devices(huawei): model OceanProtect E8000 (HDD and NVMe) from the Chinese technical white paper #1930

Merged
jaylouisw merged 1 commit from catalogue/1118-0256 into main 2026-11-18 03:12:23 +00:00
Owner

Fixes #742

The E8000's front-end census is published, in Chinese: the product technical white paper, which Huawei serves anonymously from e-file.huawei.com behind the Chinese product page's "白皮书" link and does not offer on the English page (which is why four earlier passes counted 224 /en/documents/products/storage/ pages, found two E8000 data sheets, and concluded the hardware book was login-gated).

Route, all HTTP 200 with a crawler user-agent:

What it settles, verbatim (section 3.1.3 OceanProtect E8000, p.14): a 2 U two-controller frame with 最多12个可热插拔的接口模块 per control frame and 每控2个GE管理/维护口; and (per-model table, pp.7-9) every interface module's port count - 4-port 8G/16G/32G/64G FC, 4-port 10GE/25GE ETH, 2-port 40GE/100GE ETH front ends, 4-port 25Gb RDMA Scale-Out, 4-port 12Gb SAS (HDD form) or 100Gb RDMA (NVMe form) back ends. The front-end census is now a multiplication, not a guess.

Caveat, recorded in both specs rather than papered over: section 3.1.3's first sentence opens with the X8000's model name (a vendor copy-paste slip); the heading, figures 3-10/3-11 and the 36-slot description are the E8000's, and the 12-slot sentence is the one they introduce.

Modelled: devices/huawei/oceanprotect-e8000-hdd.json and devices/huawei/oceanprotect-e8000-nvme.json - the two forms the table names - each with a stated configuration (one 4-port 10GE/25GE module in IOM 0 and one in IOM 1 per controller: 16 x 10/25GbE SFP28) plus 4 x 1GbE management ports. Not modelled and named in notes: FC and 40/100GE modules, Scale-Out RDMA (inter-enclosure), back-end modules (disk frames), serial.

Also: research/oceanprotect-e8000-cn-whitepaper.md; #742 removed from research/open-device-requests.json; registry regenerated (2,353 models); the closed-issue audit's live-row canary moved from #742 to #130.

Gates: node scripts/syntax-check.js 403 files OK; node scripts/audit-port-types.js 2,353 models, 0 offending ports; node test/hardware-invariants.test.js 26/26; importer dry run over the two specs - "Would add: 2 duplicates: 0 rejected: 0"; npm test 330 suites, 11,462 assertions, 0 failed.

Fixes #742 The E8000's front-end census is published, in Chinese: the product technical white paper, which Huawei serves anonymously from `e-file.huawei.com` behind the Chinese product page's "白皮书" link and does not offer on the English page (which is why four earlier passes counted 224 `/en/documents/products/storage/` pages, found two E8000 data sheets, and concluded the hardware book was login-gated). Route, all HTTP 200 with a crawler user-agent: - <https://e.huawei.com/cn/products/storage/oceanprotect/oceanprotect-e8000> - 360,565 B, spec table (`单节点控制器数量 2`, `前端通道端口类型 16/32/64 Gb FC, 10/25/40/100 GbE`) - <https://e.huawei.com/cn/documents/products/storage/4b552a977d394d1e802814db6a8d4a0a> - 332,980 B, *华为OceanProtect专用备份存储 产品技术白皮书 V200R001C20*, whose hidden `attachmentUrl` carries the asset path - the asset itself - `application/pdf`, **19,224,057 B, 141 pp**, document version 03 (2026-05-06) What it settles, verbatim (section 3.1.3 OceanProtect E8000, p.14): a 2 U two-controller frame with **最多12个可热插拔的接口模块** per control frame and **每控2个GE管理/维护口**; and (per-model table, pp.7-9) every interface module's port count - 4-port 8G/16G/32G/64G FC, 4-port 10GE/25GE ETH, 2-port 40GE/100GE ETH front ends, 4-port 25Gb RDMA Scale-Out, 4-port 12Gb SAS (HDD form) or 100Gb RDMA (NVMe form) back ends. The front-end census is now a multiplication, not a guess. Caveat, recorded in both specs rather than papered over: section 3.1.3's first sentence opens with the X8000's model name (a vendor copy-paste slip); the heading, figures 3-10/3-11 and the 36-slot description are the E8000's, and the 12-slot sentence is the one they introduce. Modelled: `devices/huawei/oceanprotect-e8000-hdd.json` and `devices/huawei/oceanprotect-e8000-nvme.json` - the two forms the table names - each with a stated configuration (one 4-port 10GE/25GE module in IOM 0 and one in IOM 1 per controller: 16 x 10/25GbE SFP28) plus 4 x 1GbE management ports. Not modelled and named in notes: FC and 40/100GE modules, Scale-Out RDMA (inter-enclosure), back-end modules (disk frames), serial. Also: `research/oceanprotect-e8000-cn-whitepaper.md`; #742 removed from `research/open-device-requests.json`; registry regenerated (2,353 models); the closed-issue audit's live-row canary moved from #742 to #130. Gates: `node scripts/syntax-check.js` 403 files OK; `node scripts/audit-port-types.js` 2,353 models, 0 offending ports; `node test/hardware-invariants.test.js` 26/26; importer dry run over the two specs - "Would add: 2 duplicates: 0 rejected: 0"; `npm test` 330 suites, 11,462 assertions, 0 failed.
devices(huawei): OceanProtect E8000 in both forms, from the Chinese technical white paper (#742)
All checks were successful
CI / Classify the change (pull_request) Successful in 6s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 8s
CI / Server smoke test (pull_request) Successful in 8s
CI / Docker build (pull_request) Successful in 23s
CI / Test (Node 20) (pull_request) Successful in 56s
CI / Test (Node 18) (pull_request) Successful in 1m5s
CI / Test (Node 22) (pull_request) Successful in 1m3s
CI / Arm auto-merge (pull_request) Successful in 3s
d8eb90c21f
The E8000's front-end census is published - in the Chinese product technical
white paper, not in either data sheet. Four earlier passes counted the /en/
document tree (224 storage pages, two E8000 data sheets, `slot` 0 times) and
concluded the hardware chapter was login-gated. It is: on support.huawei.com.
Huawei also publishes it anonymously on e-file.huawei.com, behind the Chinese
product page's "白皮书" link, which the English page does not carry.

  https://e.huawei.com/cn/products/storage/oceanprotect/oceanprotect-e8000
  -> https://e.huawei.com/cn/documents/products/storage/4b552a977d394d1e802814db6a8d4a0a
  -> 华为OceanProtect专用备份存储 产品技术白皮书 V200R001C20.pdf
     (19,224,057 B, 141 pp, document version 03, 2026-05-06)

Section 3.1.3 (p.14) gives the frame: 2 U, two controllers, up to 12 hot-swappable
interface modules per control frame, two GE management/maintenance ports and one
serial port per controller. The per-model table (pp.7-9) gives every front-end
module's port count - 4-port FC, 4-port 10GE/25GE, 2-port 40GE/100GE - plus the
Scale-Out and back-end modules. Front-end port count is now a multiplication.

Modelled: the two forms the table names, each as one node with a stated
configuration (one 4-port 10GE/25GE module in IOM 0 and one in IOM 1 per
controller = 16 front-end SFP28 ports) plus the 4 x 1GbE management ports.
FC, 40/100GE, Scale-Out RDMA and back-end modules are named in notes as not
modelled, so silence is not read as absence.

The section's first sentence mistypes the model as X8000 (a vendor copy-paste
slip); both specs record it verbatim rather than quietly correcting it, because
that sentence is the only vendor statement of the slot count.

Also: #742 out of the open-request ledger, registry regenerated (2353 models),
and the closed-issue audit's canary moved from #742 to #130.
jaylouisw force-pushed catalogue/1118-0256 from d8eb90c21f
All checks were successful
CI / Classify the change (pull_request) Successful in 6s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 8s
CI / Server smoke test (pull_request) Successful in 8s
CI / Docker build (pull_request) Successful in 23s
CI / Test (Node 20) (pull_request) Successful in 56s
CI / Test (Node 18) (pull_request) Successful in 1m5s
CI / Test (Node 22) (pull_request) Successful in 1m3s
CI / Arm auto-merge (pull_request) Successful in 3s
to 3681ba7317
All checks were successful
CI / Classify the change (pull_request) Successful in 8s
PR rebase / Rebase onto main and arm auto-merge (pull_request) Successful in 6s
CI / Server smoke test (pull_request) Successful in 7s
CI / Test (Node 20) (pull_request) Successful in 52s
CI / Docker build (pull_request) Successful in 20s
CI / Test (Node 18) (pull_request) Successful in 1m1s
CI / Test (Node 22) (pull_request) Successful in 58s
CI / Arm auto-merge (pull_request) Successful in 11s
2026-11-18 03:10:06 +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!1930
No description provided.