Monday, November 30, 2009

Category 6 and PoE

In reference to an earlier blog post, Why I passed on Category 6, I laid out some of the background for my opinions on Category 6 cabling. In short, it’s a great cable that never really had a home in the data networks.

I wouldn’t be addressing the topic again except the University of Florida (go gators!) is in the process of reviewing their Telecommunication Standards. I’m happy to say that I am a part of that process.

The most recent revision of the UF Telecommunication Standards was published in 2005. Many things have changed since in the past 4 years and the revision committee is in hot debate concerning a number of topics. Of critical interest is the decision concerning what cable is to be placed in new construction projects on the UF campus. The current revision of the UF Telco Standards require a plenum rated UTP cable of at least Category 5e classification.

The new version of the standard will relax the prohibition against using nonplenum cable. I’m happy to discuss the logic behind both decisions but the debate concerning the use of plenum v/s non-plenum cable has been simpler than the old argument.

While the standards currently require a minimum of Category 5e cable, most new buildings have been constructed with Category 6 cable serving as the UTP cable of choice. Beyond any technical discussions, we can easily trace the evolution of this defacto standard.
· Architecture and Engineering firms habitually install the most popular cable in new construction. With none of our new buildings being IT centered, the choice to specify Category 6 must seem simple
· Our own department of Facilities Planning has the responsibility for the construction of new buildings. While they are often looking for ways to spend money they often look to the UF stakeholders to point out ways that designs can be made to best fit UF.
· Our own IT representatives to Facilities Planning focus on issues different than cable selection. OSP issues and documentation take a front seat in any discussion concerning new construction.

So, regardless of our own local standard, all new buildings since 2005 have been installed with Category 6.

This leads us to the current discussion in which our standards committees are debating whether to increase our requirements for new network cabling to Category 6- and yes, it is a debate.

All of the older arguments have been brought out to justify the requirement for Category 6 as the mainstay of network cabling on this campus. They have been joined by a new one: PoE plus is going to require Category 6 in order to function at full capacity.

Most of the time in this industry, our principal form of education comes from our vendors. In order to sell us their products, they want to keep us informed about the latest advances in technology. In IT, this strategy works. There is always a need for the next best thing. In cabling, the world moves a bit slower but this marketing trend seems to hold. A number of cabling installations are sold on their ability to “future-proof” the building and prepare for the next application. The latest argument seems to stem from this kind of thinking.

I’ve heard this argument from a number of sales persons in the last few weeks. They have held seminars on the topic. A number of our UF staff are convinced that by clinging to older cabling standards we may be constraining our ability to serve our constituency.

But, where is the evidence? Various Google searches and requests for whitepapers have come back empty-handed. In fact, a search of online periodicals seems to imply that PoE plus is designed to work over Category 5 with Category 5e being more than capable of supporting the next standard: PoE ++?

In short, more smoke and mirrors. You need Category 6 because it is certified to run gigabit. You need Category 6 to run 1000BaseTX which is the way future electronics will be constructed. You need Category 6 to run PoE plus. All patently false.

I still hold that Category 6 is a great cable. It is a great cable without a purpose. It serves better for analog video but now more and more video is going digital.

There are times I feel trapped in the past. If any reader can tell me why Category 6 is the better cable, please pass on the information. With a good argument, I can be converted.

Until then, I will hold on to my outdated cabling. When Category 6A cabling comes down in price and 10GBaseT cards become standard in desktops I may change my tune.

Thursday, November 12, 2009

Why TIA/EIA 606-A is a Wonder

In May of 2002 the TIA published TIA/EIA-606-A, the administrative standard for telecommunications infrastructure. To say that I am a fan of this standard is a bit of an understatement. 606-A lays out a consistent method of identifying and labeling physical infrastructure components that can be applied to almost any communications plant, from a single home to a multinational campus. Without going into the details, the 606-A standard defines names by a constructing a string composed of individual chunks of data that build from the specific and gradually expound as far as necessary. The standard details a number of specific examples but the premise is simple. The entire standard builds from this elegant idea.

Again, I’m a fanboy. Committees of volunteers create standards. Multiple opinions combined with multiple points of views combined with multiple points of self-interest. Yet, 606-A comes together with a unified message and a clear set of criteria that is indicative of sole authorship. The standard accomplishes this by never defining the absolute requirements of the naming convention. 606-A communicates naming formats by using variables for each chunk of data that a name should hold. The use of variables sometimes leaves the reader thinking he’s back in high school algebra but the use of variables is what sets the standard apart: ease of adaptability.

I was lucky enough to get an advance copy of the administrative standard. I’m not even sure if I ever completely read the approved standard. I’ve worn out my advance copy many times over. My supervisor handed me an advance copy of the 606-A standard and the rest is history. I’m not sure if he ever read the advance copy.

At first the standard, like most, makes for a boring read. No one writes these documents in an attempt to get on the best seller list. Still, the more I read, the simpler my own professional problems seemed to appear.
• Identify the components you wish to administer
• Name those components
• Label those items with labels that communicate their names
• Write down any documentation you need and associate it with those names

Simple. These few simple principles apply to any form of administration. The rest of the standard deals with appropriate means of naming telecommunications infrastructure. There are some tips on what kind of information a user may want to document concerning their infrastructure but it observes that requested information may vary from customer to customer. This treatise on useful telecommunications information may be useful to the novice but can be skipped by the more experienced.

In short, if you haven’t read this standard, do it. TIA/EIA is working on the next iteration of this standard and by my advance reading, it has none of the advantages of its predecessor. The new 606 standard reads as if it were created by a committee. It allows a number of different options for data center management and in doing so ensure that there is no standard after all. In their defense, attempting to standardize the administration of a data center is as difficult as standardizing the physical layout of a data center.

For now, leave the 606-B until the committee can lock down an appropriate administrative standard. I don’t envy them the task. Reread your copy of the 606-A and make modifications as appropriate.