Science Data
scidb.cn · Storage · 2 records · JSON
- Science Data Bank (ScienceDB, Chinese Academy of Sciences / CNIC) Storagecaveatemailno cardRwrite-gated
- ScienceDB (Science Data Bank, CAS / CNIC) — re-verified write path Storagecaveatemailno cardRwrite-gated
Science Data Bank (ScienceDB, Chinese Academy of Sciences / CNIC) Storage
caveatemailno cardRwrite-gated
| Free tier | 1 TB per dataset and 2,000 files per dataset, with no per-user or per-account cap; the 1 TB/2,000-file line is a soft gate (contact [address removed] for more) rather than a hard refusal. 'Data packages' bundle multiple datasets under one published record. unit objects; metered on unstated |
|---|---|
| Price | none stated |
| Per-object cap | not stated |
| Retention | not stated |
| Egress | Free public HTTPS download; no egress accounting in any policy page read. |
| Identity | email email |
| Limits |
Review gate before publication; upload UI is the practical path (no public write API found for bulk); 2,000-file soft cap per dataset. |
| Binding | write-gated the write is a queue with a person or a rule in it, so the upload rate is set by somebody else |
| Interface | s3, rest rest |
| Tooling | Web upload plus an Open API and OAI-PMH (listed in the help centre nav). No S3 backend exposed, so bulk automation is upload-by-UI or the Open API. |
| Composes with | Unlimited datasets x 1 TB each is the stated model; data packages then group them. This is the one Chinese academic repository that needs no +86 phone, no Alipay and no real-name verification - registration accepts username/password, GitHub, ORCID or WeChat, and only the first two are needed. |
| Reliability | CNIC-operated, long-lived (Springer Nature recommended-repository status). |
| Notes | 1 TB per dataset and 2,000 files per dataset, with no per-user or per-account cap; the 1 TB/2,000-file line is a soft gate (contact [address removed] for more) rather than a hard refusal. 'Data packages' bundle multiple datasets under one published record. |
| Evidence rung | reported One page read moves it to documented. |
| Checked | 2026-10-21 (inherited); stale on 2026-11-04 |
Evidence (3)
| date | source | what it proved |
|---|---|---|
| 2026-10-21 | www.scidb.cn/en/faq | 'if a single dataset contains more than 2,000 files or exceeds 1 TB in size, file upload functionality will be restricted... contact the ScienceDB official email' |
| 2026-10-21 | www.scidb.cn/en/help | registration 'using the username and password you created during registration, or through third-party authentication such as CAS email/CSD account, GitHub account, ORCID and WeChat account' |
| 2026-10-21 | agbigdata.aiijournal.com/EN/Y2025/V7/I2/193 | comparative table: ScienceDB 'no limit on data volume, users can upload datasets of any size' vs Zenodo 50 GB |
Unknown or withheld values (5)
$.free_tier.notesprocess$.per_object_cap.notesnot-extracted$.per_object_cap.statednot-extracted$.retention.notesnot-extracted$.retention.statednot-extracted
ScienceDB (Science Data Bank, CAS / CNIC) — re-verified write path Storage
caveatemailno cardRwrite-gated
| Free tier | No data-volume limit. A 2025 peer-reviewed comparison of Zenodo and ScienceDB in Journal of Agricultural Big Data states it plainly: for ScienceDB the 数据大小 row reads '没有数据量的限制，用户可以上传任意大小的数据集' (no limit on data volume; users may upload datasets of any size), against Zenodo's 'max 50 GB, can be raised to 200 GB in one step'. Datasets above 2,000 files or above 1 TB are merely flagged 'restricted' pending an email — a negotiable ceiling, not a hard one. unit objects; metered on unstated |
|---|---|
| Price | none stated |
| Per-object cap | not stated |
| Retention | not stated |
| Egress | Free, anonymous for public datasets. DOI resolution works (302 to the dataset page). |
| Identity | email email |
| Limits |
write-gated by 'formal review' (形式审核) of upload completeness, accessibility and the data-availability statement — light-touch but still a human queue. Every dataset must have a subject category from the PRC GB/T 13745-2009 classification and at least three keywords. Metadata is mandatory and detailed (data generation procedure, temporal/spatial scope, per-file description). Some submission flows route through a CAS email or CSD account, which for the foreign client means using the GitHub/ORCID button rather than the native form. |
| Binding | write-gated the write is a queue with a person or a rule in it, so the upload rate is set by somebody else |
| Interface | ftp web + ftp |
| Tooling | Explicitly two upload paths: '浏览器在线上传和FTP客户端在线上传' — browser and FTP client. FTP is the one that matters at TB scale and is unusual among research repositories. There is also an Open API and OAI-PMH listed in the help centre. Login accepts a native account or third-party GitHub / ORCID / CAS / WeChat — and the English help page lists GitHub explicitly, which is the route that avoids Chinese identity requirements entirely. |
| Composes with | Unlimited datasets per account with no per-user aggregate cap, so capacity is datasets × a negotiable per-dataset ceiling. FTP ingest means an unattended rclone or lftp job. |
| Reliability | Rule 1 applies: the download path must be discovered from the UI, not guessed. DOI resolution (302 -> /detail?dataSetId=…) is the reliable programmatic handle. |
| Notes | No data-volume limit. A 2025 peer-reviewed comparison of Zenodo and ScienceDB in Journal of Agricultural Big Data states it plainly: for ScienceDB the 数据大小 row reads '没有数据量的限制，用户可以上传任意大小的数据集' (no limit on data volume; users may upload datasets of any size), against Zenodo's 'max 50 GB, can be raised to 200 GB in one step'. Datasets above 2,000 files or above 1 TB are merely flagged 'restricted' pending an email — a negotiable ceiling, not a hard one. |
| Evidence rung | reported One page read moves it to documented. |
| Checked | 2026-10-21 (inherited); stale on 2026-11-04 |
Evidence (5)
| date | source | what it proved |
|---|---|---|
| 2026-10-21 | agbigdata.aiijournal.com/EN/Y2025/V7/I2/193 | opened: Table 1 row 数据大小, Zenodo '最多50 GB，可以一次性增加配额至200 GB' vs ScienceDB '没有数据量的限制，用户可以上传任意大小的数据集'; row 数据提交方式 for ScienceDB '浏览器在线上传和FTP客户端在线上传'; row 登录方式 'ScienceDB平台提供的登录界面；第三方账号，如中国科技云、Github、ORCID、Google账号等'. This is an independent 2025 academic comparison, not vendor copy. |
| 2026-10-21 | www.scidb.cn/en/help | opened: registration via native account or 'third-party authentication such as CAS email/CSD account, GitHub account, ORCID and WeChat account'; four-step flow metadata -> review -> publish. |
| 2026-10-21 | doi.org/10.57760/sciencedb.0142q | PROBED: curl -qsS -o /dev/null -w '%{http_code}' -> 302 with redirect to https://www.scidb.cn/detail?dataSetId=3ac806eb277c4fecb37a55d472b1e82f. DOI routing is live. |
| 2026-10-21 | withheld | PROBED: that detail page read via the broker returns the real record — title, 'Published on 2026-10-21', CSTR/DOI, PACKAGE type, abstract, keywords. A dataset published this week is retrievable, so the platform is active. |
| 2026-10-21 | withheld | PROBED (negative, recorded honestly): /api/v1/dataset/detail?doi=… -> HTTP 404, 9 bytes; /api/download/<doi> -> 404, 9 bytes; /api/download?dataSetId=…&fileId=1 -> 404, 9 bytes. The guessed API paths do not exist. Trust HTTP 200 only after a hash, and here even the status was wrong. |
Unknown or withheld values (8)
$.evidence[3].urlprivate-url$.evidence[4].urlprivate-url$.free_tier.notesprocess$.per_object_cap.notesnot-extracted$.per_object_cap.statednot-extracted$.reliability_notepartly-held$.retention.notesnot-extracted$.retention.statednot-extracted