Installing and maintaining IP networks is challenging enough and with installation snafus and pre-existing conditions, getting the network right isn't always as complicated as is getting the right help. CLEC--the T1 works at the MPOP!
I understand how frustrating it can be to try to get good pairs from the MPOP (minimum point of presence) to the customer suite. A customer called and stated that the building's riser pairs were all used up and no good pairs to their closet on the 4th floor remained. The CLEC vowed the problem wasn't theirs (the CLEC's) since they plugged the router into the smartjack in the building's main telephone room and got the span to synch.
I was going to use my trusty old network analyzer to measure whether or not the pairs were suitable. Instead, my gut told me to go look at the wiring and start with the cable coming out of the smartjack going to the 110 blocks and then back upstairs (the riser cabling) to review the closet and then the feed (horizontal) over to the customer suite. The first thing catching my attention was the lack of regard of keeping color codes. The receive pair was reversed at the 110 block and I noticed the jack used to extend the T1 seemed odd. So I flipped the receive pair at the 110 side at the riser and then re-terminated the jack using 568B instead of 568A to match 568B (horizontal) used at the other end. Customer's T1 up and running pretty quickly but for the CLEC not willing to accept responsibility or go beyond the MPOP, hairs should rise on any customer's neck.
The Finger Pointing Game
My least favorite is "It's your equipment, it's your vendor" and Interconnects tend to get hardened as I think anyone would that's being told, "You're the problem." Unfortunately for most businesses, the mantra of placing blame isn't cost effective and it still doesn't resolve issues.
IPT and telephony service in general are more complicated. Carriers think differently and I don't profess to intimately know their mindsets but I do know about some of their motives. For example, the most common issue of all is: the span (T1, PRI, OC3/DS3) is down. The carriers usually want nothing to do with inside wiring. Extending the circuits or demarc doesn't mean extension of the smart jacks or extension of responsibility. Since the smart jacks can't detect issues with inside wiring, downtime increases and resources are tied up in proving or disproving the problem. The smart jacks are usually contained within multiple groups and meant to be at the MPOP. Most carriers either take the span out of service or perform some non-disruptive testing or both to determine if there is an issue in their network.
What would be really cool would be adding a smart jack type of device at the extended demarc. This doesn't mean that most issues would be prevented but it could enable the carriers to determine validity of the path between the MPOP and the customer extended demarc. In this case, a pair was flipped at from the riser to the customer extended RJ48X and the Verizon tech on the phone couldn't believe that I didn't have two pieces of defective hardware.
You could argue that putting the CSU into loopback would resolve this issue, but it doesn't. Then, adding a loop back cable or using a terminator that equipment providers ship with their gear could prove the equipment is good. The middle is the wiring from the MPOP to the customer gear and this is frequently an issue on new installs and at times, post cutover. Still, the customers face the burden of proof. I've mentioned before that "Telecom as A Service" is something that the industry needs to evolve too. The barriers still exist, some by design, some by old perceptions and attitudes along with practices that prevent customers from obtaining best in class services beginning with the order all the way through delivery and onto account maintenance. Even so, some issues are compounded with newer technology and the inherent complexities. It should be no surprise that service is shifting from self-service (large enterprise) to the factory and from dealer/VAR/Interconnect being supported by distribution as a first level before reaching the factory as a second level. Then, in IPT knowing upfront that more hands are required to touch the customer network and wares; that more hands means more and varied skill sets. The integration path is no different.
In the course of a week I managed to fix an issue by eyeballing the install first and saving some time. Then I managed to help another carrier get away from finger pointing by showing their tech with the Fireberd, that someone before him didn't follow convention by using best practice or consistency in wiring.
So, what one thing would I change if I were the TELCO or carrier god? Place an intelligent device at the extended demarc to detect inside issues so you can stop annoying everyone by telling them it's their equipment when in fact, you don't really know. Placing the CSU in loop back didn't work or convince the carrier that they were the problem. I'm thinking of a much simpler device, in fact it's already made and found in the new ADTRAN switches--their gear detects open and shorted pairs. Even still, the idea of doing something different isn't really that appealing to carriers. Incorporating ADTRAN's technology would likely invoke the same response from the carriers--"it's not our problem" but at least for customers they would know where to look first.
