000 tons can be saved

End vote: Does Green IT really matter? 73% yes, 19% No, 8% undecided

Key issues: increasing the efficiency of IT (which has a similar size global impact as the airline industry) is easier than increasing the efficiency of aircraft. More important still, IT is a key part of reducing an overall carbon footprint—even if the IT carbon footprint increases, it can have a 3-1 impact on an overall carbon footprint.

The Low Carbon Economy and IT's Role in It

Speaker: Simon Mingay

A 60% to 80% cut in carbon emissions will demand doing different things, not just doing things differently. The winners in a low-carbon economy will think differently.

Exposure for Green issues has exploded over the last year or so. EU is trying to work out how to create a low carbon economy that sustains growth. We are looking at the beginning of a significant and long-term change, and IT is only just starting to look at the issues. Cuts required by 2050 cannot just be on the supply side (by choosing lower carbon alternatives for energy generation). Cuts must also come from the demand side.

Key question: "Do we understand how we would thrive in a low carbon economy?"

A Low-Carbon economy:

• Greenhouse gas emissions is halted and reduced
  
• Carbon becomes a constraining resource
  
• Energy from a fossil fuels is made a constraining resource
  
• Enterprise stakeholders take action based on carbon
  
• Carbon presents both a risk and an opportunity to the enterprise
  

How can we in ISS help the university plan to reduce carbon emissions? We need to fix our own footprint and then help, through application of IT, can reduce the university's carbon footprint. Need to be able to measure advantage of initiatives based on carbon savings. Comparing video conference with in-person meeting for example. Metric used to normalise carbon abatement is Euros per metric ton. So what is the cost of putting in a video conference facility in Euros per metric ton of carbon saved—compared with (say) train travel saved for attending a meeting in person.

What IT can do to help:

• Dematerialization: the process of taking materials out of value propositions (e.g. CD – MP3 download); replacing travel with video conference.
  
• Analytics and knowledge sharing: information management, environmental management; business process analytics (looking at business process from the perspective of reducing carbon emissions)
  
• Process Control: operational technology—e.g. presence based power management; smart metering—real time metering.
  

The Building Blocks of an Environmental Program:

• Energy efficiency
  
• Material efficiency
  
• Water reduction and management
  
• Waste reduction and management
  
• Pollution prevention
  
• Emissions management
  
• Supply chain management
  
• Environmental reporting & EMS
  

Points of focus:

• Paper reduction
  
• Energy efficiency
  
• Travel
  
• Transport
  
• Buildings
  
• Waste reduction and recycling
  
• Real estate
  
• Procurement
  
• Product/service design and development
  
• ICT itself
  

KPIs:

Carbon emissions per student per lecture? Carbon emissions per graduate? Carbon emissions per hour in a lab? Carbon emissions per hour in the library? Carbon emissions per night in residence? … We need to work that out.

Steps towards reducing carbon footprint:

1. Eco-efficiency (internal): travel substitution; flexible working; remote collaboration; materials efficiency (paper reduction)
   
2. Eco-efficiency (external): product lifecycle management, remote support, remote printing (if we need 1000 prospectuses in Malaysia, why not print them there, to reduce the costs (in carbon) of shipping
   
3. Changing the mindset-doing different things: carbon accounting; challenging the principles of "just-in-time" delivery, instead aiming at efficiency of carbon use;
   
4. Changing the value chain: dematerialization; products becoming services (car to public transport; SaaS; zipcar, whizzgo, goloco); localization of the supply chain.
   

"As human numbers continue to increase, casting an evermore unsustainable ecological footprint, it is hugely important that business and industry engages with the problem." Lord Bob May

 

Taming the Rogue User: Coping with Risk, Compliance and Other Business Realities of Socialization

Presenter: Nikos Drakos

Who are the rogue users at the university? An element of "individuality and initiative" is in the definition of a "rogue user." End users finding unusual ways to achieve what they need to—how can we embrace the initiative shown by rogue users to provide additional value…

"From a dozen applications for thousands of users… …to thousands of applications for dozens of users." Highly specific tools, solving individual problems…

Change and risk is increasing as the cost of access to computing is reducing. Risk is higher and change is higher as the number of ways in which an end can be achieved increases—all of which need to be managed.

A graph showing users per application, with an elongating tail—with e-mail, office and windows at the front of the graph, wikis, blogs SaaS Open Source and mash-ups in the middle and spreadsheets, browser plug-ins and PDAs at the tail. Micro-targeted software at the tail—claim is that there is value in the micro-targeted software. 70% of informal survey says that employees should be allowed uncontrolled use of consumer devices, applications and services. More than 60% said that IT does not need to endorse every piece of software or device an employee will use. As an institution, we already live in the world described.

As long as we watch out for: security, privacy, integration, duplication, fragmentation of repositories, data ownership and portability, can we not fill tactical gaps with quickly deployed solutions? Can we not find stop-gap solutions? Could we have used "basecamp" as a project management system during the deployment of Asta Teamplan? Can we be using del.icio.us to track vendors for specific purchases? What are: pbwiki; sprout; zoho; vyew; mindmeister…? Can we continue to engage with web owners across the institution by building "mashable" application interfaces?

Given that we cannot gamble with mission critical services; cannot risk security, quality, reputation; don't want people to waste their time—but in the same way we need to engage people to build momentum for WIP by looking at and re-using existing web infrastructure—shouldn't we determine how we can continue to engage their creativity…? Rather than looking at risk as the reason to say "no", look for opportunities which have low risk and high business value.