When OSI-model layers collide
Get used to boxmakers having something to say about what you're putting in for cabling
Get used to boxmakers having something to say about what you're putting in for cabling.
While making stops at cabling-industry-related booths during an enterprise-networking trade show, I was practically accosted by a sales professional from a cable manufacturer. She wanted something done about the fact that several manufacturers of networking hardware (I'll call them boxmakers) have told her clients that Category 5e cabling is a sufficient infrastructure for their needs.
Sometime after the boxmaker leaves, the cable seller walks through the door and tries to make a case for a cabling system that exceeds the Category 5e specifications. "Not interested," she is told. And frequently, she reported to me, the conversation doesn't really get off the ground after that. The customer doesn't want to hear anything about enhanced this or headroom that. The boxmaker got there first, convinced the client, and the decision was made.
The cable seller wanted to know what I, as chief editor of a cabling-industry publication, planned to do to help her. I thought about it, and came to the conclusion that it is not my responsibility to make life easier for people who sell cable.
It is my responsibility, however, to try to make life easier for people who install and manage cabling systems. So, to that end, here is my take on the "you-don't-need-better-than-Cat 5e" argument.
The boxmakers have a compelling argument. After all, the Category 5e specifications were developed to provide a solid infrastructure for Gigabit Ethernet transmission. Therefore, anybody buying gigabit boxes should be able to rely on Category 5e. Not Category 5e plus bells. Not Category 5e plus whistles. Just Category 5e. And I know I'm preaching to the choir here, but it takes some in-depth knowledge of cabling-system nuances to recognize and appreciate the advantages that come with a better-than-standard system-a depth of knowledge that not everyone possesses.
Besides, I find it difficult to have sympathy for a vendor that:
- Participates in the standards-making process;
- Is among the TIA group that, through consensus, approves certain performance specifications as satisfactory for a cabling system;
- And then bemoans the fact that users take the standards-making group at its word-that a system meeting the standard is a sufficient infrastructure.
But putting aside that pet peeve of mine, the issue of networking-hardware manufacturers influencing cabling projects is very real. As more users entertain the idea of gigabit-speed systems, they run into the reality that a 100-Mbit to 1,000-Mbit migration comes with more, and more difficult, bumps in the road than they encountered when migrating from 10-Mbit to 100-Mbit. The advances that were made in transceiver technology to achieve gigabit speeds bordered on profound, and the boxmakers' products naturally have gotten more complex.
It was the IEEE's intention to construct the 1000Base-T specifications so that the protocol could run over the installed base of Category 5 cabling. But after field tests indicated that prospect was a crapshoot at best, the IEEE and TIA worked closely together to define just what level of performance a twisted-pair cabling system needed in order to support 1000Base-T. The result, of course, was Category 5e. This history gives even more credibility to the boxmakers' claims, but I believe it also signifies the end of the separation between the first and second layers of the famous OSI model.
What does that mean to you as a professional? For those of you who manage cabling plants and enterprise networks, get used to boxmakers having something to say about what you're putting in for cabling. In fact, don't be shocked if they want to tell you about the newest division in their company: the infrastructure-installation division. For those of you who install cabling systems, see note to cabling-plant managers above.
Patrick McLaughlinis Chief Editor of Cabling Installation & Maninenance.