devices(mikrotik): the RB260GSP's product-code entry lit five 802.3af sockets on a four-socket passive box #2415

Merged
jaylouisw merged 1 commit from rev/1121-1142 into main 2026-11-21 12:08:40 +00:00
Owner

MikroTik's RB260GSP is in the catalogue twice - as RB260GSP (#1153) and under its product code CSS106-1G-4P-1S (#405). The MikroTik PoE census (#2294) repaired the first and never looked at the second, so that copy still carried the engine's switch-wide guess: five RJ45 ports at 15.4 W of 802.3af.

The vendor page says otherwise twice: PoE-out ports Ether2-Ether5 (Ether1 is the DC jack, PoE-IN inlet) and PoE out Passive PoE. Budget is a current - Max total out (A) 2 A, 1 A per port against 11-30 V - so the entry is marked budgetUnknown rather than credited with arithmetic MikroTik did not do.

Changed: devices/mikrotik/css106-1g-4p-1s.json (eth1 poeOut: false, eth2-eth5 poeOut: true / passive, poe.budgetUnknown with the quoted limit), test/rb260gsp-poe-out-census.test.js (36 assertions, 11 of which fail on the old spec), research/rb260gsp-two-entries.md.

Source: https://mikrotik.com/product/RB260GSP (200, text/html, 344,158 B, read 2026-11-21).

node scripts/syntax-check.js: 530 files OK. npm test: 450 suites, 15,279 assertions, 0 failed. audit-poe-budget: 22 -> 21 silent, 24 -> 25 unknown.

Unresolved: whether one box should hold two catalogue entries at all (the same name/product-code split exists for hAP ac2 and Chateau 5G); and the inlet's draw, which MikroTik publishes as a voltage window only.

MikroTik's RB260GSP is in the catalogue twice - as `RB260GSP` (#1153) and under its product code `CSS106-1G-4P-1S` (#405). The MikroTik PoE census (#2294) repaired the first and never looked at the second, so that copy still carried the engine's switch-wide guess: five RJ45 ports at 15.4 W of 802.3af. The vendor page says otherwise twice: `PoE-out ports Ether2-Ether5` (Ether1 is the `DC jack, PoE-IN` inlet) and `PoE out Passive PoE`. Budget is a **current** - `Max total out (A) 2 A`, 1 A per port against 11-30 V - so the entry is marked `budgetUnknown` rather than credited with arithmetic MikroTik did not do. Changed: `devices/mikrotik/css106-1g-4p-1s.json` (eth1 `poeOut: false`, eth2-eth5 `poeOut: true` / `passive`, `poe.budgetUnknown` with the quoted limit), `test/rb260gsp-poe-out-census.test.js` (36 assertions, 11 of which fail on the old spec), `research/rb260gsp-two-entries.md`. Source: https://mikrotik.com/product/RB260GSP (200, text/html, 344,158 B, read 2026-11-21). `node scripts/syntax-check.js`: 530 files OK. `npm test`: 450 suites, 15,279 assertions, 0 failed. `audit-poe-budget`: 22 -> 21 silent, 24 -> 25 unknown. Unresolved: whether one box should hold two catalogue entries at all (the same name/product-code split exists for hAP ac2 and Chateau 5G); and the inlet's draw, which MikroTik publishes as a voltage window only.
devices(mikrotik): the RB260GSP's product-code entry lit five 802.3af sockets on a four-socket passive box
All checks were successful
CI / Server smoke and Docker build (pull_request) Successful in 30s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 5s
CI / Test (Node 20) (pull_request) Successful in 1m5s
CI / Test (Node 22) (pull_request) Successful in 1m6s
CI / Test (Node 18) (pull_request) Successful in 1m10s
CI / Arm auto-merge (pull_request) Successful in 4s
4a33e5af54
MikroTik's product is in the catalogue twice - as `RB260GSP` (#1153) and as its product code
`CSS106-1G-4P-1S` (#405) - and the PoE census (#2294) repaired only the first, because it is
keyed by file and the second was not on its list. That copy kept the engine's switch-wide guess:
every RJ45 port of a `poe-pse` switch sources 15.4 W of 802.3af unless the spec says otherwise.

MikroTik's specification table says otherwise twice over (https://mikrotik.com/product/RB260GSP,
read 2026-11-21): `PoE-out ports Ether2-Ether5` (Ether1 is the `DC jack, PoE-IN` inlet, not an
outlet) and `PoE out Passive PoE` (a fixed-voltage pair, not a negotiated 48 V standard). The
entry now declares `poeOut: false` on eth1, `poeOut: true` / `poeOutClass: "passive"` on eth2-eth5,
and a `poe` block marked `budgetUnknown` quoting the vendor's `Max total out (A) 2 A` - no wattage,
because 2 A against an 11-30 V window is arithmetic MikroTik did not do.

test/rb260gsp-poe-out-census.test.js (36 assertions) pins the port set, the class, the absent
wattage, `show power inline`'s four rows, and that the two entries agree port for port.
research/rb260gsp-two-entries.md records the sources and the open question: whether one box should
hold two catalogue entries at all.

audit-poe-budget: 22 -> 21 silent PSEs, 24 -> 25 marked unknown.
forgejo-actions scheduled this pull request to auto merge when all checks succeed 2026-11-21 11:51:44 +00:00
jaylouisw deleted branch rev/1121-1142 2026-11-21 12:08:43 +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!2415
No description provided.