devices(ubiquiti): an ES-10X that sources nothing, and three 24 V passive ports filed as 802.3af #2290

Merged
jaylouisw merged 1 commit from catalogue/1120-0952 into main 2026-11-20 10:09:25 +00:00
Owner

Refs #130 (reference only; the UDW-PRO blocker is untouched and this changes nothing about that request).

What was wrong

The EdgeSwitch XP family is Ubiquiti's 24 V passive line, and four catalogued members disagreed with the vendor about it.

devices/ubiquiti/es-10x.json claimed poeOut: true, poeOutClass: "PoE" on all eight copper ports plus the poe-pse capability. Ubiquiti's own copy for that SKU says "PoE passthrough support" where the ES-10XP beside it says "each equipped with a 24V passive PoE output", and the supplies settle the disagreement: the 10X ships a 24V/1A (24 W) adapter, the 10XP a 24V/2.5A (60 W) one, against a published 15 W per-port cap that eight sourcing ports would need 120 W for. The eight source claims are gone and poe-pse is dropped - which matters mechanically, because isPoePse() (engine/media.js:1122) promotes every copper port of a switch that declares that capability to a 15.4 W source.

ES-10XP, ES-18X, ES-5XP framed the same fixed-voltage ports as poeOutClass: "PoE", which this catalogue's vocabulary defines as 802.3af - a negotiated 48 V standard (docs/DEVICE_SCHEMA.md, "Spelling a passive port PoE is not a cosmetic slip"; the fault test/uisp-r-poe.test.js and test/ep-s16-poe.test.js were written for). All three are now passive, which is what es-16xp.json in the same line already declared.

Vendor figures, read 2026-11-20 (HTTP 200, crawler UA)

URL bytes the rows used
https://techspecs.ui.com/uisp/wired/es-10x 574,925 "PoE interface 24V DC passive PoE (Pins 4, 5+; 7, 8-)"; "Power method 24V DC, 1A power adapter (Included), 24V passive PoE"; "Max. passive PoE wattage per port 15W"; "Max. power consumption 8W (Excluding PoE output)"; meta description "PoE passthrough support"
https://techspecs.ui.com/uisp/wired/es-10xp 575,764 same rows with a 2.5 A adapter; LEDs "RJ45 ports: link/speed/activity, PoE"
https://techspecs.ui.com/uisp/wired/es-18x 580,575 "PoE interface (1) 24V passive PoE port (Pins 4, 5+; 7, 8-)"; "Max. power consumption 14W (Excludes PoE output)"
https://techspecs.ui.com/uisp/wired/es-5xp 574,944 "PoE method Passive"; "PoE Out voltage range 22-24V DC"; "Max. PoE wattage per port 11.5W"; "PoE configurable per port ... Data ports: Off/24V"

https://dl.ui.com/ds/es-10x_ds.pdf answers 302 to the tech-spec root - no datasheet is published for that SKU (udw_ds.pdf on the same host answers 200 application/pdf), and https://dl.ui.com/qig/es-10x.pdf redirects likewise.

What was deliberately not written

  • The 10X's passthrough port pair. The page gives the pinout and no port count, so which port receives and which passes through is recorded as an open question in notes, not guessed.
  • Budgets for the 18X and 5XP. Neither page publishes a total. On the 18X the arithmetic argues against deriving one: 24 V x 1 A = 24 W less the 14 W the box draws leaves 10 W, which is below the vendor's own 15 W per-port cap, so a 15 W budget would claim headroom the included supply cannot deliver. The 5XP's only figure is "Max. power consumption: 60W", whole-box - the position es-48-750w.json already takes.
  • The 10XP budget is derived, not quoted, and says so: 60 W adapter less the 8 W box draw = 52 W, flagged DERIVED CEILING in the poe.note, in the same shape as devices/ubiquiti/uisp-r.json and devices/antaira/lmp-1002g-sfp-t.json.

Gates

node scripts/syntax-check.js            489 files: all OK
node scripts/audit-port-types.js        No port contradicts its declared type.
node scripts/import-devices.js --dry-run  (only pre-existing research/incoming duplicates)
node test/hardware-invariants.test.js   26 passed, 0 failed
node test/catalogue-hygiene.test.js     25 passed, 0 failed
node test/spec-vs-description.test.js    3 passed, 0 failed
node test/doc-catalog-counts.test.js    43 checks passed
node test/import-devices-capabilities.test.js  34 passed, 0 failed
node test/es-10x-passthrough.test.js    18 passed, 0 failed   (new)

The new test locks the 10X at zero PSE ports after topology expansion (so inference cannot restore them), the XP at eight passive ports with its 52 W ceiling, and the class change on both single-PoE siblings.

docs/DEVICE_REGISTRY.md regenerated: no diff (model count unchanged at 2,355).

Refs #130 (reference only; the UDW-PRO blocker is untouched and this changes nothing about that request). ## What was wrong The EdgeSwitch XP family is Ubiquiti's 24 V passive line, and four catalogued members disagreed with the vendor about it. **`devices/ubiquiti/es-10x.json`** claimed `poeOut: true, poeOutClass: "PoE"` on all eight copper ports plus the `poe-pse` capability. Ubiquiti's own copy for that SKU says *"PoE passthrough support"* where the ES-10XP beside it says *"each equipped with a 24V passive PoE output"*, and the supplies settle the disagreement: the 10X ships a **24V/1A (24 W)** adapter, the 10XP a **24V/2.5A (60 W)** one, against a published **15 W per-port** cap that eight sourcing ports would need 120 W for. The eight source claims are gone and `poe-pse` is dropped - which matters mechanically, because `isPoePse()` (`engine/media.js:1122`) promotes **every** copper port of a `switch` that declares that capability to a 15.4 W source. **ES-10XP, ES-18X, ES-5XP** framed the same fixed-voltage ports as `poeOutClass: "PoE"`, which this catalogue's vocabulary defines as 802.3af - a negotiated 48 V standard (docs/DEVICE_SCHEMA.md, "Spelling a passive port PoE is not a cosmetic slip"; the fault `test/uisp-r-poe.test.js` and `test/ep-s16-poe.test.js` were written for). All three are now `passive`, which is what `es-16xp.json` in the same line already declared. ## Vendor figures, read 2026-11-20 (HTTP 200, crawler UA) | URL | bytes | the rows used | |---|---|---| | `https://techspecs.ui.com/uisp/wired/es-10x` | 574,925 | "PoE interface 24V DC passive PoE (Pins 4, 5+; 7, 8-)"; "Power method 24V DC, 1A power adapter (Included), 24V passive PoE"; "Max. passive PoE wattage per port 15W"; "Max. power consumption 8W (Excluding PoE output)"; meta description "PoE passthrough support" | | `https://techspecs.ui.com/uisp/wired/es-10xp` | 575,764 | same rows with a 2.5 A adapter; LEDs "RJ45 ports: link/speed/activity, PoE" | | `https://techspecs.ui.com/uisp/wired/es-18x` | 580,575 | "PoE interface (1) 24V passive PoE port (Pins 4, 5+; 7, 8-)"; "Max. power consumption 14W (Excludes PoE output)" | | `https://techspecs.ui.com/uisp/wired/es-5xp` | 574,944 | "PoE method Passive"; "PoE Out voltage range 22-24V DC"; "Max. PoE wattage per port 11.5W"; "PoE configurable per port ... Data ports: Off/24V" | `https://dl.ui.com/ds/es-10x_ds.pdf` answers **302** to the tech-spec root - no datasheet is published for that SKU (`udw_ds.pdf` on the same host answers 200 `application/pdf`), and `https://dl.ui.com/qig/es-10x.pdf` redirects likewise. ## What was deliberately not written - **The 10X's passthrough port pair.** The page gives the pinout and no port count, so which port receives and which passes through is recorded as an open question in `notes`, not guessed. - **Budgets for the 18X and 5XP.** Neither page publishes a total. On the 18X the arithmetic argues against deriving one: 24 V x 1 A = 24 W less the 14 W the box draws leaves **10 W**, which is *below* the vendor's own 15 W per-port cap, so a 15 W budget would claim headroom the included supply cannot deliver. The 5XP's only figure is "Max. power consumption: 60W", whole-box - the position `es-48-750w.json` already takes. - **The 10XP budget is derived, not quoted**, and says so: 60 W adapter less the 8 W box draw = 52 W, flagged `DERIVED CEILING` in the `poe.note`, in the same shape as `devices/ubiquiti/uisp-r.json` and `devices/antaira/lmp-1002g-sfp-t.json`. ## Gates ``` node scripts/syntax-check.js 489 files: all OK node scripts/audit-port-types.js No port contradicts its declared type. node scripts/import-devices.js --dry-run (only pre-existing research/incoming duplicates) node test/hardware-invariants.test.js 26 passed, 0 failed node test/catalogue-hygiene.test.js 25 passed, 0 failed node test/spec-vs-description.test.js 3 passed, 0 failed node test/doc-catalog-counts.test.js 43 checks passed node test/import-devices-capabilities.test.js 34 passed, 0 failed node test/es-10x-passthrough.test.js 18 passed, 0 failed (new) ``` The new test locks the 10X at zero PSE ports *after* topology expansion (so inference cannot restore them), the XP at eight `passive` ports with its 52 W ceiling, and the class change on both single-PoE siblings. `docs/DEVICE_REGISTRY.md` regenerated: no diff (model count unchanged at 2,355).
devices(ubiquiti): an ES-10X that sources nothing, and three 24 V ports filed as 802.3af
All checks were successful
CI / Classify the change (pull_request) Successful in 13s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 13s
CI / Test (Node 20) (pull_request) Successful in 57s
CI / Test (Node 22) (pull_request) Successful in 58s
CI / Server smoke and Docker build (pull_request) Successful in 28s
CI / Test (Node 18) (pull_request) Successful in 1m2s
CI / Arm auto-merge (pull_request) Successful in 4s
a959dfa42e
The EdgeSwitch XP line is 24 V passive PoE. Four catalogued members said otherwise.

ES-10X claimed poeOut on all eight copper ports and the poe-pse capability.
Ubiquiti's own copy says "PoE passthrough support" where the ES-10XP beside it
says "each equipped with a 24V passive PoE output", and the supplies settle it:
24V/1A (24 W) on the 10X against 24V/2.5A (60 W) on the 10XP, with a published
15 W per-port cap that eight ports would need 120 W for. The eight source claims
and poe-pse are gone; passive-poe is kept.

ES-10XP, ES-18X and ES-5XP framed the same fixed-voltage ports as
poeOutClass "PoE", which this catalogue reads as 802.3af - a negotiated 48 V
standard these boxes do not speak (docs/DEVICE_SCHEMA.md; the fault
test/uisp-r-poe.test.js and test/ep-s16-poe.test.js cover). All three are now
"passive", matching es-16xp.

The 10XP gets a derived 52 W ceiling (60 W adapter less 8 W box draw), marked
DERIVED in its note, in the shape of devices/ubiquiti/uisp-r.json. 18X and 5XP
get no budget on purpose: neither page publishes a total, and on the 18X the
24 W adapter minus its 14 W draw is below the 15 W per-port cap, so the two
numbers disagree.

Sources read 2026-11-20, HTTP 200, crawler UA: techspecs.ui.com/uisp/wired/
es-10x (574,925 B), es-10xp (575,764 B), es-18x (580,575 B), es-5xp (574,944 B).
dl.ui.com/ds/es-10x_ds.pdf is 302 to the tech-spec root; no datasheet exists for
that SKU.

Refs #130 only.
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-20 10:00:28 +00:00
jaylouisw deleted branch catalogue/1120-0952 2026-11-20 10:09:29 +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!2290
No description provided.