Applies to: Negotiated copper LAN rate at one link. Internet performance, Wi-Fi and server/storage bottlenecks are separate observations.
- Signal path
- Computer / USB Ethernet adapter → Patch cords + installed channel → Switch port / injector data interfaces
What the path requires
- Record the exact current link rate and both device/adapter models. Do not force a new speed or duplex setting on a managed connection.
Worked documentary examples
Worked documentary path
Documented outcome: Example with declared observations: a gigabit-capable NIC and switch report 100Mb/s through wall patching but the expected rate with a suitable direct cord. The next request is a channel report with tester/limit identity, not a claim that a lit 1-to-8 tester certifies gigabit.
Alternative or unknown branch
Documented outcome: If the adapter model is unknown or the far-end port cannot be read, keep the expected-rate premise unresolved and obtain those two capability statements.
Follow the observation
Read the result before moving to the next branch. Leave unavailable evidence unknown.
1. A path component is limited to the observed rate.
Check: Compare the documented maximum and enabled modes for the NIC/dongle, switch port and any injector.
What it supports: The cable category does not upgrade that component.
Next action: Choose a documented faster path if needed; record the limit rather than replacing intact cable on category alone.
Stop / escalation: Stop for damage or policy restrictions. No forced timing, unsupported duplex changes, powered continuity test or driver/firmware cure.
2. One side is forced or the settings disagree.
Check: Compare negotiated speed/duplex on both sides with their configured settings.
What it supports: A negotiation/configuration mismatch is a plausible cause.
Next action: Have the administrator apply the exact platform's guidance to both ends; do not force only one side.
Stop / escalation: Stop for damage or policy restrictions. No forced timing, unsupported duplex changes, powered continuity test or driver/firmware cure.
3. The requested rate appears directly but drops on the installed path.
Check: Use a suitable known-working short direct cord on the same compatible ports when permitted.
What it supports: The removed path needs channel investigation; a connector may pass basic continuity yet fail performance.
Next action: Record the whole channel and request wiremap plus an appropriate application qualification/certification test.
Stop / escalation: Stop for damage or policy restrictions. No forced timing, unsupported duplex changes, powered continuity test or driver/firmware cure.
4. Only continuity or simple pin-order pass is available.
Check: Review an existing test report.
What it supports: It does not establish required pair integrity or target bandwidth.
Next action: Identify tester model, test limit and channel/permanent-link scope; use the cable-test lesson before interpreting Pass.
Stop / escalation: Stop for damage or policy restrictions. No forced timing, unsupported duplex changes, powered continuity test or driver/firmware cure.
5. LAN negotiates the expected rate but Internet/download speed is low.
Check: Separate LAN status from actual transfer.
What it supports: The physical link observation alone does not identify the throughput bottleneck.
Next action: Preserve the negotiation result; investigate upstream service, workload and endpoint constraints with an authorized network procedure.
Stop / escalation: Stop for damage or policy restrictions. No forced timing, unsupported duplex changes, powered continuity test or driver/firmware cure.
Your next useful observation
Preserve the negotiation result; investigate upstream service, workload and endpoint constraints with an authorized network procedure.
Keep these limits
- Stop for damage or policy restrictions. No forced timing, unsupported duplex changes, powered continuity test or driver/firmware cure.
- Source-documented interpretation; example observations are hypothetical, not measurements.
Source documents, revisions and inspection methods
Configure and Verify Ethernet 10/100/1000Mb Half/Full Duplex Auto-Negotiation
Cisco · Document 10561, updated 2025-11-06
When to Use 10/100 and 1000Mb Auto-Negotiation; Appendix D interface fields
Inspected 2026-10-23 by public_page. AI primary-source text inspection; not independent human review or hardware measurement
Observation/negotiation interpretation for documented Cisco context. No live configuration, reset or erase instructed.
Read the original sourceVerification Qualification and Certification
Fluke Networks · Article dated 2018-09-14
Three kinds of test tools; closer look at verification/qualification/certification
Inspected 2026-10-23 by public_page. AI primary-source text inspection; not independent human review or hardware measurement
Test-method distinctions only; no test performed in this implementation.
Read the original sourceNetwork Engineers — cable selection and testing guide
Fluke Networks · Page revision not stated
High Performance Access Network Cabling table and channel footnote; fiber cleaning/testing sections
Inspected 2026-10-23 by public_page. AI primary-source text inspection; not independent human review or hardware measurement
Compact selected category/channel rows and test limits. No claim that a jacket mark guarantees installed performance.
Read the original sourceSource-documented education. Hardware behavior and independent human review have not been performed for these examples.