K-12 Cloud Computing and Desktop Virtualization

It’s time for K-12 schools to begin transitioning to “Cloud Computing and Desktop Virtualization”.

Why?

Cloud Computing and Desktop Virtualization have important educational and financial benefits. For the purposes of this post I am going to keep this very simple:

Whether educational tools and resources are located in the ‘Public Cloud’ or the ‘Private Virtualized Cloud’, they are accessible from anywhere, at anytime, via a web browser. This unlocks the resources of the school and makes them available 24x7x7. By doing so we extend learning opportunities and get more return on our software investment.

Schools utilizing this new paradigm are able to increase the number of computers available to students without increasing their budget because they can purchase less expensive devices. Because ‘Public Cloud’ and ‘Private Virtualized Cloud’ applications run on servers, the student device can be anything that can run a web browser, including a $300 Netbook, a $200 iTouch, or any Smartphone.

School IT support teams can be more productive because they are no longer maintaining dozens of educational applications on thousands of individual computer hard drives. The applications now reside on servers in the ‘Public Cloud’ or the ‘Private Cloud’. Since the software is on servers, software does not have to be ‘pushed out’ or ‘ghosted’ to every hard drive. The end-user’s computer accesses the servers and the new software and can use it immediately.

What first steps should I take?

Determine what applications and data you are comfortable having hosted in the Public Cloud, what applications will need to be hosted in your own Private Virtualized Cloud, and what applications will need to remain hosted on local hard drives. Remember that video and audio editing, computer programming, and some high end CAD applications may not be suited for the Public or Private Virtualized Cloud. Planning this hybrid environment is a great first step.

Make a commitment to subscribing to applications delivered from the Cloud. Begin researching alternatives to the software applications and resources that are currently loaded on your desktop hard drives. Whenever there are Web-based applications that are comparable to the hard drive-based applications, give precedence to the Web-based product even if it is not as feature rich and robust.

Begin planning the Private Virtualized Cloud by determining what desktop virtualization strategy you want to deploy:

The least expensive option called Client Virtualization allows for approximately (50) simultaneous users to share the server OS and whatever applications are being hosted.

Another option is called Virtual Desktop Infrastructure (VDI). In this approach the entire end-user machine (Desktop OS, Applications, and Storage) is virtualized on the server. Obviously, fewer VDI desktops (approximately 20-25) can be hosted and run simultaneously on each server. This requires the purchase of more servers and OS licenses than in the Client Virtualization approach.

Although these are the main approaches to virtualization there are other strategies. In reality, most schools will likely employ both Client Virtualization and VDI strategies in their Private Cloud. The planning process will help determine which users use which strategy.

One final piece of advice as you begin this journey, find some independent technical experience and advice. Be careful here since Dell, HP, IBM. Microsoft, and others have connections with specific virtualization companies and thus, their specific virtualization strategies.

Here is a very thorough and well thought out article by Brian Madden on the topic:

“When to use VDI, when to use server-based computing”

Don’t stand still. Don’t be paralyzed by the new terminology or the seeming complexity. The benefits of the Public Cloud and the Private Virtualized Cloud are too important.

It’s time to begin the process of transitioning to the new paradigm.

pete

Advertisements

Sharing Saves Money: Technology Cooperatives

images

We teach our children to share and if we remember to ‘walk our own talk’ technology sharing can save us an enormous amount of money. Some states have formalized technology cooperatives called Educational Service Agencies or BOCES (Boards of Cooperative Educational Services), others do not. If you do not have a technology cooperative in your area, think about starting one.*

Why?

1. Cooperative bidding and purchasing. Instead of asking for pricing for a small purchase of computers for your district, combine your purchase with the other districts in your cooperative. Will the price be cheaper if you are offering to purchase  3,000 computers or 300 computers? The same concept holds true for negotiating pricing for other equipment, software, and services. Join together and build ‘Economies of Scale’ and reap the benefits of ‘Volume Discounts’.

2. Disaster Recovery. Design disaster recovery plans together with other technology cooperative members. Each district can work with another to act as a ‘Hot Site’ to host a partner school district during emergencies. Working together can reduce the cost of ‘renting’ a Hot Site from a private, for profit DR company.

3. Develop a school district site or neutral site as a Network Operations Center for the technology cooperative. The idea here would be to develop a shared Internet ‘On ramp’. The districts in the cooperative would have broadband lines to the shared NOC which would have a large, scalable (hopefully redundant)  pipe to the Internet.

strencom_cloud_connect_wan_diagram-compressed

This is a conceptual diagram from private industry. Individual schools and districts connect via broadband (red lines) to the shared NOC (cloud -located at a school or neutral facility) and from there are connected to the Internet (lightning bolt).

Why is this a good idea? Once again, by combining all the Internet lines and bandwidth, the cooperative can negotiate lower Internet costs.

Also, once all the schools’ data lines come to one location before going out to the Internet, the NOC can put in a centralized firewall for all the participating districts. The same can be done for Internet filtering, spam filtering, intrusion detection, e-mail virus scanning, etc. Think about the savings both in time, effort, and money that having a centralized firewall, Internet filter, and spam filter would offer, as opposed to maintaining a firewall, filter, and spam filter in every district.

The shared NOC could also securely house Cloud Applications that don’t belong on the public Internet. These might range from web-based SIS and Financial systems to a host of educational applications.

4. Shared trainings, consultants, and keynotes. By pooling training and consulting dollars a technology cooperative would be able to offer PD or hire consultants that would be cost prohibitive for a single district.

5. Mature technology cooperatives may consider joint-hires for specialty positions. Much the same as pooling resources for training, consultants, and keynotes; members of the cooperative can find savings in sharing FTE’s that could not be justified in one district’s individual budget.

The need for sharing is there. The opportunity to share is there. The savings is there. So, how do you get started?

Begin a conversation with your colleagues. Keep it simple. Grow from there. If you need help, contact me.

pete

*Full Disclosure: For many years I directed the Lower Hudson Regional Information Center, a non-profit, educational technology consortium of (60) school districts located just north of New York City.

*Also, if you have an educational service agency in your area that is not meeting your needs, take the time to re-engage them so that together you build a more responsive cooperative arrangement. Don’t give up, it’s too important. The dollars you save can be re-allocated to student technology.

Going ‘Green’ Saves Money

When I am asked to help districts save money or financially justify the paradigm shift to One to One computing, I suggest they audit their technology energy use. Shifting from traditional desktop PC’s to laptops, netbooks, or thin clients can save significant amounts of money, to say nothing of it being the environmentally correct thing to do.

A typical desktop computer uses between 65w-250w of electricity. A typical CRT monitor uses 80w and LCD 35w of electricity. You can get the actual amount of energy usage by checking the label on the specific device, or you can use a watt-meter to measure real energy consumption.

So, if we use 158w as an average for desktops and 58w as an average for monitors our total energy use is 216w per computer.

Let’s compute the energy cost of running just ONE computer for a typical school year.

Assumptions:

1. The computer is in use 6hrs per day. (6hrs x 216w = 1296w)

2. The computer is left in power saver mode over night. (18hrs x 35w = 630w)

3. The computer is in use 200 days per year. (200 days x (1296w+630w) = 385,000w)

4. The computer is in power saver mode on weekends and holidays, approximately 100 days. (24hrs x 35w = 840w) x 100 days = 84,000w)

5. The computer uses no energy 65 days of the year.

Total yearly energy cost for ONE computer is 469,000w or 469 kilowatt hrs.

Estimated yearly cost for ONE computer @ .17 per kw hour = $80.

Energy cost for ONE computer over a (5) year lifespan = $400.

Total annual energy cost for ONE THOUSAND computers = $79,730.

Total energy cost of ONE THOUSAND computers over (5) years = $398,650.

Now, lets look at alternatives to the energy hungry desktop PC approach that is so prevalent in our schools today.

A laptop or netbook averages about 30w, most of it related to the display.

A thin client and display also averages about 30w.

Thus replacing a standard desktop with a laptop, netbook, or thin client device theoretically produces an 86% reduction in energy consumption.

Estimated yearly cost for ONE device @ .17 per kw hour = $11

(Savings =$69)

Energy cost of ONE device over a (5) year lifespan = $55

(Savings =$345)

Total annual energy cost of ONE THOUSAND computers = $11,000

(Savings=$68,530)

Total energy cost of ONE THOUSAND computers over (5) years = $56,000

(Savings=$342,650)

Even if we take the ‘best case’ desktop scenario: a 65w computer and 35w display, the energy savings for shifting to laptop, netbook, or thin client devices is 54% resulting in a savings of $227,230.

In One to One implementations, if students use battery power during the day and are required to charge their devices at home, the energy savings can be more than 95% and a cost savings of $378,717.

The yearly $68,530 savings in energy costs (ONE THOUSAND computers) can purchase:

An additional (228) netbooks, or thin clients per year. (@$300 per device)

Over (5) years a school can DOUBLE the number of devices available to students (1140) based on energy savings generated by switching to netbooks or thin clients.

If you are more interested in the traditional route you can purchase laptops and add an additional (86) devices per year (@$800 per laptop) and increase your network by (430) devices over (5) years.

Anyway you look at it there is a good case to be made to go “Green”.

It’s time to shift our technology energy paradigm.

pete

Should I Be Thinking About Moving to a One to One Model?

Without thinking about it consciously many of us change the emphasis of this question to make it a financial one that sounds like this: “Can I afford to go to a One to One model?” Our answer is generally, “No, I can barely afford the technology I have today!” When we think like this we believe we are being ‘realists’; but looking at educational technology this way shuts down many possibilities before we’ve fully explored them.

I like the approach that Bernajean Porter espouses:

Reality is too confining. If we are going to transform education, we need to let go of “reality”. If it is worth doing, then let’s do it. We should say “Yes!” first, then deal with questions of “How?” afterReality is too confining. If we are going to transform education, we need to let go of “reality”. If it is worth doing, then let’s do it. We should say “Yes!” first, then deal with questions of “How?” after.

If we keep deploying technology in the same ways we have for years, it seems to me, we are bound to continue getting the same results. It’s time for a new approach. One that puts technology in the hands of teachers and students so that they can move beyond the ‘many watching one’ model…

many watching one

….and ‘shared pencil’ approach that has dominated our classroom-based technology paradigms for decades.

girl boy sharing

How can we move to One to One financially? technically? pedagogically? There are lot’s of strategies to explore. There is no One Perfect Way to travel this path. In just the last two years the emergence of low-cost Netbooks,  Smartphones, new wireless standards, and  the availability of broadband in the home have made the initial cost of One to One more affordable.

In addition, Cloud Computing, virtualization, blade servers, and other new technologies have made One to One more easy to manage. All in all, One to One is more accessible to the average school district than at any time in the last 30 years.

Many visionary districts have found the answers to their questions and have created dynamic new One to One environments for learning. Many are beginning the journey with a single grade level or a single pilot. If you aren’t exploring and plannng for One to One, you should be.

It all starts with saying,”Yes!”

pete

Pilot Opportunities


One-to-One  & Ubiquitous Computing

‘Pilot’ Opportunities

Ed Tech Journeys is seeking school districts interested in exploring the possibilities of One-to-One computing and desktop virtualization. Right now we are seeking districts for the 2009-2010 school year. To learn more contact me by e-mail:  preilly@edtechjourneys.com

You can get a good idea of what a proof of concept might entail by downloading a copy of one of the final Pilot Reports below:

Deer Park USFD One to One project completed in June 2009


The Shoreham-Wading River  CSD  One-to-One project completed in June 2009

The Wethersfield Schools (CT)  Virtualization project completed in June 2009


The Niagara-Wheatfield CSD  One-to-One project completed in March  2009

If you are interested in learning more, contact me:
preilly@edtechjourneys.com
pete

K-12 Cloud Computing: The Private Cloud

605px-Cloud_computing.svg

Cloud computing is getting a lot of play in the k-12 community and there’s no doubt that there are some wonderful benefits to this model for schools.

Web-based software can be the road to 24×7 access from any location with Internet access.

Software as a service off-loads the costs of servers and the ongoing cost of maintaining them by an already overburdened tech support staff.

Web-based software gets updated centrally and insures that all students and teachers are using the same version.

However, there are some limitations to the public cloud:

Not all applications run in the public cloud.

Storage of sensitive student data is not under the complete control of the school district.

Software becomes an annual subscription and not an outright, one-time purchase. Over time, software budgets will grow as we add more subscriptions.

We still need the local network for policies, printing, grouping students, web filtering, and local storage

We can address these issues by developing ‘private clouds’ within the district.

350px-Cloud_computing_types.svg

We can virtualize desktops and applications and run them from servers in the ‘private cloud’ so that little or no software remains on the students’ or teachers’ computing device. Combining Public and Private clouds provides the best of all world’s:

All the benefits of the public cloud as stated above…

plus

The school can run applications that are unique to their environment.

Access these applications 24×7 from any device with Internet access.

Become device independent (apps are running on servers) thus allowing the purchase of thin clients, netbooks, and other low cost computing devices.

Install and manage applications centrally.

Slow the replacement cycle (software runs on servers so no need to buy new computers every few years.

There is no doubt that we are at a unique crossroads in educational technology. There are huge changes in the ed tech paradigm that are about to take place. The Public and Private clouds are a means to an end and not the end itself.

As we shift to this new paradigm we also make it possible for each student to have their own device and to access their learning resources and files from anywhere, at any time.

pete

Virtualization, Thin Clients, and Energy Consumption

In previous posts I documented a financial strategy that allows the average district to afford ubiquitous and/or one-to-one computing. This week I want to broaden the strategy to energy savings. It’s amazing what a significant savings replacing traditional ‘fat clients’ with ‘thin clients’ can be.

The University of Pennsylvania produced the energy graph below which shows the average PC drawing more than 100 watts during moderate use. This doesn’t include the monitor which on average can draw approximately 75 watts. So, for the sake of today’s post, let’s settle on 175 watts for the average energy use of a typical desktop used in our schools.

penn-u-energy-chart

Below is another chart, from the Mr. Electricity blog, showing ranges of PC energy use.

energy-use-chart

The chart below, from Steve Greenburg, President of Thin Client Solutions, shows the average energy used by several models of WYSE Thin Clients. The 3630 model uses more energy because it has a built in monitor. The other energy readings are without monitors.

thin-client-power-usage

From the same report, notice that the amount of energy consumed by Thin Clients is significantly less compared to the traditional PC.

thin-client-v-pc-graph

What kind of savings can we expect by implementing a Thin Client solution? Let’s look at a district with 1,000 computers.

Assumptions:

1. 175 watts used by each computer.

2. Each computer in moderate use 6 hours per day; 185 days per year.

3. Computers left on overnight and during the summer use approximately 35 watts.

4. The Thin Client solution uses 6 watts plus 75 watts for the monitor.

5. A utility rate of $.14 per kilowatt hour

Using these assumptions, the total amount spent on energy for our 1,000 computers is $64,680.

Now, let’s do the same calculation with the Thin Client solution

The approximate savings by implementing Thin Clients for our 1,000 computer network is $29,291 per year; a 45% savings in energy costs.

The 5 year savings = $146,455

$146,455 can be used to purchase quite a few new $450 devices.

BTW, it’s not only cost effective; but the right thing to do for our environment

pete

Note: Obviously, energy use can vary based on many equipment and usage factors. The savings shown here are illustrative only.