devices(ubiquiti): the Enterprise PoE pair sourced 802.3af watts on 802.3at ports #2388

Merged
jaylouisw merged 1 commit from catalogue/1121-0722 into main 2026-11-21 07:42:37 +00:00
Owner

Refs #130 only - the UDW-PRO request is still blocked (re-measured this run: techspecs.ui.com/unifi/cloud-gateways/udw-pro -> 307, 1 byte; store.ui.com/us/en/products/udw-pro -> 404). This PR does not answer it.

Two switches that source power on every copper port shipped no poeOut on any of them, so the engine's derivePortHardware() fallback made each port an 802.3af PSE at 15.4 W:

model vendor says catalogue said now
USW-Enterprise-8-PoE 8x RJ45, PoE/PoE+, max 32 W/port by PSE, 120 W budget 8 ports, no PoE direction -> 15.4 W 8 x poeOut: 32, poeOutClass: "PoE+"
USW-Enterprise-24-PoE 24x RJ45 (12x 1G + 12x 2.5G), PoE/PoE+, max 34.2 W/port by PSE, 400 W budget 24 ports, no PoE direction -> 15.4 W 24 x poeOut: 34.2, poeOutClass: "PoE+"

Both rows now also declare poe-pse, 802.3af, 802.3at, and each copper interface declares connector: "rj45" instead of leaving it to be inferred from the name. The SFP+ cages are untouched and stay non-PSE.

Why it shows: web/faceplate.js prints PoE out <poeOut> W in the port tooltip and web/app.js prints PSE, <poeOut> W out in the detail panel, both from the derived hardware. On a box the datasheet rates 802.3at, the socket reported 15.4 W - the one number an operator reads when asking whether it can power a 25.5 W AP.

Sources (all read 2026-11-21, every field traces to one of these):

New regression test test/usw-enterprise-poe-census.test.js (26 assertions) pins the port counts, the per-port PSE watts, the class word, the AC power method (no poeIn) and that the engine's derived hardware reports 32/34.2, not 15.4.

Gates: syntax-check 517 OK; audit-port-types 0 gated contradictions; audit-devices 0 hard defects (2,352 models / 29,768 ports); audit-poe-budget 0 orphans; hardware-invariants 26/0; check-control-bytes clean; check-doc-citations clean; docs/CONNECTORS.md regenerated, not hand-edited (974 -> 976 port shapes, because two rows that drew unnamed copper ports now draw declared RJ45).

Refs #130 only - the UDW-PRO request is still blocked (re-measured this run: `techspecs.ui.com/unifi/cloud-gateways/udw-pro` -> 307, 1 byte; `store.ui.com/us/en/products/udw-pro` -> 404). This PR does not answer it. Two switches that source power on every copper port shipped no `poeOut` on any of them, so the engine's `derivePortHardware()` fallback made each port an **802.3af** PSE at 15.4 W: | model | vendor says | catalogue said | now | |---|---|---|---| | USW-Enterprise-8-PoE | 8x RJ45, `PoE/PoE+`, max **32 W**/port by PSE, 120 W budget | 8 ports, no PoE direction -> 15.4 W | 8 x `poeOut: 32`, `poeOutClass: "PoE+"` | | USW-Enterprise-24-PoE | 24x RJ45 (12x 1G + 12x 2.5G), `PoE/PoE+`, max **34.2 W**/port by PSE, 400 W budget | 24 ports, no PoE direction -> 15.4 W | 24 x `poeOut: 34.2`, `poeOutClass: "PoE+"` | Both rows now also declare `poe-pse`, `802.3af`, `802.3at`, and each copper interface declares `connector: "rj45"` instead of leaving it to be inferred from the name. The SFP+ cages are untouched and stay non-PSE. Why it shows: `web/faceplate.js` prints `PoE out <poeOut> W` in the port tooltip and `web/app.js` prints `PSE, <poeOut> W out` in the detail panel, both from the derived hardware. On a box the datasheet rates 802.3at, the socket reported 15.4 W - the one number an operator reads when asking whether it can power a 25.5 W AP. Sources (all read 2026-11-21, every field traces to one of these): - https://dl.ui.com/ds/usw-enterprise-8-poe_ds.pdf (200, application/pdf, 526,714 B, 2 pp): "Networking interface (8) 1/2.5 GbE RJ45", "SFP+ interface (2) 1/10 Gbps SFP+", "PoE interface (8) PoE/PoE+ (Pins 1, 2+; 3, 6-)", "Max. PoE+ wattage per port by PSE 32W", "Max PoE power 120W" - https://dl.ui.com/ds/usw-enterprise-24-poe_ds.pdf (200, application/pdf, 397,353 B, 2 pp): "(12) 10/100/1000 MbE RJ45 / (12) 100/1000/2500 MbE RJ45 / (2) 1/10G SFP+", "PoE interface (24) PoE/PoE+", "Max. PoE+ wattage per port by PSE 34.2W", "PoE budget 400W" - https://techspecs.ui.com/unifi/switching/usw-enterprise-8-poe (200, 648,955 B): "2.5 GbE RJ45 = 8 (All PoE+) (2.5G/1G/100/10M)", "PoE+ = 8", "Total PoE Availability = 120W" - https://techspecs.ui.com/unifi/switching/usw-enterprise-24-poe (200, 650,004 B): "1 GbE RJ45 = 12 (All PoE+)", "2.5 GbE RJ45 = 12 (All PoE+)", "PoE+ = 24", "Total PoE Availability = 400W" New regression test `test/usw-enterprise-poe-census.test.js` (26 assertions) pins the port counts, the per-port PSE watts, the class word, the AC power method (no `poeIn`) and that the engine's derived hardware reports 32/34.2, not 15.4. Gates: `syntax-check` 517 OK; `audit-port-types` 0 gated contradictions; `audit-devices` 0 hard defects (2,352 models / 29,768 ports); `audit-poe-budget` 0 orphans; `hardware-invariants` 26/0; `check-control-bytes` clean; `check-doc-citations` clean; `docs/CONNECTORS.md` regenerated, not hand-edited (974 -> 976 port shapes, because two rows that drew unnamed copper ports now draw declared RJ45).
devices(ubiquiti): the Enterprise PoE pair sourced 802.3af watts on 802.3at ports (refs #130)
All checks were successful
CI / Classify the change (pull_request) Successful in 14s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 16s
CI / Test (Node 22) (pull_request) Successful in 1m2s
CI / Test (Node 18) (pull_request) Successful in 1m3s
CI / Test (Node 20) (pull_request) Successful in 1m4s
CI / Server smoke and Docker build (pull_request) Successful in 26s
CI / Arm auto-merge (pull_request) Successful in 4s
d92f117935
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-21 07:33:07 +00:00
Sign in to join this conversation.
No reviewers
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
jaylouisw/packetforge!2388
No description provided.