edgecore: the ECS4120 access-switch family, five members from the R16 datasheet #1700
No reviewers
Labels
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
jaylouisw/packetforge!1700
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "catalogue/1116-1755"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Adds Edgecore's own ECS4120 L2+/L3 Lite access switches, which the catalogue did not have at all (only the bare-metal AS-series was modelled).
Port census, from the ECS4120 Series datasheet R16 (EC-DS-0324-R16, 2024-03-12)
https://www.edge-core.com/download/ecs4120-series-ds-r16/?wpdmdl=1696
Both fibre members also carry a 1x GE RJ-45 out-of-band management port and an RJ-45 console; the two copper members have a console and no OOB port (the R16 table prints No for that row).
Where the numbers come from, and what is not from R16
poe.noterather than pretending each port can run Class 4 at once.notessays so instead of implying a printed list.l3-forwardanddhcp-serverrest on. No member has NAT, and none declares it.Validation
node scripts/syntax-check.js- 356 files, OKnode scripts/audit-port-types.js- exit 0, "No port contradicts its declared type"test/hardware-invariants.test.js26/26,spec-vs-description3/3,catalogue-hygiene25/25,faceplate-catalogue14/14--all): 5 scanned, 5 duplicates (already written), 0 rejecteddoc-catalog-counts43/43,doc-citations11/11,canonical-urls52/52node scripts/regen-registry-doc.js: 2,252 -> 2,257 models, 898 port shapesUnverifiable / left alone
The catalogue had Edgecore's bare-metal data-centre boxes (AS4610, AS7726, AS9716) and nothing from the ECS line - the vendor's own-software L2+/L3 Lite access switches that the ECS4120 series belongs to. Census from the ECS4120 Series datasheet R16 (EC-DS-0324-R16, 2024-03-12), which prints a per-model port table: ECS4120-28T 24x 1G RJ-45, 4x 10G SFP+, 128 Gbps, 95 Mpps, fanless, 20 W ECS4120-52T 48x 1G RJ-45, 4x 10G SFP+, 176 Gbps, 130 Mpps, 60 W ECS4120-28Fv2 20x 100/1000 SFP, 4 combo, 4x 10G SFP+, 1x GE OOB, DC, 0-50 C ECS4120-28Fv2-I same, -10 to 65 C, dying gasp ECS4120-28P 24x 1G RJ-45 all PoE+ at 30 W a port on a 370 W budget, 4x 10G SFP+ - the PoE member, EOL 1 January 2019 Nothing is guessed. The 28P is not in R16; its 370 W budget and 30 W a port come from its product page and the Web Management Guide R03, and its 128 Gbps / 95 Mpps from datasheet R10, all cited in the spec notes. Port numbering on the fibre members is fixed at both ends by the CLI reference ("Ports 1-24 (1000BASE SFP)", "Ports 25-28 (10G SFP+)"); the four combo positions are placed at 21-24 from the datasheet port order, and the spec says so rather than implying the vendor printed a 28F port map. Firmware facts recorded: V1.2.2.30 is the current image with u-boot v2.0.0.2 as a separate upgrade; flash 256 MB, DRAM 512 MB, MAC 16 K, jumbo 9 KB, DHCP server 8 pools / 512 addresses, static routes plus RIP v1/v2, which is what the l3-forward and dhcp-server capabilities rest on. No member has NAT. Gates: syntax-check 356 files OK; audit-port-types exit 0 ("No port contradicts its declared type"); hardware-invariants 26/26, spec-vs-description 3/3, catalogue-hygiene 25/25, faceplate-catalogue 14/14; importer dry run over the five specs: 5 scanned, 0 rejected; doc-catalog-counts 43/43, doc-citations 11/11, canonical-urls 52/52.