TechTalk Columns
Back to Current Tech Talk Columns
December 2008
If You’re Not on the List, You Don’t Get into the Club
Before
we get into our main topic, we just want to comment on a recent study of spam
by researchers from two
Turns out – based on their study, which involved the temporary takeover of some existing spamming systems – that the response rate for spam is about one in 12.5 million messages. Over 26 days the researchers sent out about 350 million spam messages, and ended up with 28 suckers who would have bought what they were selling, had they really been selling something.
The bad news is that even with such a microscopic response rate, spammers can still make a fair amount of money (we refuse to refer to it as “earning a living”) just based on the sheer volume of garbage they peddle.
Now, on to the main topic, the protection of the network. Given the battle that everyone wages to keep bad stuff out of the network, a new approach sounded interesting: focus only on letting the good stuff in. Bottom line is that only approved applications get through the castle gates, with everything else falling harmlessly into the moat.
It’s called application whitelisting, and its advocates make a good case for this approach, as opposed to blacklisting, in which all energy is devoted to identifying, rejecting, and cleaning up network threats. Blacklisting is the traditional approach, with firewalls and anti-threat solutions that determine what is bad and keep it from entering and harming the network.
But, the whitelisters say, blacklisting has shortcomings. The biggest, they say, is that it is a reactive strategy. It requires that a solution recognize a threat. Thus solution vendors have to identify new threats daily and immediately update their software to combat them. But in the time lag between the first strike and the software update, networks are vulnerable. Also, the list of threats continues to expand almost exponentially, requiring significant processing power at the endpoints to stay vigilant against everything the bad guys throw at the network.
Whitelisting, on the other hand, is targeted at controlling what you know, the authorized applications. You maintain a very short list of approved programs, and you identify and validate the authorized program as it loads, checking its file digest, file location, and file size. Any program other than the approved ones is not allowed to load or execute. A threat may mimic an approved program, but because its core properties will be slightly different from the legitimate program, it goes nowhere. Same goes for malicious attempts to establish rootkits to control remote systems and injections of bad code via buffer overflows.
Whitelisting, its supporters say, can even protect unpatched vulnerabilities in application software from exploitation, which relieves the IT department of the intense pressure of immediately implementing patches.
There are a number of application whitelisting vendors out there, whom you can find easily with a bit of online searching. Caution: Not all whitelisting approaches are created equal, so be a smart shopper.
November 2008
For Most, the Big Change on Campus is Wireless
Most of you won’t be surprised when we tell you that according to the latest ACUTA survey, the explosion of wireless networking is not only the biggest communications trend in the last few years, but it is a trend that is continuing, with nearly three out of four schools planning to expand their networks over the next two years.
This latest survey was done
in part at the Fall Seminar in
This survey asked members to identify one Big Change in their cabling and wiring infrastructure over the last three to five years. For 60 percent, that change was deployment of wireless networks. Dwarfed by that super-majority were the 13 percent who installed fiber-optic cable and another 13 percent with major rewiring projects for technology upgrades.
Two out of three respondents said it was the demand for “connectivity anywhere” that drove their Big Change, while 40 percent said the evolution of communication styles was a major factor. Interestingly, both of these are clear drivers for wireless. Meeting growing capacity needs, and migration to Voice over IP and Unified Communications, were additional drivers cited, at 33 percent and 23 percent respectively.
The single greatest benefit of their Big Change, respondents said, was network access anywhere and anytime (42 percent cited this), while user convenience came in at 23 percent, network efficiencies 17 percent, and greater bandwidth 10 percent. On the downside, 56 percent said the cost of implementing this change was the biggest challenge. Another 21 percent said their biggest hurdle was the locating and installing of the many wireless access points needed for coverage.
What’s next? Well, 71 percent said they plan expansion of their wireless network – or installation of one if they haven’t put one in already. Another 19 percent are planning additional rewiring projects. For a timeline, 39 percent of respondents expect to take these next steps within six months, and 34 percent in six months to a year.
Asked how the Big Change affects themselves and their departments, respondents said the highest-impact issues are ever-tightening budgets (67 percent), a greater need for long-range planning (63 percent), the need to learn new technology skills (54 percent), the fact that growing campuses mean more responsibility for their departments (44 percent), and the challenge of finding employees with the right mix of skills (35 percent).
In half of the surveyed schools, it was the IT/telecom department driving the major change. Surprisingly, in about one-fourth of the cases, students got the credit for pushing the change, while the remainder was evenly split between the administration and faculty/staff.
Some other interesting results from the survey:
- 90 percent of schools say they have a “copper/fiber/wireless” mix on campus, with just 2 percent indicating their campus is “mostly fiber.” Not sure what we were expecting there, but it just seems a bit more lopsided than expected.
- Only 4 percent of respondents said their Big Change had decreased their ongoing operating costs when asked to point to the biggest benefit. So much for ROI.
- Being forced to “do more with less” due to budget-forced staff reductions was reported as a long-term trend for 19 percent of survey respondents, but conversely, another 19 percent said their budgets are going up. Almost makes you want to track that money trail, doesn’t it?
October 2008
Why Not Share That Extra Computing Power?
Extending computing resources cost-effectively has been a campus challenge pretty much since the days of punch cards and computers the size and weight of delivery trucks. There just never seems to be quite enough budget to do all that you want to do.
There may be a solution out there for you, one that’s as simple as what Mom told you and your sister when you both wanted that last cupcake: Share!
A
The company is NComputing (www.ncomputing.com), and it offers two types of sharing. One, designed for garden variety PCs, allows up to seven people to share a single PC’s computing power. This is via a direct connection, with the PC located within 10 meters. For a server or high-end PC, up to 31 people can share its resources via an Ethernet connection.
The way it works is that each user gets a small device that connects via cable or Ethernet connection to the main computer. The user’s keyboard, mouse, monitor, and peripherals all plug into this device, which has no CPU, memory, or moving parts. Virtualization software running on the PC creates identities for each of the users so they can share applications, storage space, and other resources.
According to the company, more than 250 colleges and universities – including some ACUTA members – are using this approach. It stretches budgets, makes upgrades less overwhelming (rather than have to upgrade seven computers for seven users, for instance, only one computer needs an upgrade to serve seven users), and is actually environmentally beneficial as well. The device to which users connect consumes only 1 to 5 watts of power, compared with about 115 watts for an actual PC. That not only cuts power usage, but generates less heat, which can create a secondary benefit by reducing cooling needs.
We talked to the company about issues such as software licenses, performance, and user access rights, and it all seems pretty straightforward. Software licenses are really no different than they would be if you had independent PCs or terminals on a server. Performance in situations where you have all the users working at the same time is designed to keep everyone as consistent as possible. If there are 30 users, each user is allotted essentially one-thirtieth of the computing power, so if one person is using graphics-heavy applications, for instance, only that person is affected, rather than taking “bandwidth” away from anyone else. And as far as user access rights, setting them up using the company-provided software sounds pretty simple.
September 2008
ACUTA Member Schools Are Mostly in the Green
Based on a survey taken at the ACUTA Annual Conference, two out of three member schools have gone, or are in the process of going “green.” And most of the ones who haven’t say they really want to. So for that, we tip our hat (our organic, earth-friendly fair trade hat) to you. We were going to say “more power to you,” but it’s actually about “less power” these days.
By the way, we appreciate those of you who took the time at the conference to fill out the survey form. The more people who do that, the more accurate our surveys are.
With that said, 65 percent of survey respondents reported that their schools have bought new equipment, launched distance learning and online education programs, instituted energy saving policies, and are otherwise moving to reduce energy usage and aid the environment.
Among the 35 percent who haven’t yet gone green, three out of four of them are looking into ways to save energy, operate more cleanly, and cut waste out of their information communications technology operations. They’re just being held back by budget limitations, difficulty in finding energy-efficient equipment and proving future cost savings, among other obstacles.
The single most widespread pro-environment step taken by the already “green” schools is to recycle computer and networking equipment, diverting it away from landfills. Eighty percent of the “green” schools are doing this. Seventy-three percent reported that they have purchased more efficient, energy-saving equipment and 63 percent said they have implemented a policy of reducing the amount of printing.
“Powering off” whatever equipment they can whenever possible is a practice at 55 percent of the green schools, while 29 percent have revamped their data centers and 20 percent have simplified their networks, both with energy savings in mind.
Surprisingly, 27 percent of the “green” schools say alternative energy sources are providing some of the power on campus (impressive figure!) and 25 percent say there is at least some telecommuting going on among faculty or staff. Also, 22 percent and 18 percent have implemented or expanded their distance learning and online education programs, respectively.
The survey indicates that while energy and cost savings aren’t easy to document, most schools do believe their efforts have provided or will soon provide a return on investment; also, 35 percent of the “green” schools say they have already seen a payoff in the form of an enhanced environmental image.
Growing up, I always dismissed my late mother’s Depression-era philosophy of “use it up, wear it out, make it do, or do without” as mere cheapness. She’d be most pleased to see we’re all coming back around to her way of thinking at least when it comes to energy. Excuse me while I go open some windows and turn off the air conditioning.
August 2008
Yes, Your Disks Will Fail Sometime, But How Do You Know When?
Now that it’s August, school kicks into high gear soon, and one’s thoughts turn to … storage. That may seem like an odd connection, but as someone who has ferried his share of teenagers to school, I always look at the rental trailer and think, “Where is she going to put all this stuff?” You have to be thinking the same thing when you see all the vehicles flock to campus and jockey for the parking spaces nearest the dorm doors.
At least that’s the case for the cars with girls in them. Sometimes it seems like all the boys bring is a computer, a PlayStation or something like it, and a gym bag. One trip up to the room and they’re done. Maybe they should lease out some storage space to the young ladies.
Anyway, that’s why I can’t help but equate this time of year and storage. And along those lines, there was an interesting piece online a few weeks ago about an issue that affects storage and the management of it – assessing just how confident you should be about the durability of your hard disks.
Those of you who deal with disks and disk arrays certainly know the acronym “MTBF.” Nope, it’s not text-speak for “My Two Best Friends”; it means “mean time between failures” and is usually expressed in hours of service. This is how disk makers rate their products, so you know how long they’ll last.
Problem is, like the overly optimistic figures we see when it comes to auto mileage, these hours of service figures can’t be taken literally. The conditions under which they were assessed may bear scant resemblance to your real-life conditions, unless perhaps you work in clean suits in places where the temperature never varies more than half a degree and static electricity and dust don’t exist.
For many disks, the MTBF figure would seem to indicate that the unit should be lasting for decades, but that won’t be happening. Even if they did last anywhere near that long, new technology would displace them anyway. But it’s critical that they don’t fail when you least expect it or are otherwise unprepared.
Bottom line is that MTBF should be viewed more on a comparative basis than as some sort of guarantee. If you have a choice between disks or drives, and other factors (such as price and reputation for quality) are consistent, go for the one with the higher number. So, like a Prius will get you better mileage than a Hummer, but neither one may get what it says on the sticker, you at least know which one will sip, rather than gulp, $4-plus gasoline. Then you won’t be stuck (stretching our automotive metaphor even further) with a Yugo or a Ford Pinto of a disk.
June 2008
If Only Aspirin Worked for Network Headaches
In the last two months, my esteem for ACUTA members has climbed to new heights. In that time, our little public relations agency has put in remote networking capability, moved our office to a rural area, and installed a sophisticated voice mail system. Looking back on the challenges we encountered along the way, I have to marvel at how ACUTAns can run networks of the scale that you do.
From server software that wouldn’t support enough remote log-ins to phone wiring done incorrectly in the new office to a router that wouldn’t recognize the Internet connection, yours truly has had enough technology headaches to last for years. In fact, even as I write this, the router problem hasn’t been completely solved, and we have had to resort to a slightly updated version of Sneakernet to get things done.
Sneakernet, for the benefit of anyone unfamiliar with the term, was the practice of copying files onto floppy disks (remember those old 5¼ -inch, fragile beasts?) and taking them to the co-worker who needed them. Because you were walking all the files from place to place, the network was based on putting miles on your sneakers, hence the term. It was sort of a play on words on “Ethernet,” which was only a young’un back then.
Anyway, being our agency’s IT and Phone Guy, I have felt the worst pain of all these networking challenges. It may go without saying, but I write and do public relations far better than I do IT. So we’ve become good friends with the phone system installer and our by-the-hour outside network guru. Not so good friends with the electricians, though, who we have had to keep summoning back in order to get the wiring right.
Certainly, as each of you looks around your campus and your data center, you see things that could be done better. Perhaps you feel like the whole thing is barely contained chaos that could unravel any minute. But it’s working, and you should take pride in that. You certainly have my respect and admiration.
But while we’re on the
subject of looking for ways to do things better, the ACUTA Annual Conference is
coming up next month, and it is filled with great sessions that can help you
improve your knowledge, skills, and your personal network. If you can make
yourself available and get to
Back in my editor days, I always looked forward to covering the ACUTA conference. Our magazine (Communications News) was always seeking good user stories from various schools. And of course, we found them. In retrospect, though, perhaps I should have listened more intently to the technical details – maybe it would have made me a better IT and phone guy today.
May 2008
Femtocells: Designed to Give Cellular Coverage a Boost
One of the next big areas in the world of mobile phone usage – at least in residential and small office situations – is going to be handy little devices known as femtocells. While they probably won’t have a huge impact on the college and university scene, there is a chance that femtocells will show up there and become part of the campus network.
Femtocells are essentially very compact radio base stations, and they are designed to supplement a service provider’s cellular coverage, particularly in areas where coverage is spotty. If you are one of those people who find your cellphone signal dropping to unacceptable levels at your home, forcing you to stand by a window or even go outside to use it effectively, you know it can be frustrating. The same goes for people who live near sports venues, shopping malls, or places where large numbers of people can load up the cellular network at times and make a connection difficult.
Users in these cases have two choices. Either they can wait for their carrier to build another tower closer to them, or they can purchase a femtocell from the carrier, which in effect puts a “tower” right in their home.
The femtocell, with a range of about 250 feet, handles all the radio traffic for up to six phones – many more with an “enterprise” version of the device – and then connects to the user’s broadband network (DSL, cable modem, or Ethernet). Then the voice or data traffic from the cellphone is transported via the wired network into the carrier’s core network. It never goes out to a physical tower in what is known as the carrier’s “macro” network until the user gets out of range of the femtocell.
A company we have been working with, AirWalk Communications, makes femtocells, and expects this market to take off in the coming months. Poor coverage is one of the biggest causes of “churn” for cellular providers, so it’s in the providers’ interest to keep customers from defecting because of less-than-desirable coverage at home. Femtocells will cost a couple of hundred dollars or so from the carrier, with a small monthly usage fee. The units are programmed to accept only traffic from certain phone numbers, so an entire family, a group of roommates, or a couple of dorm rooms could use them.
Where it starts to affect you and your networks is when someone does buy a femtocell and attaches it to the campus data network. This could be students trying to share a single femtocell (and possibly get reduced per-minute rates from their carrier), or it could be someone in a distant campus building (like the cow barns at an ag college) trying to boost coverage.
This shouldn’t cause any network problems for you, because the amount of traffic generated from these femtocells isn’t likely to stress the network. If you can stand up to all those YouTube video downloads and file sharing, this won’t be anything to worry about at all. Still, it is always good to be familiar with what is being, and might be, connected to your network.
April 2008
We’ll ask again: Is this a phone you want on your network?
That’s the question we raised in this corner last August, not long after the iPhone stormed the market. At the time, it redefined “cool,” and frankly, it still may be redefining it. Not being particularly cool in the first place, I’m not really the one to judge that.
And like an Apple-shaped snowball rolling downhill, gaining momentum and heft at the same time, the iPhone continues to be a consumer favorite – as well as an issue when it comes to enterprise networks. The question in the headline may end up merely a moot point, since whether you want it there or not, the iPhone will be part of your network. Way back in August, our advice was to start planning how to handle the iPhone, and we hope you followed that, because the urgency is increasing.
To its credit, Apple understands the reluctance of the people who run the networks to open them up to devices that are so much beyond their control. Security and management are serious issues, and the latest moves by Apple reflect that.
Last month, Apple introduced a number of new iPhone features that address the needs of enterprise users. For one, iPhone users now can gain access to Microsoft Exchange ActiveSync, which will deliver secure e-mail, contacts, calendars, and global address lists. This also enables IT departments to establish password policies, set up virtual private network settings, and even wipe out data on any iPhones that are lost or stolen.
Also, Apple says the iPhone will soon support a secure Cisco VPN protocol designed to guarantee highly secure IP-based encryption. These changes would seem to directly address the most serious arguments that experts have directed against the iPhone in terms of its applicability and utility to enterprise networks.
The latest “upgrades” will be part of a software revision that also includes an iPhone software development kit, which will aid developers in creating new, and presumably enterprise-oriented, applications for the device.
So it looks like the incorporation (or infestation, depending on your point of view) of the iPhone into enterprise networks is an inevitability. The good news is that it now appears that the unit itself and the intelligence contained within it are moving in the right direction. So when it happens, at least you’ll end up connecting a device that’s considerably more friendly than it might have been at one time.
March 2008
Chargeback Systems Get a Mediocre Grade
At
the recent ACUTA Winter Seminars in sunny
The latest survey focused on the issue of chargebacks, and while the survey sample was relatively small, it still yielded some interesting findings.
For one, on average this survey suggests that you are only moderately satisfied with the systems that are currently in place at your schools to charge back the communications network costs to departments and to students. If you were grading those systems on the four point scale, the aggregate grade would be a 1.92. Less than impressive, to be sure.
The systems themselves are not youngsters, by any means. Forty percent of surveyed attendees said their system was five to10 years old, and 24 percent had systems more than 10 years old. They are about evenly split between having been built in-house or by an outside vendor. It is worth pointing out that despite the general lack of enthusiasm reflected in that aggregate 1.92 grade, fewer than half of the surveyed attendees felt that their system had any negative impact on their budgets.
One very interesting finding was that more than half of the survey respondents said their biggest chargeback problems were outside the systems themselves. That is, their headaches are caused by inadequate cooperation from departments and the school administration, and a lack of backing at the executive level.
However, the biggest single “pain point” for respondents was the performance of the systems when it came to allocating the costs of data usage. Some of that is no doubt related to the age of these systems, since the issue of tracking data usage, particularly in a converged network environment, is a relatively new concept compared with the ages of some of the chargeback systems.
One thing we always find in these surveys is that ACUTA members are generally optimists. A total of 84 percent of respondents said they are likely to make a change in their systems, and most of them predicted that change would happen within a year. The most desirable benefits of a new system, the respondents indicated, would be to provide more accurate cost-of-service data, allocate service costs more fairly, and enable better ongoing analysis of their cost environment.
February 2008
What Kind of Experience Does Your Website Provide?
Is your school’s website doing its job? Sure, it may be pretty, with lots of views of the campus in the fall and of your highly attractive students. It may even be easy to navigate and provide scads of informative content. But as Dr. Phil might ask, “How’s it workin’ for you?”
The topic came to mind as we were doing a one-day “media training” project for a company that tests the usability of websites. Media training, by the way, is instruction in ways to better communicate their messages to the news media, whether the medium in question is newspapers, trade magazines, bloggers, or TV.
Anyway, as the people from this company describe what their company does and how they do it, you can’t help but wonder how your own website would hold up to scrutiny. What they do is analyze a website from the experience of the “customer” – essentially any individual who comes to your website with a purpose. The all-important question becomes how well the website serves that purpose.
Whether a school or a company hires a third-party organization for a real in-depth analysis is not so much the point as the need to view your own website from the critical perspective of its users.
All of us visit websites for information, to shop, to buy, or just to browse. What’s your experience with most of them? If it’s anything like mine, the majority of the time you end up frustrated. In so many cases information is hard to find, the pages are a maze of visual elements, or the navigation appears to have been designed by visitors from another planet. I will say that in my own experience, most college and university websites are pretty well done. The worst seem to be those companies that don’t sell directly to consumers, yet consumers visit them to find out information about products or where to buy them.
But even if the obvious potential shortcomings of a website are addressed, there will always some level of disconnect between the design, content, and apparent usability of the site and the needs of its visitors. In your school’s role as provider of the site, it is a huge challenge to determine whether you are meeting the needs of all your audiences. Students, prospective students, faculty, staff, and parents are just some of the people that your site needs to serve.
For at least some of those audiences, usability is part of the equation, but the overriding issue is the quality of the “customer experience” that the site provides. Does the “average visitor” come away from your website feeling positive or negative about the site and therefore your school? For some of your audiences, that is clearly a very important bottom line, but one that is difficult to analyze thoroughly and objectively. That probably explains why a company like the one we worked with has succeeded as well as it has.
Too often, the “is my website doing the job?” query falls into the same category as the “do these pants make my butt look big?” question. Honest answers may be hard to come by, but you really do need to know.
January 2008
Play the Network Game and Track Mystery Signals
As if you didn’t have enough to worry about with your campus wireless networks, what with coverage gaps, capacity, security, logon issues. Now comes a new potential threat to the network’s performance – video game signals.
Yes, it actually happened at
Morrisville State College in
In places where the video game signal was particularly strong, the IT folks at the ACUTA member college had trouble getting their 2.4 GHz Bluetooth headsets to sync up with their phones. They believe their wireless network equipment is working around the interference issue effectively, but they were planning to run a more extensive test using a number of video game consoles in one particular dorm that has a large number of active notebook clients.
Matt Barber, Morrisville’s network administrator, described the situation to Network World’s John Cox. After suspecting radio frequency issues were at work, the IT staff did some spectrum analysis and found a signal that was difficult to identify because it was “jumping all over the spectrum band,” Barber told the magazine. When on a hunch Barber plugged in a video game console for a signal comparison, it matched the characteristics of the mystery signal.
Pending further testing, the culprit appears to be an embedded 2.4 GHz radio in the console that communicates with a handheld wireless controller for the video game. The offending signal appears even when the game console is powered off; just the fact that the console is plugged into an outlet causes the radio to send out a signal in search of its companion controllers.
This is one of those problems that you probably can’t do much about, except to be aware of it and watch for any problems of your own. Students, particularly of the male variety, aren’t about to give up their video games any more than they’ll give up their cellphone or iPod, so it isn’t something that will solve itself anytime soon. However, if the problem grows, and if enough complaints are lodged, it could push Microsoft to come up with a solution.
For now, it’s an anomaly. With the next generation of video games, who knows?
By the way, Happy New Year!