Industry luminaries tackled a range of subjects at a spirited panel moderated by CNS Editor Paul Barker. Topics ranged from 'hits and misses' to the fibre vs. copper debate.
January 1, 2008
BARKER: It has been a wild 10 years. Describe in your own words how the structured cabling industry has changed over the last decade?
FRANC: The way I see it is has been a period of huge growth culminating with IP technology everywhere. 10 years ago, we were worried about integrating not just IP and IS together, but telephony and data. Today, we’re talking about building automation, lighting, video, just about everything. Anything you can think of can be IP enabled. I think that’s had a huge impact on our industry.
SHARP: 10 years ago everyone was talking about wireless as a possible technology that was going to be deployed. Now it’s here and we are seeing not only the IP environment expanding, but also the wireless environment.
So many elements of information sources and information destinations are obviously finding that their connectivity strategy is there. But that does not make life easier. It just makes it more difficult. The expectations of the public are that much greater as well. The IT directors have now got facility management to deal with and if not, they should be. The idea that IP brings access to so many areas now just makes life more complex and perhaps more difficult.
KISH: The biggest change I’ve seen is in the networking demand. That relates directly to how cabling has evolved. I remember in 1995 we were talking about Category 5 cabling and people at that time said that’s at the limit of what copper can do.
Today, we’re talking 10 Gigabits. That’s 100 times the speed over a cabling system and it is still copper. But if you look at fiber, if we stayed with the fiber we had in 1995, it would only support 20 metres at 10 Gig. So fiber has evolved as well. We’ve got better grades of fiber.
STEVENSON: One of the biggest things that has changed in the last 10 years is the impact of standards on what people are doing in the market. I think in a way it has begun to delay innovation to a certain extent because vendors are not going to put out anything that’s proprietary. Customers are educated enough now that they’re not going to roll out and spend a lot of money on something that doesn’t meet a standard that they know is coming. Everybody is sitting on the standards.
BOYD: It’s more of a commodity today than it was 10 years ago. 10 years ago, when we were in the build cycle, we were primarily dealing with IT directors and you could talk about technology to them. Today, a lot of our business comes through construction and through engineering. In a way, I think that has dragged the margins in the business down for everyone.
WEEKES: I would echo that, Rick. I’ve been talking with a significant number of contractor/owners recently in support of the training school that we’ve got going, and the market is exceedingly competitive out there and that is changing a lot of the dynamics. It’s very hard to attract young people into the network cabling industry. And when they do come through our program, more times than not the first question they ask — how much money am I going to make coming out of the back end of this program?
BARKER: Let’s turn our attention to hits and misses, of which there have been many. What stands out in your mind?
HORNE: The VF-45 connector is the number one flop over the last 10 years as far as technology goes. I remember when it first launched and a colleague of mine in Ottawa was hired on to promote it and I said to him, “Boy, have you got a job ahead of you.” It was interesting to see a company like 3M who really was by far the number one leader in connector technology and how they put all their eggs in that basket and it was a real mess, unfortunately. It was interesting technology, but I think sometimes when you leap too far forward, when you push things too far and there isn’t enough acceptance in the market, it just isn’t going to work.
WEEKES: I was working at Panduit around the time the VF-45 was launched and here’s exactly the note I wrote: “The failure of the industry to select a small form factor fiber connector (is) creating confusion in the marketplace. And I went to those standards meetings and I came away very disappointed because as an industry we said, “OK, let the strongest guy win.” That’s the person who has the most clout in the marketplace. Because we all had our instructions — parochial instructions — as to how we were going to vote. I’ve seen a lot of my clients put in MT-RJs because that’s the way the Cisco ports went initially on a lot of the switch equipment. And then LC comes along and you adjust your pricing accordingly as to which product you want to sell, and now we have to go back to the customers and say “You know what, we’re going to have to slowly take off all those MT-RJs because all of the new electronic equipment now is coming with LCs.”
And so, to a certain degree, because we didn’t make a decision at those meetings, we created some confusion with our customers.
KOSTASH: One of the more notable misses in terms of what the networking industry was going to do, how it was going to evolve and how that would impact structured cabling was the whole question of asynchronous transfer mode, or ATM. We certainly pushed a product in the mid-1990s as being capable of supporting the 622 Mbit/s ATM, and our friends at Belden have products that are named after ATM speeds, but clearly IP has been the winner and a lot of the marketing hype around ATM never materialized.
KISH: I think the market wanted a technology such as Ethernet that could evolve. The key point here being that it’s backwards compatible so that the next generation, can still communicate with the previous approach. That’s a key factor.
FRANC: In terms of misses, there were a lot of prognosticators that brought out their crystal balls 10 years ago. Copper had a limit. Fiber is limitless. Well now the fiber that we installed 10 years ago has a limit. ATM is the future. FTDI is the future. I find it funny looking back 10 years ago. All the crystal ball-type predictions failed and today we have people out there saying the exact same thing: “You will never need this” or “You will always need that.” I think we should have learned from our past and never say never.
KOSTASH: The only one that has held true has been Moore’s Law.
HORNE: Business drives the market ultimately. We could come up with new products, but the bottom line for manufacturers is leveraging the installed base. If they can get products that will work over older technology, bump up their speeds, push the throughput on existing infrastructure, they will do that because there’s a huge installed base out there. Let’s not forget the VHS-Beta lesson from years ago. You can have great products and great technology, but if the market does not accept it or if it’s not cheap enough or acceptable enough for their needs, they won’t pick it up. The limits of copper came out of the needs of the manufacturers to push those existing limits because they knew that manufacturers of equipment are going to put stuff on copper before fiber. I think that leads to a push forward.
BOYD: 10 year ago with the whole wireless debate, the conversation was we aren’t going to need wires by 2008. Wireless is all over the place, but it’s not replacing the cabling, thank God, and adding a lot more services and connectivity that customers need.
BARKER: Data networking has evolved substantially from the network engineers’ perspective. In your opinion, what major changes have you seen in the past 10 years and what do you feel the future looks like in the next decade?
MASTERSON: There has been a major change for the network engineer out there in how they design their networks so the design of the structured cabling plant has to keep up.
We started seeing it on unusual items such as medical devices and HVAC devices and building controls and door entry systems. In the past five years we’ve found ourselves in the middle of all sorts of interesting discussions.
The expectation when you develop a building now is that it’s going to come turnkey with a network and the network will be both wired and a wireless that’s going to carry protocols that will support in some cases life-critical applications related to health care.
We were involved in the launch of the William Osler Hospital in Brampton, Ont. where they’re putting everything over that network from the door entry controls to the life support systems for the patients.
From a networking perspective, it probably simplified the wiring a little bit because you used to have a wire for every application and every protocol. And as those protocols made their way onto IP and also through encapsulation techniques with the multi-protocol label switching or MPLS where you can put non-IP protocols over Ethernet through MPLS, you could still carry even some of ATM through MPLS over IP.
It’s an interesting approach and there’s a way you can do it. You then raise the bar for resilience and security, because you’ve got all sorts of stuff going over one wire.
FRANC: The network engineer and the network administrator have to become integrated with the construction process. When projects fall down, it’s usually due to a lack of communication between the facilities people or the applications people and making sure that the planning is done right at the outset.
I remember being involved with the new airport terminal at Pearson International.
The network administrators had to deal with things that they’ve never dealt with before — baggage handling systems, signage, flow control and access control. How do you handle all those things? How do you plan for them?
KISH: A couple of points that were triggered by what Brantz said. The network will slow down, we all realize that, but then there are a lot more applications running on the network. The thing that you’ll find though is that a couple of parameters are going to be very important: quality of service, and having the bandwidth that you need to be able to support all those applications. And some applications are not that tolerant for bit errors because you get dropped frames and poor quality of service.
SHARP: The concept of convergence is greatly misunderstood. I think the idea that convergence should necessarily mean that all your data traffic, no matter what the source, should travel down the same conductor is misplaced although it’s a very popular notion. Many get the idea that we can compromise or the building manager or building owner can economize by putting in less.
They’ve got to transfer those capital dollars into operating dollars and try to maintain an IT department that can understand the difference in quality of service, the difference in product, and now you’ve transferred something that was static and deterministic into something that’s volatile and not deterministic. And I think that’s going to be a future nightmare for property managers.
You go through the commoditization process, you can get IP-enabled or Web-ready enabled devices to be extremely cheap, very low cost and that goes for all of the facilities that go into a traditional building today.
The ability to manage that information and manage that technology without the trades or the skill set that the various departments responsible for running the operation or a business, have fallen far behind the abilities of the people at this table to deliver a product.
KISH: Do you advocate redundancy to some extent?
SHARP: You need parallel redundancy, absolutely. You’ve got to have strategic redundancy as well. I don’t think it’s sufficient to say that this fiber can carry all the data traffic in the world so it’s going to be sufficient for me. But on the other hand, now I’ve got to administer the priorities of that traffic in order to make the applications function properly. I just don’t think the skill sets are a) there and b) I don’t think the people who are managing the buildings and are managing the IT departments are aware of the fact that there are these differences.
And that doesn’t even address the issue of at what point do you put a network into a building when you’re constructing a new one? That’s a serious problem in the construction industry. It has not caught up, and never will in my opinion, with the technologies that are available to go into the network.
MYERS: I’ve seen this very symptom with a number of customers who have moved in one application that we haven’t talked about yet, which is voice.
The rigor at which you treat a voice application and the need of a phone is different from the rigor that you say, give to a printer on the floor. And so, if the IT department wins the battle for voice, it needs to enhance their rigor around the voice application set. Because the phone people would never have anybody be without a phone set working on their desk for more than a few minutes or an hour, but a printer could be down for a few days.
HORNE: Peter’s point is well taken. In construction, we’re all dealing with property managers and building operators. Historically, the systems involve base building systems that run a building, and then there is the network that the tenants within that building operate. Will the two overlap? I think we’re seeing convergence in this. Specific companies are taking advantage of these new Web-enabled or IP-enabled devices and integrating a lot of the building core systems into a unified system.
They’re still proprietary in many ways, and that’s the nature of the beast. The manufacturers want it proprietary so that they can keep people locked into their products, but we’re seeing this opening up now into multi-vendor situations. I think building systems are going to naturally converge into a system that will manage multiple building systems and devices.
MYERS: The real business returns don’t come from putting these applications on a single wire. In fact, if you want to put them on separate wires, that’s fine.
For example, if you’ve got a customer information database that is tied to the CRM system, you now want to tie that to the phone that’s on their desk so that you can do automated dialing or if someone calls you can bring that record up and interact much better and elevate the level of competitiveness through customer service. That happens when you put those things on the same network. That’s the business value. It’s not the low cost of one wire versus multiple wires.
MASTERSON: The network, if you look at the evolution, really was a way for people back years ago to get their printing done and basically just a method to do some work. Today, the network is integral to the business and business goals now have to be aligned with network management or the goals of the network. If the goals of the network and the people driving the networks do not align with the goals of the business model, it will create problems.
SIEMON: When it comes to quality of service, the past 10 years have focused on bit error rate. In the years ahead the trend toward convergence and highly interactive applications will change the way we look at network performance.
Our industry is always looking for the next new thing and frequently refers to disruptive technologies as “killer apps” — innovations that have an impact on our daily lives by improving the way we access and interact
with people, places and things.
Killer apps also drive revenue growth in nearly all sectors of the IT industry. In the future, I wonder if the focus will shift from killer apps to app killers.
By app killer, I mean the type of change that forces us to rethink fundamental aspects of network performance. Will trends in network utilization, operating cost or environmental considerations take us in new directions that bring latency and power consumption to the front of the line as the primary drivers for future networking technologies? Over the next decade we will either see Ethernet adapt or new networking protocols will emerge that more effectively address these needs.
STEVENSON: From a network engineer perspective there are a lot of other things that now impact the decisions apart from the applications and equipment you need. Now you’ve got power and HVAC considerations that can very critically impact what you’re doing. We’ve got people who are going in and saying “Oh great, I can fit all these blade servers in a rack,” and they’re ordering them and bringing it in to a computer room, and then somebody comes in and says “Why is it 90 degrees in here? What are we doing about this?” These are the types of things that network engineers have never really thought about to the degree they must do now.
SHARP: I think that the absence of skills and capability is never more obvious than in an organization’s ability, or lack thereof, to manage their data centres. I find it absolutely staggering that people will willingly turn a blind eye to many of the issues, which are impacting the most important thing, which is the information that they’ve amassed. And the ignorance that they are inflicting on their organization and putting all of their systems at a high level of risk is absolutely astonishing. I think part of that is because we’ve got this traditional divide between facility personnel and the users.
WEEKES: The data centre manager who looks after these network engineers, so to speak, hasn’t come up through the budgeting wars through the years and does not necessarily know how to play that sort of game within the organization. With a few exceptions, unless I get to the president of some reasonable-sized organization to have that little discussion, nobody seems to take that up to the level where it needs to be, where you have that what-if. It’s a very simple discussion: “What if your data centre goes off-line for two days? You’re the president.” We just saw a CitiBank executive lose his job because of some sub-prime loans somewhere else that he could probably do nothing about, but if somebody had said that to him a couple of years ago — “What If” that may not have happened.
FORTUNE: During a previous computer buiding project I worked on, there were constant conflicts about who actually owned it. Did facilities own it? Did IT own it? And when it got right down to it, it was still a struggle right to the end, when you finish building the thing.
IT had to install tons of equipment and they had to install it fast and they’re just shoving it in. Meanwhile, the operations side was trying to take care of all the added power and cooling requirements.
It’s an interesting issue to be involved in, but when you get back to what we talked about earlier, if the governing standards boards couldn’t make decisions on what type of connector to use, how do you really expect these two sides to come together with all the issues they have to deal with and have a common front going in one direction? It’s almost impossible.
BARKER: Will the fiber vs. copper debate still be around in, say, 2018?
KISH: I definitely think so. Before we see 10 Gig to the desktop, we’re going to have to wait five years, but I certainly could see widespread use of 10 Gig to the desktop in 2013. At the same time you’ll find fiber’s has its place too and it’s going to have to run at 100 Gig on the backbone. There’s a need for copper and what’s driving the need for copper to the work area is the manufacturers of computing equipment
FRANC: Media choices are going to get harder especially now that we have copper, fiber and wireless. Copper and fiber are great if you have a fixed data point, but what if you have mobile requirements? People aren’t satisfied with only accessing their information when they’re at a desk or when they’re at a fixed terminal. How many people around this room have Blackberries? We want our data all the time, every time, and we always want more data, so I think the media choices are going to get far more complex.
HORNE: The economics of the market will always drive copper to the desk. As long as those media are there to support it, it will be there. But it is going to be interesting to watch what wireless does. Can they push the speeds up on wireless? Can they address the security concerns and issues that haunt wireless?
KISH: VoIP is a natural application for wireless because it does not use that much bandwidth, but you can very quickly choke up a wireless system by putting in high-definition video or high-resolution video. And if its users are accessing that same access point, it’s going to choke up because it doesn’t have sufficient bandwidth. So I think it’s a great technology for people who need the portability, but it doesn’t have sufficient bandwidth for the really intensive applications. That’s why I think we still need wires.
SIEMON: I agree. For the debate to end there must be a decisive winner. 2018 is not that far away. Over the next 10 years fiber will continue as the media of choice in the core of the network – especially for link segments greater than 100m. Copper will be the preferred choice for the last 100 meters.
MASTERSON: I think at some point copper and fiber always have to be there. But who thought they were going to get 10 Gigabits over copper? What is the limitation of wireless? Do we really know right now? Do we know what the future holds? I don’t know if we do. If somebody told me 10 years ago that they’re going to run 10 Gigabit over copper, that’s going to be on a twisted pair to my PC or my server, I would have told them they were crazy. So how do we predict what’s going to happen? It’s very difficult.
BARKER: How has the role of the cabling installer evolved since 1998?
SIEMON: The guys that got their training with lunch and a couple slices of pizza aren’t good enough anymore. When Category 6 came out, they discovered they couldn’t do it the same way they did DIW (D-Inside Wire) back in the AT&T days. These guys require skill in design and implementation. They now have better capabilities and tools with which they can provide a more reliable, higher performance cabling infrastructure than ever before.
MASTERSON: The termination methods and skill level required today are quite high. We found that out quickly, based on the questions we would get: “Why is this tester failing the cable?” as opposed to “Is the cable failing or am I installing it wrong?” By the fact that the testers are out there and able to give installers some feedback and automatically give them fault info in terms of where the problem is, allows the quality of workmanship to go up, I believe. So that’s been a big change, the quality of workmanship required.
SIEMON: And it’s based on the instant feedback they get from testers. The end user will not let them leave with a failing result.
STEVENSON: The expectations of the installers are a lot higher because in general, the end users and the customers are a lot more educated than they used to be and they know what they’re expecting to get. You can’t get away with doing installations and not testing them. So you’re going to have to do those test results. You’re going to have to show that your work is performing the way it’s supposed to be. And
you’re going to have to be able to answer a lot of questions.
BOYD: From a contractor’s perspective, it’s become easier to train people to do the work because pretty much all we do now is UTP. 10, 15 years ago, we had guys that had to know Wang, Twinax and IBM Token Ring, Type 1 and Type 2, and Ethernet. From the perspective of training the technician, I think it’s a lot easier to do. Also, the technicians have evolved with the training that comes from manufacturers and organizations such as BICSI.
Still, there are still a lot of people out there doing poor quality work. It’s not core cabling professionals, but more the phone installations firms who have guys who will run cable for them, but do not use support structures. There’s still a lot of that going on, unfortunately and typically it’s driven by the dollar.
When a customer’s putting in a new telephone system, and their phone guy says “I can put in a Cat 5E for you for 80 bucks.” Cool. We come in and say it is $120 because we’re putting in J-hooks or velcro tie-wrapping and we’re doing all these things you should have for cable management. It’s hard to explain to them.
FRANC: My concern with installation isn’t so much the pre-installation as the post-installation. We made everything plug-and-play with RJ-45s and LCs and different types of connectors. Installers do a fantastic job installing it.
But if you look at a lot of the installations afterwards, you’ve got this incredible asset you paid good money for, but it is managed poorly.
If you look at most data centres, you’ll see the best of times and the worst of times. Somebody’s mission-critical data centre is an institution where the thing falls apart, they lose their data. And just because it’s an RJ45 or it’s an LC connector, somebody will plug something into a cabinet, come out the door, go four cabinets over and plug it back in, which means there are three cabinets in between you can’t access now, because somebody’s gone and patched something.
I saw that in a Fortune 100 company’s data centre, and unfortunately, the link that was connected in that way was their main WAN connection to the outside world. So in order to fix the problem, you have to disconnect it from the outside world.
I think the role of the cabling installer’s has to change. It’s not just about the installation, but managing the asset afterwards.
FORTUNE: A lot of times that’s not the cable installer, it’s the local IT guy in the company or it’s the guy that the company pays a little bit of money to come in and do all their menial patching.
It’s part of the engineering design. It’s documented and probably never ever done. It’s the patch cable management piece. When I was at the bank, it was our biggest problem too. You did all the cabling infrastructure, designed it all properly and people would always come back and blame us for all this horrible patching. We’d say “Come here,” and I would take them around to the back of the cabinet and say, “Now see that, that’s what we do.” We provide the patch cords or the contractor provides the patch cords of a certain length and then whoever would come in and just plug in whatever he wanted to.
If he had a 30-foot patch cord to do a three-foot patch cord job, he’d use a 30-foot patch cord.
HORNE: That’s more of an operational issue, probably a whole other set of discussions. I definitely think that the role of the installer has been marginalized and they have become a commodity.
Unfortunately, we see huge demand for qualified people, and a lot of people have moved on to other trades because they just can’t get paid accordingly. I think that’s where it sits right now. You’re going to see this work just being naturally picked up by construction/electrical world as far as the installation. Now we are talking a lot about maintenance. It’s not being viewed as a qualified trade, a wholly functional, recognized, respected trade, which is unfortunate.
Maybe Bill has something to say about that?
WEEKES: We’re trying to train. Our ability to attract young adults into our particular programs is terrible. We have people lined up out the door on waiting lists for the electrical program. We have people lined up for the plumbing program, and in the HVAC program. The two programs that are taking it on the chin are network cabling and the alarm security. The kids aren’t interested. And it’s further compounded by the fact if you go on to Workopolis and Monster, you’ll see installer ads where companies are looking for people with three years experience, know Cat5, the 6, 6a, fiber terminations and most start at between $18-20 an hour. They gag when they look at that. And entry level positions in our industry, through nobody’s fault — it’s all through market pressures — are $12-14 an hour.
STEVENSON: Maybe that is the biggest way that the role has changed. It’s been marginalized. There are lots of people available, it’s all about the market’s ability to pay them what you need to pay them. All these companies are telling Bill they can’t find people. They can’t find people who want to do the work for what they feel they can afford to pay them.
BARKER: Why is that?
WEEKES: Well, the Two Bobs In The Truck destroyed the whole bottom end of the market. To Rick’s point exactly, if there’s a guy going in selling telephone systems and he says “I can put in those 50 drops and he doesn’t even look at the project, it’s automatically $80 a drop? Because he doesn’t have an employee, he’s then going out to somebody else who he’s paying a flat rate, right? Rogers and Bell got into that mess years ago. Everybody’s under contract to Rogers. Everybody’s under contract to Bell.
FORTUNE: But we don’t do that with electricians. We had to unionize non-union electricians, so why isn’t there something put in place for cabling installers?
WEEKES: I want to stay far away from the politics. We moved into these self-regulating industries. You could make your arguments about the Law Society of Upper Canada, how effective they are at regulating lawyers, and you could make examples of the medical association of Ontario, and how effective they are at kicking out bad doctors. And the real estate agents and everybody now kind of belongs to this self-policing, self-regulating organization of some sort.
I’m a certified engineering technologist, whoopdee-do, but the point is there is some legislation that backs that up a little bit and I can have my CET revoked.
Electricians got a lot more discipline-oriented this year, but this industry is the complete opposite. Rightfully or wrongfully, we said “we don’t want regulation.” We don’t really want the government overlooking what we’re doing. We don’t want to go out and get a permit if we need to install 20 or 30 drops. So if we don’t want those things, how do you change the perception on the part of the end users about what a data cabling person is really worth?
BOYD: I’m generalizing but, with installers, it’s a blue-collar job. You’re not going to get a guy who could be a lawyer or a doctor wanting to go into that job. We target people with a high school education, who don’t want to go away to university, don’t want to go to college for whatever reason, and they don’t know what they want to do.
And my sales pitch to them is it’s one of the few industries where you can start there and end up as a salesperson and make six figures. The biggest problem that I see is kids coming out of high school today have expectations that they’re going to be driving Porsches within two years because that’s what their parents do.
They’re not willing to put in two or three years to learn the trade. I’ve got installers that I pay over 60 grand a year. I mean that’s a decent living for a blue-collar guy who’s got no education other than high school — in my opinion.
I know I might get shot down politically, bu
t that’s fine.
BARKER: I would be remiss, if I did not delve into the shielded vs. unshielded cabling debate. How do you think this will unfold?
KISH: I guess this whole unshielded and shielded debate has come up right now because of 10 Gig and 10GBASE-T. We have strong proponents on both sides of the issue. We have some manufacturers that are advocating using shielded cabling and other manufacturers like ourselves that are actually saying, “UTP cabling supports 10 Gig and does an excellent job.” When you look at IEEE the 802.3an standard what it actually says is “to support 10 Gig, here are the recognized media” and one of the recognized media is Category 6a UTP. And they also recognize Category 6 and Category 6a shielded, which is called screen twisted pair.
So they will both work and I think the debate should probably end there, but it doesn’t. We have a lot of manufacturers that are going and basically telling customers “Oh, this is a better product. You should really be putting in shielded.” But there are issues with shielded cabling that are really have not been fully addressed.
I mean, shielded cabling is widely used in Europe and they’ve established a way of terminating it, and how to ground their building and facilities properly. In North America it’s relatively new, it’s like less than 2% of the total cabling infrastructure. And when we talk to installers, there are not that many who are adequately trained on how to terminate the shielding.
From a performance point of view, for the shield to work effectively at high frequencies, and even that’s controversial these days, is that you need to have it well grounded at both ends.
You also have to pay attention to the grounding issue but also there’s another important parameter that’s important for both UTP and STP, and that is pair balance, and how the actual balance of the pair effects how much is actually transferred or converted to noise that the receiver sees. And today’s cables, especially Category 6a UTP, have excellent pair balance characteristics, so that’s a positive thing. And I think down the road what we’ll see is that’s going to be the important parameter that’s going to dictate how one cabling system performs better than another. But today, we’re just getting into this area.
SIEMON: If the question is which one will win — shielded versus unshielded, I think shielded cabling will win. It’s just a matter of time. That said, we have been making products for twisted pair cabling for over 100 years. Since we are a family business, UTP put me through school. It put my father through school and his father through school. We recognize its importance to our business and to our industry and will continue to support it as part of our core product offering.
The reason I feel that shielded systems will eventually displace UTP is that transmission theory favors shielded cabling in terms of channel capacity. UTP will support 10GBASE-T, but screened 6A cabling is a better choice for 10G because of its inherently better immunity to external noise, including alien crosstalk.
Pair balance is equally important for screened and unscreened systems. In the case of individually shielded pairs, it’s possible to hide the sins of poor pair construction with a shield. I believe that balance for both screened and unscreened systems should be specified in industry standards. Unfortunately, doing so is challenging because balance measurements produce different results depending on the presence of a shield. Since the shield acts as the reference plane in one case, but not the other test results can be quite different even when intrinsic pair balance is the same. Standards bodies need to recognize that screened and unscreened systems are not the same and to develop specifications that reflect their respective capabilities. When this is done, the technical merits of screened systems will become much more apparent than they are today.
Also, more education is required regarding proper screen termination and grounding. Incidentally, they are not the same thing. The screen is not a ground conductor. It is a high frequency transmission element of the cabling system, which requires the same level of care in handling and termination as the pairs themselves. Optimally, the screen is bonded to ground in the telecommunications room only, and screen connections that meet standard requirements for high frequency performance are maintained up to and including connections to active equipment.
STEVENSON: One of the big drivers is going to be cost. The cost comparisons we have done to date is that we can’t do shielded for the cost of unshielded. Unless something happens to eliminate that difference, I don’t see what the driver’s going to be for people to go to shielded.
HORNE: It’s definitely out there. We had a client in Montreal, a law firm that was taking over six floors of a tower, and they were considering a 15-year lease. They were looking at all the options. One was definitely shielded. They seriously considered it.
Ultimately, it went unshielded because of the way that they wanted to distribute their infrastructure, which was with a zone-type distribution system, which didn’t favour shielded cabling. When it came to one Gig, the clear winner was unshielded, but definitely I think 10 Gig and beyond it is going to be a tight race.
FORTUNE: It’s not just the cost of material and the cost to terminate the material. The cable’s also smaller, so pathways are smaller. You can potentially look at cost savings there to offset the price of installation.
KOSTASH: I think really there’s a question too of the timing of shielded twisted pair cables in the marketplace. The product’s only 2% market share right now as Paul Kish referred to earlier. In the timeframe that we’re looking into going into 10 Gig to 40 Gig to 100 Gig, and the timing of the conversation we had earlier, about when is fiber potentially going to take over from copper, I see a fairly narrow window there for sure for shielded twisted pair cabling to grow into significant share in the marketplace. I think that fiber’s coming faster than shielded twisted pair.
The cost of copper rod went up fivefold in three years. That type of thing can have a dramatic effect on how quickly people start to move towards fiber optic infrastructures, right? If cable costs are going up by a factor of 15-20% because of the commodity material used to make copper cabling, that could potentially be a good thing for fiber’s entry into the market.
BARKER: It seems like the world is turning greener with each passing day. Does this unstoppable movement have the potential to radically change the structured cabling and networking industries and if so, how?
KOSTASH: One of the things we have to recognize is that structured cabling by its nature is green. It can eliminate multiple different systems or has the potential to increasingly eliminate multiple different systems proprietary systems. The whole idea behind structured cabling was one platform that could run multiple applications and evolving applications over time. Perhaps inadvertently our industry has stumbled across a product or a design philosophy that is inherently green. But there is still a lot of work to do in cable construction and so on to improve the cabling and manufacturing processes as well.
HORNE: I honestly find that a bit of a stretch. I understand and I agree with the overall concept of it, but I think cabling as green is a big misnomer We deal a lot with the building owners and property managers on this side of the world and they’re running a number of different programs, trying to green their buildings.
I know our government’s driving a lot of new building construction and they seem to be building to meet a certain minimum of green standards.
And the big dirty secret behind cabling is that it’s made with a lot of toxic c
hemicals and where do these cables go when they’re abandoned? Well often times they end up going on a ship to China where they get burned into the atmosphere along with the copper. I would like to see a lot more pressure on manufacturers to come up with other materials and recycling programs for their cables.
KOSTASH: I don’t disagree in any way. All I was saying was there has to be a way of dealing with multiple proprietary cabling systems backing up over years in the cable tray, and a structured approach was, and remains a better design.
MYERS: What’s on the end of the cable is possibly a bigger problem. You’ve got no lead in the cable but you’ve got lead in the computing systems that are sitting in the data centre, so I think the good news here is that the move toward virtualization actually helps a lot of the green buildings because it’s more economical to put more applications on these very powerful one-use servers.
At the same time, they’ll extend their useful life hopefully because you’ll be using a server for more applications than maybe one that only runs for a couple of years and then you have to move on.
So you can load a server up to its capacity, push that computing load around using modern virtualization tools to get the most out of that asset and grow the computing volume as you need, maybe on month-ends and quarter-ends. And then ultimately the life of those assets means hopefully fewer systems put in the trash or recycling systems that we have today.
WEEKES: I spoke about LEED and the RCDD at the BICSI Fall Conference in Las Vegas, Nev. in mid-September and I was shocked by the number of questions that I received after the presentation.
I was also overwhelmed by the number of people who came up to me after the presentation to talk about a significant number of ideas and also what they think about our industry.
If you have experience working with LEED buildings or the United States Green Building Council and how they put their requirements together on a new building, it really asks you to think a lot differently about a lot of different things.
On the one hand, we can talk about all the positives that our technology is creating, and I’m right there with you, and we need to spend more time and energy communicating that out to the world.
On the other hand, the Toronto Star ran an article recently that said 10 to 12 people a week die in the coalmines in China. And that’s acceptable in China because they desperately need that coal to support their generating stations, which support a lot of the products that we buy.
What I tried to do at the presentation was to say that “here are the rules that architects have to follow. These are the rules that electrical engineers and mechanical people have to follow on LEED buildings.”
Lucky for us, telecom is completely off the map when it comes to LEED, but that doesn’t give us an excuse not to follow the initiatives that they’re trying to set. And so after I gave that presentation, a guy from one contracting firm came up and and said, “Oh, I’m so glad you talked about that. My wife makes me bring all the boxes, all the cable boxes back, and we fold them down, we put them in the green bin.” And I go “That’s wonderful.” I mean that’s a start. But that’s how deeply ingrained this is in our new society that we live in.
What are we doing to make things a lot greener? My presentation was really, unfortunately short on answers. It was about awareness and going back to BICSI and saying “Look, guys, I need your help!” I’m getting the questions. I’m working on a building for Department of National Defense. They’re going for LEED gold, now they’re thinking about going for platinum. And they’re coming back to me and they’re going “Bill, you gotta help us get some innovation going.” And I’m going “OK.” To get innovation going, I have to go back to my manufacturers who I’m working with and say, “What are you doing to help me to get something so that they can get some innovation going on this project? What have we done?”
Now we follow certain steps on this particular project and we are going to apply, I hope, they’ll get the one or two innovations going they need to push from gold into platinum. But it is an issue for a lot of different people as to “How do we make these cables?” “Where are the cables being made?” “What does it mean when it’s made in a jurisdiction where we don’t know about the environmental regulations of that other jurisdiction?” Pete, you must have some experience with that as well.
SHARP: We are engaged in designing a building in which the owner has asked for LEED status. And the contribution that cabling can make is so minimal that it’s just not worth bothering. The only component that would even have a bearing on the catchment of that building is what you do with the bits that you cut off. How do you dispose of the waste products during construction? And if you’ve got a documentable, auditable process whereby you can dispose of it in a green sense, then you contribute to a larger pool that will give you one point.
So all of the waste disposal thing, if the data cabling installer were to simply cut the stuff off and burn it on the street, it really wouldn’t affect the LEED status of that building.
I think, in honesty, the real contribution to greening is to not slash and burn the installation every five years. And that has a plus and a minus side.
BARKER: True or False: Structured cabling installed today will last for at least 20 years?
SHARP: Everything that’s being installed today will be useful in 20 years.
FORTUNE: One problem is churn rate, whether they’re reconfiguring all locations and therefore having to rip out the wiring.
SHARP: It’s not the churn rate it’s the attitude towards the churn. I would argue that if the designer’s done the job properly and if the owner has allowed the designer to do the job properly, the very fact that it is, by definition structured cabling means that it’s going to support multiple protocols.
Therefore it is going to be useful for generations to come. So therefore it’s going to be useful for 20 years. The question is is the political will there? Is the economic will there? Is the organizational business will there in order to allow the designer to do the job according to LEED or green philosophy? Invariably, every owner believes that their applications are unique, special, and require that kind of approach to the cabling.
It’s a big challenge and I don’t think I’m the only one who experiences this. The issue really is that you know instinctively what the right approach is in order to economize on the materials that are going to a) be installed and b) that are going to be thrown into the landfill in five or 10 or 20 years. Very seldom are you allowed to exercise that aspect of your judgment in the design process because it invariably does not produce the lowest cost design.
FORTUNE: The reality is that things happen in those locations. They have 90 square feet per staff and then they turn around, they want to squeeze in more people, they get it down to 70, or down to 60. They restructure their organization because they’ve got 20% spare space on these floors so they then have to redevelop that whole floor space, re-cable it.
SHARP: As an example, take a building where you’ve got a 30,000 square foot floor of space. You’ve got a common core, which carries the elevators, the staircases, the washrooms, electrical and communications room. If you were single tenants on this floor, it’s a no-brainer to run all the cables back to that central room and then administer from there.
But if you’re going to divide the tenancy of this floor, which in this particular building could easily be divided into 10 tenants a floor, how are you going to cable each one of those and make use of that one centre core room? First of all, that core room bel
ongs to the landlord and you’ve got to petition rights to use it and your rent is going to go up based upon the benefits that you accrue.
If you’ve got two or more tenants in a common space, they’re each going to cable according to their own home run criteria, and that home run is going to be somewhere on the tenant floor.
So all of a sudden, you may have a structured cabling scheme, but because it got different cabling points, focus points, it does not survive a different tenancy. If you have another tenant that comes in and puts up a wall or takes down a wall, all of a sudden your structured wiring scheme, although it’s structured in principle, doesn’t apply to that application.
HORNE: I agree absolutely. I think it gets into dealing a lot with building owners, property managers. We strongly advise them not to allow anybody to put any equipment in any of the corridors. In fact, when we’re building new buildings, we tell their techs not to build any communication rooms.
We advise them to build a riser room only for the passenger cable, and all commercially built by the tenant, in their space to their specs. Whether it’s a single tenant or not, I consider the whole issue of common space and electrical a common area. But I think that one of the things that we’re talking about is life cycle. And it gets into it but a single-tenant/multi-tenant issue which is common to… the building industry itself. When managing a building, they life cycle everything. There’s a life cycle on the windows, on the roof, on the HVAC system. There should be a life cycle on the cabling infrastructure based on when you’re putting it in, how long you expect it to last, and if you budget and plan accordingly for the replacement of it … IT departments don’t do that. It’s all business-driven. “I’ve got a pile of money,” or “We’re upgrading applications, oh we have to upgrade the system or we should upgrade. Let’s just rip it out, put in something new. But if you treat it more like an asset that should be life-cycled like a building, then you can plan for an expected life span and plan for a replacement in 15 years.
If something comes along in the interim that forces you to change that, well then you deal with that at the time, but otherwise, life-cycle it, plan for the replacement, upgrading in a certain period of time and budget for it. IT departments aren’t or have not been historically good at doing that, whereas the building industry is very good at life-cycling and managing the replacement of something over an expected period of time.
BARKER: Time to delve into standards. Does the current process work or should it be changed?
FRANC: I think it works fine. I hear the frustration. I’ve been an end user, a consultant, a contractor, now working for a vendor. Being involved with standards gives you a different perspective.
You realize that a lot of people provide their input. Look at the discussion we just had, I think we’re all in agreement that there’s no right answer. The problem is everybody’s looking for a magic bullet. They want a standard. They want a document. They want a page number they can point to and a checklist that will take away all their problems. I don’t think the world is that simple.
I don’t think our industry is that simple. So I think the way we’re moving towards consensus, the fact that we can have a 10 Gig cabling draft standard closely following within a year or so of an electronic standard, I mean that’s wonderful. I think people are frustrated, perhaps because there’s not enough involvement in the industry within standards.
SHARP: The only flaw in the standards process is that the users aren’t involved enough. And I think the problem is that involving yourself in the standards process is an extremely costly process. If you’re selling widgets, then it makes sense for you to standardize on widgets. But if you’re using widgets, chances are you’re only using one or two of them. And so what’s the benefit for you to engage in that?
I know, my experience in the standardization process was rich and fulfilling because I got a great deal of knowledge and connections, and acquaintance with some of the ebb and flow in the thinking and the tendencies within the industry, which I think provides me the mechanism of providing a better service to my clients when I have to try and look in the crystal ball and see where the industry is going to go in two years.
Also, I think there’s a big misunderstanding and that is if you follow the standards then you’ve done it right. Well no, if you haven’t followed the standards, you’ve done it wrong, but the corollary is not necessarily the case. I think you’ve got to use the standards as an intelligent guide to the manner in which you’re using these particular commodity products and the application for which they were intended. And I think you have to understand the philosophy behind the standard to be able to successfully use it
HORNE: They do a lot of good, but you have to recognize they’re not written by end users per se. They’re written by manufacturers to promote their products.
FRANC: If you look at the international bodies, it’s not populated by vendors. It’s one vote per country. Within TIA, you have one vote per organization. And there are a significant number of end-user consultant organizations south of the border involved in TIA.
In fact I think you’ll find that the huge issue we have in Canada is a lack of awareness and a parti- cipation issue.
KISH: I’ve been in TIA since 1989 and although there is a large percentage of vendors it has to be a balanced organization, so we do get participation from end users and consultants. But the manufacturers have a large share of what goes into it because a lot of the detailed specifications that go into it, I would s say 80% of the document, deals with requirements on the cable and the components.
SIEMON: I’ve seen the level of complacency in standards work especially with respect to IT infrastructure, which I believe relates to an imbalance in representation between cabling system manufacturers and their customers. The question is how do you develop a meaningful standard that provides lasting solutions without having customer’s needs represented as an integral part of the approval process?
Complacency is also evident in that successive generations of cabling standards provide diminishing returns in terms of delivered system performance. The industry should take a hard look at the level of value provided by the incremental solutions, from 5E to 6 to 6A.
For future standards, the goal should be to provide significant, measurable improvements that address practical problems associated with the installation environments like data centers and enable newly installed cabling to support multiple generations of applications.
Category 6a is lagging behind 10GBASE-T. Standards for cabling infrastructure need to get out in front of the applications and user needs. To attain balance, increased representation from IT equipment developers and IT professionals that represent the enterprise community is required. Having this input will help to ensure that future cabling standards address installation issues such as the congestion in cabling pathways while minimizing complexity in active requirement, which drives up power and latency.
SHARP: One of the plusses in the process that we have today, which should never be removed is democracy. Ironically, that also is its underbelly in that sometimes individual initiatives can be hijacked by those who would prefer not to see the initiative have the light of day in the standardization process, the debates that occur. There are a number of examples that I won’t quote that are pretty obvious where organizations, manufacturers or user groups just simply either want to promote or hijack the development of a particular form of technology. And it’s very easy to do tha
t in a democratic environment.
FRANC: That’s a good point, Peter. If I as an individual wanted to participate in ISO, could I? No. I’m not a country. Could I participate in BICSI standards? Theoretically, yes, after submitting an application and having it reviewed by a separate committee, yeah. To Peter’s point, the TIA standards, although we’re always frustrated by them, I think we’re all agreeing to it — they’re probably the most open. If I want to represent myself as Henry Franc, I can vote on any document. Paul Barker can vote on any document. The fact that there’s a heavy weighting of manufacturers at this point in time speaks to, not nefarious purpose as a vendor, but to be perfectly honest, the lack of commitment or willpower from the other interested parties in the industry to become part of the process.
I spoke to Paul Kish well before we worked in the same company, I worked for an end user, and I had the same frustrations everybody else had, but I wanted to become involved. What do I have to do? As to Bill’s point, Paul said “Well, the next meeting’s in Montreal. Go there.”
I know there’s that perception out there, but it’s not a vendors club. Anybody who wants to can participate.
BOYD: I agree with everything people are saying, but I do think the standards process is the single biggest factor in commoditizing our industry. Because now, people just say “I want Cat 6. Give me Cat 6.” Well, yeah. 10 years ago, you’d be in there trying to tell Lucent PowerSum because nobody heard of it before and it was sexy and it was cool and you’d get people to buy into it. Now it’s Cat 6.
SIEMON: Several issues back, CNS magazine touched on the point that product claims and product performance are two very different things. In other words, labels often lie. While I agree that standards accelerate the move toward commodity status, the bigger problem is that there is no meaningful oversight to prevent products from being falsely labeled as standard compliant. Products made by companies that are not actively involved in industry standards and trade organizations pose a higher risk. The only effective countermeasure today is to educate the market on risks associated with adopting a price-driven strategy for cabling infrastructure products and installation.
WEEKES: I have often wondered why the TIA meetings and standards meetings never coincided with a BICSI meeting. You know, I always had to make two separate trips. You go to Orlando for BICSI, and sure enough, six, nine month, a year and a half later you’re going back down to Orlando for a standards meeting. I would think it would be absolutely fantastic if the TIA had a standards meeting at the same time BICSI was on and, to Robert’s point, all of a sudden, a thousand, 2,500 people at a BICSI conference get sucked in to this little vortex of standards and they’re actually made a little more aware of the process and what it is all about.
BARKER: Is anyone in this room concerned about the pending IT skills shortage and if not, why?
MYERS: It’s a vast problem. The skills required today in small medium enterprise are very much under-served. There’s a long list of specialty skills, a growing list of specialty skills. Each one of these network conduits touches so many points about applications and security and privacy and fire and long wide area and local area and wireless.
The list is massive and growing and we just aren’t seeing enough people coming out of school and embracing these skills. It’s not an opportunity they want to pursue.
It’s a supply and demand problem right now.
BOYD: Everybody was going through the Y2K thing. IT people were paid exorbitant amounts of money and all of a sudden that went away and you lost the market. That, in my mind, is what killed the apprenticeship program in Ontario. We were all busy as all get-out, we needed guys like crazy. We had this thing really cooking along. Then all of a sudden 2002 came along, we had all these people in school, and everybody said: “We’re actually laying off. We don’t need people right now.” And we lost people out of our industry, because for a two-year period we didn’t have the jobs available.
We lost good installers, we lost people coming into the industry, because we just didn’t need them. So you lost maybe a generation of university students who looked at the market and said “these people are overworked and underpaid. I’m going to do something else.”
MYERS: If you look at the computing science programs today, they’re producing people that are coming out with great skills around kernel operating system development and compiler optimization skills. There are a handful of people in the whole wide world who do that work for a living. They need to adjust their curriculum from a pretty esoteric PhD-level skill set to one that I believe is far more practical and ubiquitous.
BARKER: Final question: Do you feel that trade organizations such as BICSI are serving the best interests of both their members and the IT industry? How can they be improved? I cannot speak for the U.S. but certainly in this country they need to pump up the membership. I mean, for an organization of that size to only have 811 members in a country of 33 million is, to me, a little astounding.
BOYD: They put on a hell of a golf week every January. Thank you BICSI.
HORNE: I think overall they do an excellent job. I think people get out of BICSI what they put into BICSI. We complain about it, but the positive ones are going to say they aren’t involved in it. I don’t think I’ve ever gone to a conference I didn’t get something good out of.
That being said, we need to be constantly questioning why they’re there and what they’re doing for us as members. We shouldn’t rest on our laurels and we shouldn’t be too complacent.
FRANC: I think that’s part of the benefit and part of the challenge of BICSI. To Rick’s point, about the golf week, you know, people make fun of the golf week. One of the most important things that BICSI does is give people a forum to interact with their peers, not just in a professional capacity but an interpersonal capacity. They have mandates varying from education and certification to writing standards.
They’re trying to be a marketing organization. They’re a lobbyist organization. They’re providing guidance not just on design, but installation and best practices. They’re trying to do a multitude of things and I think it’s because, as Rob says, because it’s a member-driven organization, different member profiles within BICSI have different needs, so it takes on multiple different roles.
Whether that’s the right thing or the wrong thing, that’s up to the membership to decide, but it’s their great benefit and probably their greatest liability as well. It’s really going to dictate how things move over the next 10 years.
SHARP: The unfortunate thing is that yet again, BICSI is a business and they can only survive as a business. And unfortunately that has its down side. I remember in Chicago maybe a year and a half ago, the conference was a completely different setup. For three days, there were workshops. And the workshops were intense. They were excellently delivered.
The attendance was fantastic and when I asked BICSI would they be continuing this, they said that this was such a resounding success, we will never do it again. Because the vendors who were on display there had absolutely nobody attending the displays. They were all at the workshops and from a BICSI point of view, it didn’t make sense. As an attendee, it was fantastic. Unfortunately, BICSI has to survive as a business.
SIEMON: I think that’s the primary challenge. BICSI is serving two different constituencies. One group designs, installs and maintains working IT installations. The other group is the supply base for the first. Ye
s they serve the same industry, but as Peter has pointed out, the best interests of both groups do not always intersect. The key to future success will depend on their ability to clearly identify the differences and address them in a way that balances the interests of IT industry suppliers and those that design, install and use those products and systems.
No other organization in our industry has demonstrated such a strong commitment to continuing education. As long its members and leadership continue to recognize the value of this role and to support ongoing training and education, BICSI will continue to play a central role in the growth and stability of our industry.
KOSTASH: Since BICSI is thriving they are obviously doing something right, but as John mentioned in trying to address two constituents and bring them closer together, the challenge to remain relevant is ongoing.
SHARP: Seeing the products is very important as well. Maybe if they took the workshops idea and then had days that were absolutely dedicated to just products, then maybe that would be a good balance. It must be a very challenging exercise to try and satisfy all the disparate groups that belong to BICSI.
BARKER: That’s a wrap. Thank you very much and I hope you enjoyed it as much as I did.