Huawei NetEngine 8000 F1A: two 1U edge routers, 56 ports each, and the chassis the 1.2 Tbit/s figure belongs to (#790) #1534

Merged
jaylouisw merged 1 commit from catalogue/1115-1955 into main 2026-11-15 20:17:17 +00:00
Owner

Fixes #790

What

The catalogue had the fixed M-series NetEngine 8000 boxes (M1A/M1C/M1D/M1D-B) and no F-series at all. This adds the two 1U fixed F1A variants, both from Huawei's own hardware descriptions rather than a marketing table.

model ports 10GE 25GE 100GE switching typical power
NetEngine 8000 F1A-8H20Q 56 28 (0-27, SFP+) 20 (28-47, SFP28) 8 (48-55, QSFP28) 2.4 Tbit/s, 453 Mpps 325 W
NetEngine 8000 F1A-C 56 28 (0-27, SFP+) 20 (28-47, SFP28) 8 (48-55, QSFP28) 2.4 Tbit/s 336.3 W

Two files and not one because the panels are identical and the port maps are not: the 8H20Q splits its 25GE range into 28-35 (25GE/10GE/GE, two port groups whose mode changes together) and 36-47 (25GE/10GE only), and it is the only one of the two that breaks a 100GE port out to 4x25GE or 4x10GE. The F1A-C datasheet answers breakout No, interface rate auto-sensing No, FlexE No, MACsec No. Modelling them as one device would have invented a port mode on the C.

The discrepancy, stated

The F-series marketing page gives the F1A "a maximum capacity of 1.2 Tbps" (1200 G/U capacity density) while the V800R023C10SPC500 hardware description for this chassis states 2.4 Tbit/s switching capacity. The hardware description wins here and is the sourceUrl; the marketing figure is recorded in notes so the next reader does not "correct" the spec back to 1.2 T. Same disagreement on the F1A-C.

Sources

Management ports are recorded in notes and not modelled: MGMT-ETH (RJ45), Console/AUX RJ45 hybrid at 9600 bit/s, TOD and CLK RJ45 clock ports, USB, and the two reserved SFP 1G cascade ports GE0/GE1. The data-port list is exactly ports 0-55, which is the number Huawei states as "maximum number of physical ports on the entire device".

Validation

  • node scripts/import-devices.js (PF_INCOMING on the batch, --apply): Added: 2 duplicates: 0 rejected: 0, no warnings. Re-run is idempotent (Duplicates (skipped, existing entry kept): 2).
  • node scripts/audit-port-types.js: No port contradicts its declared type. Both boxes resolve to 28x sfp+, 20x sfp28, 8x qsfp28 — checked through devices/registry.effectiveInterfaces + engine/media.derivePortHardware.
  • node scripts/audit-devices.js: No hard defects.
  • node scripts/syntax-check.js: 321 files, all OK. scripts/check-control-bytes.js: clean.
  • Targeted tests: port-count 15/15, importer-roundtrip 26/26, port-roundtrip 32/32, wired-port-has-media 36/36, device-aliases 13/13, catalogue-hygiene 25/25, import-devices-capabilities 29/29, faceplate-catalogue 14/14, inspector-port-count 10/10, doc-catalog-counts 24/24.
  • node scripts/regen-registry-doc.js: 2,045 → 2,047 models, routers 369 → 371, Huawei 82 → 84; the page was regenerated in this branch rather than left behind.

Not done / unverifiable

  • F8 (#789) is a different problem: 13U, eight service slots, forwarding capacity that differs per slot (400 Gbit/s in 3-6, 200 Gbit/s in 1, 2, 7, 8) and parent-child LPU/PIC population. Picking one population would be a guess, so it is left for a run that reads the board list properly.
  • No Huawei figure found for an F1A PoE budget because there is none: neither chassis sources power on a data port. Nothing was invented to fill the gap.
Fixes #790 ## What The catalogue had the fixed **M-series** NetEngine 8000 boxes (M1A/M1C/M1D/M1D-B) and no **F-series** at all. This adds the two 1U fixed F1A variants, both from Huawei's own hardware descriptions rather than a marketing table. | model | ports | 10GE | 25GE | 100GE | switching | typical power | |---|---|---|---|---|---|---| | NetEngine 8000 F1A-8H20Q | 56 | 28 (0-27, SFP+) | 20 (28-47, SFP28) | 8 (48-55, QSFP28) | 2.4 Tbit/s, 453 Mpps | 325 W | | NetEngine 8000 F1A-C | 56 | 28 (0-27, SFP+) | 20 (28-47, SFP28) | 8 (48-55, QSFP28) | 2.4 Tbit/s | 336.3 W | Two files and not one because the panels are identical and the port *maps* are not: the 8H20Q splits its 25GE range into 28-35 (25GE/10GE/**GE**, two port groups whose mode changes together) and 36-47 (25GE/10GE only), and it is the only one of the two that breaks a 100GE port out to 4x25GE or 4x10GE. The F1A-C datasheet answers breakout `No`, interface rate auto-sensing `No`, FlexE `No`, MACsec `No`. Modelling them as one device would have invented a port mode on the C. ## The discrepancy, stated The F-series marketing page gives the F1A "a maximum capacity of 1.2 Tbps" (1200 G/U capacity density) while the V800R023C10SPC500 hardware description for this chassis states **2.4 Tbit/s switching capacity**. The hardware description wins here and is the `sourceUrl`; the marketing figure is recorded in `notes` so the next reader does not "correct" the spec back to 1.2 T. Same disagreement on the F1A-C. ## Sources - F1A-8H20Q: https://support.huawei.com/enterprise/en/doc/EDOC1100382055/ce5c8c83/netengine-8000-f1a-8h20q (port table 1-5, technical spec table 1-7) - F1A-C: https://support.huawei.com/enterprise/en/doc/EDOC1100465707/7044fd1d/netengine-8000-f1a-c (port table 1-5, table 1-7 for CR8B0BKP03E0) - Platform role (SR/BNG/CGN/IPSec → `nat`, `firewall`, `dhcp-server`): https://carrier.huawei.com/en/products/fixed-network/data-communication/router/NetEngine-8000-Series-Router/NetEngine-8000-F-Series Management ports are recorded in `notes` and **not** modelled: MGMT-ETH (RJ45), Console/AUX RJ45 hybrid at 9600 bit/s, TOD and CLK RJ45 clock ports, USB, and the two reserved SFP 1G cascade ports GE0/GE1. The data-port list is exactly ports 0-55, which is the number Huawei states as "maximum number of physical ports on the entire device". ## Validation - `node scripts/import-devices.js` (PF_INCOMING on the batch, `--apply`): `Added: 2 duplicates: 0 rejected: 0`, no warnings. Re-run is idempotent (`Duplicates (skipped, existing entry kept): 2`). - `node scripts/audit-port-types.js`: `No port contradicts its declared type.` Both boxes resolve to 28x `sfp+`, 20x `sfp28`, 8x `qsfp28` — checked through `devices/registry.effectiveInterfaces` + `engine/media.derivePortHardware`. - `node scripts/audit-devices.js`: `No hard defects.` - `node scripts/syntax-check.js`: 321 files, all OK. `scripts/check-control-bytes.js`: clean. - Targeted tests: `port-count` 15/15, `importer-roundtrip` 26/26, `port-roundtrip` 32/32, `wired-port-has-media` 36/36, `device-aliases` 13/13, `catalogue-hygiene` 25/25, `import-devices-capabilities` 29/29, `faceplate-catalogue` 14/14, `inspector-port-count` 10/10, `doc-catalog-counts` 24/24. - `node scripts/regen-registry-doc.js`: 2,045 → 2,047 models, routers 369 → 371, Huawei 82 → 84; the page was regenerated in this branch rather than left behind. ## Not done / unverifiable - **F8 (#789)** is a different problem: 13U, eight service slots, forwarding capacity that differs per slot (400 Gbit/s in 3-6, 200 Gbit/s in 1, 2, 7, 8) and parent-child LPU/PIC population. Picking one population would be a guess, so it is left for a run that reads the board list properly. - No Huawei figure found for an F1A **PoE** budget because there is none: neither chassis sources power on a data port. Nothing was invented to fill the gap.
Huawei NetEngine 8000 F1A: two 1U edge routers, 56 ports each, and the chassis the 1.2 Tbit/s figure belongs to (#790)
Some checks failed
CI / Classify the change (pull_request) Successful in 23s
CI / Test (Node 20) (pull_request) Successful in 3m9s
CI / Test (Node 18) (pull_request) Failing after 3m57s
CI / Server smoke test (pull_request) Successful in 29s
CI / Docker build (pull_request) Successful in 50s
CI / Test (Node 22) (pull_request) Successful in 3m4s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Failing after 13m20s
cd27635bb9
F1A-8H20Q and F1A-C from the V800R023C10SPC500 / V800R024C10SPC500 hardware descriptions: 28x 10GE/GE SFP+, 20x 25GE SFP28, 8x 100GE QSFP28. The 8H20Q breaks 100GE out to 4x25GE and can run ports 28-35 in 10GE/GE mode; the F1A-C datasheet says breakout No, auto-sensing No - same panel, narrower port map, so they are two files and not one.
jaylouisw scheduled this pull request to auto merge when all checks succeed 2026-11-15 20:17:06 +00:00
jaylouisw force-pushed catalogue/1115-1955 from cd27635bb9
Some checks failed
CI / Classify the change (pull_request) Successful in 23s
CI / Test (Node 20) (pull_request) Successful in 3m9s
CI / Test (Node 18) (pull_request) Failing after 3m57s
CI / Server smoke test (pull_request) Successful in 29s
CI / Docker build (pull_request) Successful in 50s
CI / Test (Node 22) (pull_request) Successful in 3m4s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Failing after 13m20s
to ab83ef2fd5
All checks were successful
CI / Classify the change (pull_request) Successful in 17s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 17s
CI / Test (Node 18) (pull_request) Successful in 3m18s
CI / Test (Node 20) (pull_request) Successful in 3m2s
CI / Server smoke test (pull_request) Has been skipped
CI / Docker build (pull_request) Has been skipped
CI / Test (Node 22) (pull_request) Successful in 3m1s
2026-11-15 20:17:08 +00:00
Compare
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!1534
No description provided.