TechTalk Columns

Each month, Dux Public Relations Vice President Kevin Tanzillo writes a column on emerging technologies for the online newsletter of the Association for Communications Technology Professionals in Higher Education. Here are his most recent columns:

December 2009

What’s the Next Really Big Thing?

Surely you saw all the news items about a month ago, regarding the 40th anniversary of the first Internet message. Quite a year, 1969. Moon landing, New York Mets winning the World Series, Woodstock, and the birth of the Internet. But while three of those events may be the best remembered, it is the one least remembered that had the greatest impact on America and the world.

No one back then – not even the writers for Star Trek – could have possibly anticipated what would eventually unfold from a simple attempt to link computers. Universities, of course, were the midwives and pediatric care nurses for the birth of the Internet, just as colleges and universities lead the way in technology adoption and refinement today.

Is there anything on the horizon today with the potential for anywhere near the impact that the Internet has had on our lives? We’ve already had the Rise of the PC and the Rise of the Wireless Device in the interim, both of which are closely tied to the Internet’s ascendancy. The PC was essential to broadening the “audience” for the Internet, and wireless devices let us take the Internet anywhere.

A couple of things come to mind in terms of large-scale effects in the future, although we have doubts that even they would be anywhere near the Internet scale. One is Unified Communications (UC) and the other is the PAN, the Personal Area Network.

UC – which we’ll define here as a convergence of all communications – fosters collaborative work, more “aware” communications among individuals, the ability to work at any time from anywhere, and eliminates many of the existing barriers to contact among members of teams, employees, and other groups that need to be connected. UC has the potential to alter the way work is done and the ways people interact in the workplace. It won’t happen tomorrow or even five years from now. Rather, we see an evolution in which people will someday look back and not be able to imagine how anyone ever worked effectively otherwise.

A PAN is an environment in which devices (think Bluetooth headsets as one example) with low-powered radios will communicate with other devices within a small radius, maybe 25 feet. Your PAN would involve elements of presence (making sure friends and colleagues know where you are) or environment (adjusting temperature, lights, music, etc., to your entry into a cubicle or a room) as well commerce, communications, and more. While some of this would be purely personal, there is real workplace potential as well. Like iPhone apps, if you can conceive of a PAN app, someone can create it.

Do you see any other Really Big Thing out there? We’d like to hear your thoughts.

November 2009

Forecast: Cloudy, with a Slight Chance of Lost Data?

I might have mentioned previously that I am rarely an early adopter. I own no HDTV, no smartphone, I’ve never Tweeted, and I actually still use a wired telephone for 98 percent of business calls. This disclosure is to help you assess whether the following comments about cloud computing may be colored by what might be seen as “resistance” to newer technologies. Hey, I get around to them; I’m just not first in line.

Cloud computing is definitely a hot new technology, and with good reason. There’s a tremendous appeal in knowing that a service provider will take care of all the software, all the data storage, and users only need to log in through a web browser and work with any of your applications through the online cloud service. Software updates and upgrades? No problem. Backups? No problem. Well, at least in theory, but we’ll get to that.

Cloud computing (perhaps it needs a snappier name, like “clouding”) makes computing a utility, like electricity. The more resources you use, the more you pay. But a lot of the “burden” of maintaining all those applications and backing up data is shifted to the service provider. Cost savings doesn’t tend to be the major driver so much as the administrative relief.

The downside for many potential cloud computing candidates is the loss of control. Yes, it must be nice to not worry about changing software throughout the user base. But now your users are doing their work “out there” and that’s where your data is. You may have no reason to believe the data isn’t secure, but you aren’t in charge of it, either. If you’re anything like me, such uncertainty is, shall we say, discomforting.

A recent report by the Yankee Group industry analyst organization noted some of the concerns on the part of would-be cloud computing customers. They worry about network congestion, and want to be sure that any and all of their data is immediately available, when they need it, without delay. They’re also looking for a measurable financial return if they move to cloud computing, beyond the “soft” benefits of reduced administration.

Reliability is an obvious concern as well. And an early October problem with one wireless carrier’s cloud service will only feed that worry. It was somewhat of an isolated incident, affecting a couple of million wireless users, but as of this writing, the extent of data loss hadn’t been determined. Isolated or not, though, anything like this makes people think twice when it comes to putting their data at even the slightest risk.

Yet technology does march on, and the Yankee Group report also found that a majority of mid-sized to larger organizations are planning to move to cloud computing in the next two years, most of them starting first with their non-mission-critical applications. This makes sense; proving that cloud computing can be trusted is pivotal to the wide-scale implementation of this strategy. As for me, you can be sure I’ll still be doing our company’s backups and software upgrades myself for a while yet.

October 2009

To Blog or Not to Blog …

Should you blog? Now, we’re not talking about something you might do in your off hours relative to your interests or hobbies; instead, this is about writing something, on the IT/telecom department portion of your school’s website, concerning what you and your department do.

This may be worth looking at in terms of increasing visibility for the good works your department performs on campus. Speaking as someone who is increasingly “ghost-writing” blogs for various clients, I can attest to the “awareness” value of putting your experiences, your thoughts, and your perspective out there.

No doubt you read some blogs, and perhaps you may be a loyal follower of some specific blogs. Every trade publication features blogs, and many industry experts have, for a variety of reasons, started their own blogs. So you no doubt have a pretty good sense of what’s good and what’s bad, which are valuable or at least entertaining, and which are useless.

On the positive side, blogging would offer you a means of spreading the word about projects your department is working on, and these projects’ progress, along with advice and perspective on network security and usage, as well as scads of timely topics unique to your school and campus. Chances are a blog would be a far more effective means of communication than simply posting some IT bulletin on the website and hoping people wade through it. On a personal and personnel basis, it also raises the profile for you and your department, enabling you to let more people know the value of the work you do both on a daily basis and via major projects.

On the negative side, blogging means you have to sit down and write something. Regularly. Frequently. Dependably. And for some people, that is torture. If you are such a reluctant writer, seeing anything beyond a text message or a five-line e-mail as painful, then blogging is probably not for you. But if you can deal with the writing, and you are willing to craft a blog maybe every other day (to keep things fresh), this has potential.

By the way, another option is to have someone in your department who might enjoy writing become the blogger. Or find someone on campus (an aspiring journalist?) who understands technology to do the ghost-writing, at your direction and under your name. You are hardly limited to one blogger, either; depending on the size of your department, you could have several. That would help ease the burden on each individual writer when it comes to frequency of posts.

You would have to determine university policies regarding blogs, but considering how much blogging is likely going on from the faculty and administration sides, there shouldn’t be anything major standing in your way.

A couple of tips for blogging success: Keep posts short, generally aiming for under 300 words (that’s only a little more than the first four paragraphs of this column), and be fairly informal, allowing at least a bit of your personality to show through. Who knows, maybe your blog will get so popular on campus you’ll take the next step, becoming a Twitter tweeter and developing a following. But that’s a topic for another time.

September 2009

Wake Up, Little Computer

One thing that at least 99 percent of us can agree on is that we don’t want to use more electricity than we have to. We can disagree about whether there is global warming or global cooling and what should be done in response, or whether the utilities are stiffing us or are simple enablers of our electronic lifestyles. But when it comes to saving a few bucks on the monthly electric bill, we definitely find common ground.

So whatever your motivation for conserving power, good for you. Every kilowatt counts. Speaking of which, are you old enough to remember “Reddy Kilowatt?”
Pictured here, he was the mascot for hundreds of U.S. electric companies throughout much of the 20th century, as a friendly electric-bolt stick figure who wanted you to use as many electric appliances as you could. I always liked Reddy, but then I was just a kid and not paying the electric bills.

Anyway, to get to the point here, there’s a technology that you might find useful in your efforts to conserve electricity. It’s called Wake on LAN, and despite the unusual name, it is basically a means of putting computers on the network into some form of power-saving sleep mode, and then easily rousing them remotely via a network message. This way, computers don’t have to remain fully powered on for network technicians to troubleshoot them, upgrade software, do maintenance, etc.

From what we’ve learned, Wake on LAN is pretty easy to enable, and while we don’t need to walk you through the process step-by-step (you can find details online) it essentially involves setting the properties of the network card. It sounds so straightforward that even I could handle it.

The waking-up process involves sending a message to the dozing computer via what’s called a Magic Packet. You know, if I remember right, a couple of guys at my dorm in college tried to sell me something like that back in the 70s. There were a lot of magic packets floating around then, which explains a lot about that decade’s music and, for that matter, clothing styles and our choices of U.S. presidents.

Seriously, though, it is the Magic Packets that are broadcast over the broadcast address for a particular subnet – or the entire LAN if you take the extra trouble to configure it. The network cards of the computers, which are either in sleep mode or may even be completely powered off, check the packets. If the request is valid and contains the appropriate network cards’ MAC addresses, the computers proceed to wake themselves up and get ready for service. Even if the computer is completely powered off, the network card can draw enough power from the motherboard to service a Wake on LAN request.

In enterprise-level environments, Wake on LAN is typically used along with a management system that stores the subnet and MAC address information.

We read that simply turning off PCs at night can save up to 60 percent of an organization’s IT energy consumption costs. Whether that’s an appropriate figure in your case or not, the bottom line is that you can save money – and maybe the planet – by powering down whenever possible. Wake on LAN is a tool that can help you do that.

August 2009

Technology and the Death of the Vacation

Okay, maybe the headline is a little strong. Technology hasn’t killed the vacation, but it sure has wounded it.

This seemed like an appropriate topic for August, since this is about the time when people are coming back from, or heading off on, a vacation before the school year really cranks up.

Maybe I’m just getting crabby in my old age, but I can’t help but look around and wonder whatever happened to the true vacation. Perhaps those of you over 40 remember them – you would leave the office for two weeks, and no one would hear from you, or contact you, during that entire period.

By the way, does any white-collar worker take a two-week vacation anymore? I don’t think I’ve had one since sometime in the early 90s.

Technology is a wonderful thing, in so many ways. The ability to be in touch at any time is truly a marvel, and if you’re a person who travels a lot on business, the advances of the last dozen years have been a salvation to you, allowing you to keep up with your “office” work even when you’re on the road. And as a tourist abroad, having a mobile phone in a pinch, and not needing to figure out how to make a call from a payphone when you don’t know the language, can be a lifesaver, literally and figuratively.

But now the flip side. Technology has also made it all but impossible to “get away.” You are reachable by cellphone almost wherever you go, and you are connected to e-mail via a smartphone or a laptop even when you make a vain attempt to take a vacation.

Part of this is the nature of our jobs. In my case, clients and news media folks have to be able to reach me, and in yours, the people who work for you, and those to whom you report, have to be able to get in touch if there are network problems. I suppose that makes us “important,” and that’s good, but it also keeps our batteries draining at a time when a vacation is supposed to recharge them. The other part of the blame lies with ourselves, since who among us can say we aren’t addicted to being in touch, at least at some level?

About the only way to take a “real” vacation anymore is to take a cruise, since it’s so hard to stay in touch there, or do something incredibly remote, like climb Mt. Denali in Alaska. Otherwise, you find yourself checking e-mail at least in the morning and evening, and keeping the phone by your side.

But with the right attitude, you can cope. Once you’ve checked e-mail in the morning, that still gives you at least 12 hours of free time before you’d feel the need to check it again. Turn the phone off and only check voice mail at lunchtime. If you can really relax your mind and spirit in the times inbetween, then it’s a vacation and it has done its job. At least that’s what they tell me.

July 2009

Preventing Data Leaks is Better than Plugging

If you’ve spent any time shopping for big-screen TVs, and we use the acronym DLP, you’re going to think we’re talking about the projection technology behind some models, and wonder if this has become a home entertainment column.

Well, don’t worry. If there’s anything we aren’t concerned about as we write this column, it’s entertainment. Nope, we’ll happily bore you so long as we can edify and inform you in the process. Granted, these aren’t mutually exclusive objectives, and maybe someday we’ll learn to write for entertainment (until then, that book deal just isn’t going to happen). But for now, we’ll settle for conveying useful information.

So let’s get back to DLP. The acronym we’re talking about is Data Leak Protection, known also in some circles as data loss prevention. We prefer “leak” over “loss,” because it just sounds more fixable. You can stop a leak, but a loss seems like a done deal.

What DLP refers to is systems designed to centrally manage your data, whether it’s moving from point to point, is currently in use, or is sitting in storage. The idea is to prevent unauthorized access or use of confidential or sensitive information.

This isn’t just about hackers, whether they’re inside or outside your campus. It can also be about making sure that only the users with proper authorization are perusing or moving particular data. Or that someone isn’t absent-mindedly forwarding a sensitive e-mail trail just because the item on top is related to some other topic. This is clearly an issue when it comes to medical or classified research, but it can affect an institution in lots of less obvious ways.

DLP systems – of which there are two general types – typically use what is known as “deep content inspection” to accomplish their objectives. That means taking an extremely close look at all the data and analyzing key words, patterns, and other cues, then associating those with user rights and authorizations for the sake of keeping things kosher.

DLP systems are available as network- or gateway-based and as host-based. Network-based are usually dedicated platforms that tend to be simple to install and do a reasonably good job for their cost. Host-based systems run on users’ workstations or on servers, and have an edge in that they can monitor and control access to physical devices – for instance, mobile devices with data storage capabilities – and in some cases can access information before it has been encrypted. But their downside is that they need to be installed on every workstation in the network; yet they can’t be used on mobile devices or non-networked equipment.

In a world where regulatory requirements continue to get more stringent in terms of what data needs to be retained and what needs to be kept confidential, many institutions can afford precious little tolerance for data leaks. It may be in everyone’s best interest to make sure the data bucket is absolutely leakproof rather than stockpiling corks to plug the holes later.

June 2009

How Safe Do You Feel?

Do you feel safer than you did five years ago? We’re not talking a crime-in-your-neighborhood or global-terrorist-threat sense of safety, but rather your campus network. With all the threats out there – an ever-growing list of dangers and nastiness – we would have thought that most ACUTA members see their networks as more vulnerable now than in the past.

Not true! In a security survey conducted at the Annual Conference in Atlanta, 84 percent of ACUTA members said they believe their networks are more secure now than five years ago. Good news, to be sure, but it still took us by surprise. Overall, when asked to rank their own networks from 1 (poor security) to 5 (best security), the aggregate average was a 3.7, further indicating a healthy amount of confidence among ACUTA members in their network security.

But confidence doesn’t mean they are ignoring the dangers, as is reflected in their responses to the survey’s inquiry about specific network threats. The area of their networks that the largest percentage of respondents consider most vulnerable is the usage of mobile devices. Thirty-five percent of respondents see these devices as a weak link in their security chain. Other top vulnerabilities are internal controls (named by 30 percent of respondents), student downloads (29 percent), student hackers (23 percent) and Internet access (20 percent).

Asked to rank several threats based on their level of danger to the network, the conference respondents laid them out as follows:

Viruses and malware
Internal controls
Phishing or vishing
Student or external hackers
Student downloads
Physical access
Spam
Mobile devices
Internet intrusion

The most common approach to dealing with security threats is education of students and staff, which 58 percent of respondents said they are doing. Tightening internal controls and addressing problems through new and improved systems were strategies each named by 44 percent.

About half of the survey sample still bears bruises and scars from the security battlefront. “Significant” security breaches were reported by 47 percent of them, and of those who were hit, 71 percent said the damage was at least limited to only minor exposure of confidential information, along with some inevitable campus and public embarrassment. No doubt this kind of experience strengthened their resolve to make their networks more secure and keep a second breach from happening.

Among other interesting findings from the survey, 72 percent of respondents said their schools do have individuals or teams dedicated to security, but only 18 percent of schools use enhanced login, in which a user must type letters from a graphic or identify an icon to gain access. Also, the number of schools that have seen the need to delay any technology projects out of concern over security is relatively low, at 24 percent.

“Winning” the security battle may be an impossible dream, but at least this survey shows that ACUTA members don’t see themselves losing it, either.

May 2009

Revisiting the iPhone in an Ever-Careful Way

Early adopter is not a term you would apply to me. Never have been a gearhead, and probably never will. Heck, I’m still considering whether to finally get a microwave oven. (Just kidding!) But in my role as a humble reporter, I have to make sure that I don’t let my next-to-last-in-line tendencies color my writing.

So just about two years ago, when I wrote in this corner about the new Apple iPhone and the need to be careful and judicious in considering it as an enterprise communications tool, I had to wonder. Was I being TOO cautious in my “advice?”

This time, it appears not. Lots of smart people were suggesting the same thing then and over the past couple of years, so it probably wasn’t bad advice. But in fairness, now that time has passed, it is only fair and balanced to point out that the analyst firm Forrester Research has declared the iPhone to be just about enterprise-worthy.

In a report last month, Forrester concluded that concerns over security no longer are a huge obstacle to adoption of the iPhone as a legitimate enterprise device. Apple has apparently addressed these concerns sufficiently, particularly in the latest version, the iPhone 3.0 operating system, available this summer. Among its advancements is a requirement that the user sign into the VPN (virtual private network) each time, rather than be automatically signed in. A small, but important, change.

In the Forrester report, three enterprise iPhone-using companies were profiled (though none in higher education), and while challenges remain, there is clear indication that the iPhone and the way that IT departments are reacting to it have both positively evolved since the introduction of the device two years ago.

The Forrester findings suggest that Apple has improved in its openness when it comes to applications and operating system development, paying more attention to enterprise input and concerns. And with tens of thousands of applications, it has a lot of appeal to developers.

So integration of the iPhone is obviously worth a much closer look these days, but just remember, it never hurts to be careful and deliberate. Oh, there I go again.

April 2009

After a Long Dry Spell, We Find a Good Acronym

It has been a while since we’ve had a really good acronym – longtime readers of this column know my weakness for them – but once we saw this one, we knew we had to write about it. Serendipitously, it also happened to be tied to a topic we were planning to write about anyway.

The acronym is VoLGA, like the river in Russia (the longest river in Europe, by the way, for you geography buffs). It stands for Voice over LTE via Generic Access. So not only is VoLGA an acronym, but it contains other acronyms. It’s a veritable AL’s (acronym lover’s) dream.

The term does require a bit of explanation. First, take LTE, which was to be our original topic for this month’s column. LTE is Long Term Evolution, the mobile communications protocol of the (near) future. It’s kind of a Promised Land of mobility – beyond all the 2G/3G/4G labeling – in the form of a potent and highly efficient network for carrying all types of broadband traffic. Migration to LTE is in the plans of pretty much every major wireless carrier, although we won’t really start seeing commercial LTE until sometime in 2010.

Once it is available, LTE could be a good solution for mobile broadband access for higher education organizations and other enterprises. But of course, in order to be a solution worth considering, it needs to be able to deal effectively with voice traffic as well. People do still talk to each other at least occasionally. Hence VoLGA, whose goal is a standard to support circuit-switched voice and messaging services over LTE.

To do this, VoLGA supporters want to leverage the existing Generic Access Network (GAN) standard, better known by its commercial name of Unlicensed Mobile Access. Those supporters include a number of well-known names in the mobile world, and they are looking to enable providers to migrate wireless voice and messaging traffic from emerging LTE networks onto the existing infrastructure, to avoid a future one-network-for-data and one-network-for-voice scenario. Seems reasonable, considering that most everyone is moving to VoIP for the very same reason – why do something in two networks if you can do it in one?

There are other options for achieving the same end goal (though none with as good an acronym), so VoLGA may or may not turn out to be the path the industry takes. And to a large extent, this will all be driven by what makes the most sense for the wireless service providers anyway, since it’s their delivery networks that are most affected. So it isn’t like we end user types can have a lot of influence. But in the meantime, if anyone asks you about VoLGA, now you know that it’s not just a river.

March 2009

FCoE May Figure into Your Storage Future

Who would have thought that when Bob Metcalfe came up with Ethernet back in the early 1970s that it would still – nearly four decades later – be the basis for so much of what we do in networking and communication? It’s like the Princess phone evolving as the core of all our voice telephony today.

But then, the early 1970-ish timeframe was also when optical fiber was developed, and that has certainly served us well in the intervening years. So with a tip of the hat to Mr. Metcalfe and his fiber-optic friends, we say thanks for coming up with technologies so versatile, and with so much evolutionary potential, that we can still be marveling 40 years later about the great things we can do with them.

On that note, we’ll take up the topic of one of the latest Ethernet developments, Fibre Channel over Ethernet. FCoE, as the acronym lovers know it, is one of the newest breeds of storage technologies, and even as you read this it is moving through the standards process. FCoE’s purpose is to enable network convergence and cost-effective storage area network (SAN) expansion in the data center

The idea is to move native Fibre Channel traffic over 10 gigabit per second Ethernet, alongside standard network traffic. In this instance, in order to meet the demands of Fibre Channel, the Ethernet is a still-in-the-works enhanced Ethernet, with greater flow control, dynamic bandwidth allocation, and better communication between end nodes and switches, designed to overcome latency and packet loss issues.

The benefits touted by FCoE’s backers, most notably Cisco, are seamless extension of Fibre Channel SANs, increased IT efficiency, unified management, and overall reductions in the total cost of ownership.

Other options have their defenders as well, who say that storage technologies such as iSCSI or InfiniBand (which we wrote about here way back in July 2003) could work equally well, particularly by using the same enhanced Ethernet that is still being worked on in the standards bodies. These folks say that while FCoE might have the edge on speed and power, it would also be more expensive than iSCSI.

So if changes to your storage network are in your short- or long-term future (should your budgets permit), now is the time to begin watching these developments closely. Some of your decisions will depend on how much of your data is “mission critical,” as well as your existing data center infrastructures.

February 2009

It Sounds Scary, But “VoIP is Dead” Declaration Isn’t Cause for Worry

Don’t know how many of you find time in your busy schedules to monitor the “blogosphere,” but there has been a bit of a mini-debate going on in the last few weeks about the health – or lack thereof – of VoIP. This debate picked up momentum when one of the more influential bloggers gave some credence to the contention that “VoIP is Dead.” Since then, there has been lots of commentary on this idea, as you might imagine.

Rest assured that this crazy talk shouldn’t even remotely affect your network convergence plans, whether you’re bringing a unified network to campus for the first time or expanding an existing setup. It’s an interesting debate, but one that is peripheral to the larger issue of convergence.

If you look at VoIP in a vacuum, as a standalone application isolated from all the other things that IP enables, you can see the logic behind the VoIP-is-dead argument. Few organizations today would embrace network convergence for the sole purpose of being able to carry voice over the data network. That just isn’t the way the higher education world, or the corporate world, are converging.

Today, it’s more about EoIP – Everything over IP. And that includes voice. You have an entire laundry list of applications as well – conferencing, collaborative work, remote learning, unified messaging, location-based services, fixed-mobile convergence, and traditional voice telephony. These are all being merged for transport over a robust IP infrastructure.

Despite the overheated VoIP-is-Dead rhetoric, the debate is really a healthy thing at this point in time. It has become too easy to lump all the above-mentioned forms of network convergence under the misleading heading of VoIP, rather than focus on each application and the particular benefits that it brings to the organization and its users.

In the debate, we favor the side that says voice (VoIP) as one of the EoIP applications is quite healthy, but simply not as much of a novelty as it may have been a few years back. But that’s pretty typical as technology advances. If you go back a dozen years, e-mail was quite a novelty, and half a dozen years before that, faxing was Hot Stuff.

The EoIP progression is akin to the development of the iPhone, in which actual voice calls are only a small subset of all the things that handheld device, and others like it, can do for you. Just as hardly anyone buys a mobile phone today only for its voice capabilities (“Jitterbug” customers perhaps being the exception), moving to IP simply for the sake of voice is as unrewarding as it is illogical.

January 2009

A Clean Slate, and a Clean Network, for the New Year

Well, here we are in 2009. Normally we’re incredibly optimistic at the start of a new year, but with the economic uncertainties, a sunny outlook is a bit hard to come by this time around. Still, because we consider pessimism a contagious disease, we’re going to stay away from the dark side. They say “technology” as an industry will be less affected than some other areas; we can hope They are right, and that we’ll all soon be back on our way to a less treacherous economy.

In the meantime we can each cope by optimizing our opportunities, our outlook, and even our networks. Which is our not-so-smooth way of leading into this month’s topic, wide area network optimization. It just seemed timely – in the same way that we clean the slate at the start of a new year – to talk about unclogging the network pipes as well.

With WANs, whether you’re connecting a number of campuses in a region or state, or linking small satellite facilities to the main network, it’s all about performance. That means making the best use of bandwidth and minimizing the transmissions and applications that jam things up, particularly in the busiest times.

One way to aid in this is caching. The more you can hold documents, files, and applications locally, the less you have to chew up bandwidth going across the WAN to pull them each time your distant users need them.

There are essentially three categories when it comes to caching – Cold (actually, not yet cached), warm, and hot. Cold documents or applications have to be pulled across the WAN when demanded, and the idea is to cache them after the first request. Obviously, this initial request consumes the greatest bandwidth. But once documents or applications are in the local server, they can be held there on a warm or hot basis for further use.

With warm caching, after a user requests a document or application, the local server responds with the locally held version but checks across the WAN for any updates before delivering it to the user. Hot caching is the most efficient, particularly for frequently accessed materials. With the proper application of a continual refresh algorithm, it ensures that the cached object is continually updated. Thus when the request is made, there is little cross-WAN bandwidth used and the user does not have to wait for the update.

How much bandwidth can be kept off the WAN by caching depends on the type of work done by each site on the network. But if you do the math with file and application sizes, you can see how caching helps reduce overall WAN bandwidth. The result is better performance for everyone, as well as cost savings, since it reduces the need to add circuits.

Savings really can add up via small increments. For instance, we could reduce oil imports and our carbon footprint significantly if each of us just used one less gallon of gasoline a week. By the same token, using just a bit less WAN bandwidth each time an application is launched or a document downloaded can make a sizable difference in performance and in holding off the need for additional circuits.

duxpr