The problem: a label that made sense on the day you wrote it often means nothing to the next person, and sometimes nothing to you. I am Dwayne, retired from a long career in telecom and enterprise networking, and I want to compare notes on how people label patch panels and cable runs so the system survives staff changes and bad memory.
Here is the convention I settled on after pulling a lot of cable and cleaning up after other people's installations.
Label the thing, not the purpose
Purposes change. A port that feeds the front office printer today feeds a camera next year. If the label says "printer", it is wrong the moment someone moves it. So the physical label names the location, and the purpose lives in the documentation where it can be edited.
One ID, the same at both ends
Every run gets a single identifier, printed on both ends, and it encodes where the two ends are. Mine looks like this:
PP1-24 <-> WP-B07-2
PP1 = patch panel 1, in rack 1
24 = port 24 on that panel
WP = wall plate
B07 = room B07
2 = jack 2 on that plate
The rule is that someone holding either end can find the other end from the label alone, with no lookup needed. The lookup is for the purpose, the switch port, and the VLAN.
The record behind the label
The binder, or a spreadsheet if you prefer, has one row per run:
- Run ID, as printed on the label
- Both end locations
- Cable type and approximate length
- Switch and port it lands on
- VLAN or purpose, with the date last changed
- Test result, if you tested it
The date column matters more than it looks. A row with no date is a row you cannot trust.
Habits that keep it honest
- Label before you terminate. Nobody labels afterward.
- Print labels. My handwriting did not age well, and neither did ink on cable jackets.
- Put the label where it can be read without moving the cable, usually within a few inches of the connector.
- Patch cords get labels too, at both ends, naming the two ports they join.
- Any change to a run gets written down the same day. A binder that is two weeks behind is worse than none, because it is confidently wrong.
The cost is real: it adds time to every install, and a label printer and supplies are a small ongoing expense. I think it pays back the first time something goes down and you can trace a run in a minute instead of an hour. A layer 1 problem is usually the real culprit, and you cannot check the cable if you cannot find it.
Two questions for the group. Do you encode location in the ID, or use sequential numbers and keep everything in a document? And how do you keep the records current once the person who wrote them has moved on?
Report back.
