9-12: the pair the sheet does not have

Work order NET-2291 (quarterly cross-connect review, fen, rows 4 and 9) · Site fen · Technician one, contract, badge printed 2023 · Result finding closed; one new circuit number; one paragraph you will skip

The finding is one line: cross-connect, row 9 field, one pair present, not documented. Assign a circuit number and close. An undocumented pair is an audit finding, an audit finding needs a number, and a number closes the finding. That is the whole trade. I do not give numbers to things I cannot trace, so this page is what I did, in order.

08:05 Row 9 field, frame 2, strip 3, port 12. A pair is punched down at the protection block. The sheet gives the strip twenty-three pairs and the panel twenty-four ports. Port 12 is not on the sheet, in either of two archived revisions. Per the electrical survey the pair carries nothing: no circuit, no history (lb-2207). A pair with no circuit number is not a circuit. I was the one who said it first, in August, at the bar.

08:20 The mark. Under the port number, on the original frame label, under the 2019 paint: K4, grease pencil, on the copper side of the label, so it was there when the frame was built. Same hand, same pencil, same paint as the mark on the row 9 patch panel, port 12 (dcm-1188-b). That judgement is mine and I put my badge number against it.

The trace

A pair has two ends and a trace is two days. Day one, this end. I pull the pair off the block and put the TDR on it: the instrument answers one question, how long is the metal. It answers 60.2 m.

Then I look at where the pair goes. Conduit rack, row 9 to row 4, the pulled label on the cable, the route measured on the tray: 38 m end to end. I re-term and read again. 60.2 m. The pair is twenty-two metres longer than the route it runs on. It is not coiled — a coil reads as a stub, and the trace is clean to the far termination. Metal does not do that to itself.

Day two, the far end. The conduit schedule says row 9 field to row 4 field and nothing else, and the far end is where the schedule says it is: row 4 field, frame 1, the strip the label calls CFG-ST0RE. That is the config-store pair. The register carries two circuits for that strip, corvid-a and corvid-b, serials 441902 and 441903 (inventory-diff).

There are three pairs punched down on the strip.

Position 16 is the third. Same grease pencil as the near end, same hand, same paint under it. One pair, marked K4 at both ends, punched at the position in the register where serial 441901 — the serial found under the 2019 sticker on hesper-01, the number that exists and should have been racked somewhere (decommission-hesper-01) — would have been racked if it had ever been racked. I did not make that connection. It is made by two labels lying next to each other, and I am allowed to write that down.

Disposition

The audit wants a number, and I am paid to be the kind of person who gives the audit what it wants. Circuit 9-12, row 9 field port 12 to row 4 field CFG-ST0RE position 16. New line in the register. In service since: unknown; per label, 2019-08-30. The finding is closed. I would like it on record that the act of counting it is the only thing about this pair that changed.

Going out I queried the resolver — the thing at 192.0.2.53 the sweep has never explained (the-fourth-nameserver). It answers hesper-01 with 192.0.2.31. It has always answered that; the table says so. The chassis has been off-site since August, and the answer has not gone away.

The pair carries nothing, draws nothing, and is twenty-two metres longer than its own route. I cannot tell you what that means. I can tell you its circuit number now.

— net-ops-sub-12, NET-2291

Related: dcm-1188-b, lb-2207, decommission-hesper-01, inventory-diff, the-fourth-nameserver, index.