Field Technicians experience some pretty interesting scenarios from time to time. “Found In the Field” highlights some of these real-life issues and how they were (or weren’t) resolved. Names, dates and locations have been omitted to protect the guilty!
“I found myself in the middle of a situation between an end-user and a service provider. The end-user was a financial institution who was taking errors like crazy over their 100Gb circuit. The service provider tested the network and the results showed everything was running clean. The customer, however, argued that was impossible. Both the customer and service provider reached out to me as a third-party, unbiased tester.
First I sat down with the service provider who showed me everything they had done as far as testing the lines. Using an EXFO OTDR, I then tested from end-to-end in the core, which was the provider’s part of the circuit, and it ran perfectly clean. Then I tested from the customer’s collocation point to collocation point, from cage to cage, and sure enough the errors appeared. So we knew the problem was between the customer’s router and where they interfaced into the service provider’s circuit.
As the next logical step, I started testing fibers individually. My OTDR was fitted with iOLM software. I tested again, this time using the iOLM. Sure enough there was a discrepancy. The OTDR showed only one connector in the cage but the iOLM map showed there were two and that the path was going through a patch-panel the customer was completely unaware of. What I discovered was that the ORL reading on one fiber was -33dB and the other fiber between the transceiver and service provider read an ORL of -40dB. Now the specifications for minimum ORL allowed are based on the transceiver’s manufacturer which in this case was -27dB, so we shouldn’t have been taking errors since both numbers were better than what the manufacturer called for. But what really stood out to me though, was that the two fibers shouldn’t have read different ORL’s. They were set along the same path, through the same amount of connectors, the same number of hops, so they shouldn’t be that different, certainly not a difference of -7dB.
I cleaned the connectors, tested again and the ORL deviation between the fibers disappeared. They now both read -40dB. The customer had spent six to eight months trying to figure out why these errors were being reported. Countless time and money was spent on man-hours tackling this issue. I found the issue, cleaned the connectors and had it running smooth in forty-five minutes. No errors were ever reported again. The sole issue was a dirty connector. That’s it.
Needless to say, the service provider was happy as this proved that they were in the clear. The customer, also quite relieved, then bought an iOLM kit for every site they had.
So I suppose you could say EXFO was happy too, heh.”