devices(poe): six budgets that were only in prose, a GS724TP that mixed two revisions, and a Moxa whose ports were inferred at 802.3af #2295

Merged
jaylouisw merged 1 commit from rev/1120-1812 into main 2026-11-20 18:38:58 +00:00
Owner

Six specs sourced power, had no poe.budgetW, and already quoted a wattage in their own description - the cohort scripts/audit-poe-budget.js calls "prose". Every one is now sourced to the vendor's own document:

spec figure
aruba/2540-48g-poe-4sfp-plus-jl357a 370 W (HPE QuickSpecs)
netgear/GS724TP 190 W (GS724TPv2 data sheet)
moxa/EDS-G4012-8P-4QGS-LVB-T 150 W (24 VDC row; 180 W @48 V, 62 W @12 V)
dahua/NVR5216-16P-EI2 130 W
hikvision/DS-7616NI-K2/16P 200 W
hikvision/DS-7608NI-K2/8P 80 W

Audit: 61 of 411 PSEs silent, 6 in prose -> 55 silent, 0 in prose.

Two were more than a missing number:

  • GS724TP quoted the 2007 802.3af v1's 192 W (225 W supply, 2 combo copper/SFP slots) beside the PoE+ revision's ports (24 PoE+, 2 dedicated SFP). The PoE+ layout's budget is 190 W, which is what the entry is now; the v1 figures stay in poe.note.
  • Moxa EDS-G4012-8P-4QGS-LVB-T declared no PoE at all, so the engine inferred 15.4 W 802.3af on all eight copper ports of an IEEE 802.3bt switch rated at 90 W a port. The ports are now declared (poeOut: 90, PoE+++, poe-pse/802.3bt) and the 150 W budget written.

No engine code changed. test/poe-budget-prose-six.test.js pins the six figures, their vendor-host URLs, poeBudgetSource === 'research', Available:<n>.0(w) from show power inline, that each budget is below the sum of its own port ceilings, and the two special cases (47 checks).

Unresolved: the Moxa's supply-dependent total (62/150/180 W) needs poe.budgetVariants from docs/plans/poe-budget-variants.md; whether GS724TP should also exist as a v1 SKU with combo ports. Details: research/poe-budget-prose-six-remaining.md.

Test totals: npm test - 412 suites, 13674 assertions, 0 failed.

Six specs sourced power, had no `poe.budgetW`, and already quoted a wattage in their own `description` - the cohort `scripts/audit-poe-budget.js` calls "prose". Every one is now sourced to the vendor's own document: | spec | figure | |--|--| | aruba/2540-48g-poe-4sfp-plus-jl357a | 370 W (HPE QuickSpecs) | | netgear/GS724TP | 190 W (GS724TPv2 data sheet) | | moxa/EDS-G4012-8P-4QGS-LVB-T | 150 W (24 VDC row; 180 W @48 V, 62 W @12 V) | | dahua/NVR5216-16P-EI2 | 130 W | | hikvision/DS-7616NI-K2/16P | 200 W | | hikvision/DS-7608NI-K2/8P | 80 W | Audit: **61 of 411 PSEs silent, 6 in prose -> 55 silent, 0 in prose.** Two were more than a missing number: * **GS724TP** quoted the 2007 802.3af v1's 192 W (225 W supply, 2 *combo* copper/SFP slots) beside the PoE+ revision's ports (24 PoE+, 2 *dedicated* SFP). The PoE+ layout's budget is 190 W, which is what the entry is now; the v1 figures stay in `poe.note`. * **Moxa EDS-G4012-8P-4QGS-LVB-T** declared no PoE at all, so the engine inferred 15.4 W 802.3af on all eight copper ports of an IEEE 802.3bt switch rated at 90 W a port. The ports are now declared (`poeOut: 90`, `PoE+++`, `poe-pse`/`802.3bt`) and the 150 W budget written. No engine code changed. `test/poe-budget-prose-six.test.js` pins the six figures, their vendor-host URLs, `poeBudgetSource === 'research'`, `Available:<n>.0(w)` from `show power inline`, that each budget is below the sum of its own port ceilings, and the two special cases (47 checks). Unresolved: the Moxa's supply-dependent total (62/150/180 W) needs `poe.budgetVariants` from `docs/plans/poe-budget-variants.md`; whether GS724TP should also exist as a v1 SKU with combo ports. Details: `research/poe-budget-prose-six-remaining.md`. Test totals: `npm test` - 412 suites, 13674 assertions, 0 failed.
devices(poe): six budgets that were only in prose, a GS724TP that mixed two revisions, and a Moxa whose ports were inferred at 802.3af
All checks were successful
CI / Classify the change (pull_request) Successful in 28s
CI / Test (Node 22) (pull_request) Successful in 1m9s
CI / Test (Node 20) (pull_request) Successful in 1m13s
CI / Test (Node 18) (pull_request) Successful in 1m19s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 12s
CI / Server smoke and Docker build (pull_request) Successful in 34s
CI / Arm auto-merge (pull_request) Successful in 5s
9a56f8797f
The audit's prose cohort named six specs that source power, carry no poe.budgetW and already
quote a wattage in their own description. All six vendor documents state the figure; each spec now
carries it with a source URL, so show power inline answers with a budget instead of 'budget not
modelled'. Audit: 61 of 411 PSEs silent / 6 in prose -> 55 / 0.

Two were more than a missing number. GS724TP's 192 W is the 2007 802.3af v1's figure (225 W supply,
2 combo copper/SFP slots) sitting beside the PoE+ revision's ports; the PoE+ layout is 190 W, which
is what the entry is now. The Moxa EDS-G4012-8P-4QGS-LVB-T had no declared PoE at all, so the
engine inferred 15.4 W 802.3af on all eight copper ports of an 802.3bt switch rated at 90 W a port;
the ports are declared and the 150 W budget written, with the 12 VDC (62 W) and 48 VDC (180 W) rows
kept in the note because one field cannot hold them.

research/poe-budget-prose-six-remaining.md records the quotes, the revision split and what is still
open. test/poe-budget-prose-six.test.js pins the six figures and the two special cases: 47 checks.
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-20 18:25:39 +00:00
jaylouisw deleted branch rev/1120-1812 2026-11-20 18:39:01 +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!2295
No description provided.