Introduction
Both tests are valid and both produce a "PASS". But they do not measure the same thing, and if you accept the wrong one as your basis for handover you may take delivery of infrastructure that has no real margin.
What Permanent Link measures
The fixed, permanent part of the infrastructure: from the patch-panel port to the wall outlet, including the horizontal cable and both terminations. Patch cords are not included. This is exactly what the contractor installed and hands over, which is why it is the correct basis for handover.
What Channel measures
The entire real path, including the patch cord at the rack and the one at the user end. This test shows the true state of the service and is useful for fault-finding, but patch cords are replaceable and are not part of the contractor’s permanent delivery.
Why the limits differ
Because the length and the number of connections differ. A channel has more components, so its insertion-loss limit is more relaxed. The important consequence: a link that passes as a Channel may fail as a Permanent Link. If a contractor only gives you Channel results, you are seeing a more flattering picture of the infrastructure than it deserves.
The number to look at
Do not just look for the word PASS. Look at the margin on each parameter. A link that passed with 0.3 dB of margin can become a real problem with a little heat or a poor patch cord. A healthy margin is one that will still be there in a few years.
What to ask the contractor for
Permanent Link results for every port, in the tester’s native file — not just a printed summary. The port numbers in the report must match the labels on the patch panel exactly, or the report is effectively unusable.