InvenioRDM platform
data.caltech.edu · Storage · 2 records · JSON
- CaltechDATA (InvenioRDM, California Institute of Technology) Storagefails: identityinstitutionno cardR
- InvenioRDM platform class — CaltechDATA, TU Graz, TU Wien, TalTech, KTH, Bamberg, MSD-Live Storagefails: identityinstitutionno cardR
CaltechDATA (InvenioRDM, California Institute of Technology) Storage
fails: identityinstitutionno cardR
| Free tier | 1 TB of free storage per user, announced 2026-06-26, 'made possible through a National Science Foundation Campus Cyberinfrastructure grant' and backed by the Open Storage Network. Additional storage $200/TB one-time. 550 files max per record. unit objects; metered on stored-bytes |
|---|---|
| Price | none stated |
| Per-object cap | not stated |
| Retention | not stated |
| Egress | Unmetered — no egress terms; downloads stream straight from data.caltech.edu (I saw no redirect). Content-Security-Policy references caltech2.osn.mghpcc.org, i.e. storage sits on the Open Storage Network, an NSF-funded academic fabric |
| Identity | institution derived from the record's prose, not from an identity field, and no signup was executed: e $200/TB one-time. 550 files max per record. Anonymous API reads work and are generous (X-RateLimit-Limit: 500 observed on /api/records). The 1 TB allocation is explicitly funded |
| Binding | none the quota is real and unreachable: identity is the binding constraint, and it is not one of the six limits kinds |
| Interface | rest rest |
| Tooling | InvenioRDM REST API (/api/records, files under /api/records/<id>/files/<key>/content); purpose-built client caltechlibrary/caltechdata_api (github.com/caltechlibrary/caltechdata_api, 13 stars); the whole instance's source is public at caltechlibrary/caltechdata (11 stars) so the quota config is auditable |
| Composes with | 1 TB is the largest single free allocation I found in this niche and it is per user, not per record. Records themselves are not account-capped, so 1 TB is a floor rather than a ceiling. Institutional deployments like this are exactly the 'more generous than Zenodo' pattern the brief predicted. |
| Reliability | Anonymous API reads work and are generous (X-RateLimit-Limit: 500 observed on /api/records). The 1 TB allocation is explicitly funded by a time-bounded NSF grant, so the number has a funding horizon attached. |
| Notes | 1 TB of free storage per user, announced 2026-06-26, 'made possible through a National Science Foundation Campus Cyberinfrastructure grant' and backed by the Open Storage Network. Additional storage $200/TB one-time. 550 files max per 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 | libanswers.caltech.edu/faq/163212 | read live (Last Updated Jun 29, 2026): 'All users with an access.caltech account get 1 TB of free storage through CaltechDATA. Additional storage requires a one-time charge of $200 / TB ... Individual records can include a maximum of 550 files.' |
| 2026-10-21 | library.caltech.edu/blogs/news/caltechdata-expanded-storage | read live, dated 06/26/2026: 'All researchers at Caltech can now upload up to 1 TB of data for free ... possible due to the Open Storage Network' |
| 2026-10-21 | data.caltech.edu/api/records?size=1 | probe: GET -> 200 with X-RateLimit-Limit: 500; GET /api/records/mhmjd-4pr39/files/... content endpoint -> 200, files.enabled: true, per-file links resolve |
More fields (1)
{
"dead_end": {
"failure_class": "quota",
"why": "1 TB of free storage per user, announced 2026-06-26, 'made possible through a National Science Foundation Campus Cyberinfrastructure grant' and backed by the Open Storage Network. Additional storage $200/TB one-time. 550 files max per record.",
"retry_when": null,
"first_recorded": "2026-10-21"
}
}Unknown or withheld values (6)
$.extras.dead_end.retry_whenprocess$.free_tier.notesprocess$.per_object_cap.notesnot-extracted$.per_object_cap.statednot-extracted$.retention.notesnot-extracted$.retention.statednot-extracted
InvenioRDM platform class — CaltechDATA, TU Graz, TU Wien, TalTech, KTH, Bamberg, MSD-Live Storage
fails: identityinstitutionno cardR
| Free tier | Per-record, NOT per-account: RDM_FILES_DEFAULT_QUOTA_SIZE = 10 GB and RDM_FILES_DEFAULT_MAX_FILE_SIZE = 10 GB, with APP_RDM_DEPOSIT_FORM_QUOTA maxFiles = 100. Records are unlimited. This is the structural inverse of Dataverse: InvenioRDM ships a real quota and it is per-object, so total space is unbounded while each record is capped at 10 GB. unit GB; metered on unstated |
|---|---|
| Price | none stated |
| Per-object cap | 10 GB/file (10000000000 bytes) |
| Retention | not stated |
| Egress | Cloudflare-fronted on every instance checked. |
| Identity | institution institution |
| Limits |
per-object-cap 10 GB/record and 10 GB/file, 100 files/record. No per-account cap exists in the platform, so no instance can report one. |
| Binding | none the quota is real and unreachable: identity is the binding constraint, and it is not one of the six limits kinds |
| Interface | rest, rclone, cli REST (/api/records), InvenioRDM REST API v1 |
| Tooling | curl / invenio-cli / rclone via REST |
| Composes with | Excellent for the archive workload if registration were open, because 10 GB x unlimited records is unbounded in aggregate. It is not open on any instance tested. |
| Reliability | All seven instances 404 on /signup/, which in InvenioRDM means SECURITY_REGISTERABLE is off; only Zenodo (already counted elsewhere) and the CERN demo return anything other than 404. |
| Notes | Per-record, NOT per-account: RDM_FILES_DEFAULT_QUOTA_SIZE = 10 GB and RDM_FILES_DEFAULT_MAX_FILE_SIZE = 10 GB, with APP_RDM_DEPOSIT_FORM_QUOTA maxFiles = 100. Records are unlimited. This is the structural inverse of Dataverse: InvenioRDM ships a real quota and it is per-object, so total space is unbounded while each record is capped at 10 GB. |
| Evidence rung | reported One page read moves it to documented. |
| Checked | 2026-10-21 (inherited); stale on 2026-11-04 |
| Parsed numbers |
|
Evidence (10)
| date | source | what it proved |
|---|---|---|
| 2026-10-21 | github.com/inveniosoftware/invenio-rdm-records/blob/master/i | RDM_FILES_DEFAULT_QUOTA_SIZE = 10 * (10**9); RDM_FILES_DEFAULT_MAX_FILE_SIZE = 10 * (10**9); RDM_FILES_DEFAULT_MAX_ADDITIONAL_QUOTA_SIZE = 0 |
| 2026-10-21 | github.com/inveniosoftware/invenio-app-rdm/blob/master/inven | APP_RDM_DEPOSIT_FORM_QUOTA = {'maxFiles': 100, 'maxStorage': RDM_FILES_DEFAULT_QUOTA_SIZE} |
| 2026-10-21 | data.caltech.edu/api/records?size=1 | 200, records served; /signup/ — 404 |
| 2026-10-21 | repository.tugraz.at/api/records?size=1 | 200; /signup/ — 404 |
| 2026-10-21 | researchdata.tuwien.ac.at/api/records?size=1 | 200; /signup/ — 404 |
| 2026-10-21 | data.taltech.ee/api/records?size=1 | 200; /signup/ — 404 |
| 2026-10-21 | datarepository.kth.se/api/records?size=1 | 200; /signup/ — 404 |
| 2026-10-21 | fd-repo.uni-bamberg.de/api/records?size=1 | 200; /signup/ — 404 |
| 2026-10-21 | data.msdlive.org/api/records?size=1 | 200; /signup/ — 502 |
| 2026-10-21 | inveniordm.web.cern.ch/signup | 308 (registration enabled on the demo); https://zenodo.org/signup — 308 |
More fields (1)
{
"dead_end": {
"failure_class": "quota",
"why": "Per-record, NOT per-account: RDM_FILES_DEFAULT_QUOTA_SIZE = 10 GB and RDM_FILES_DEFAULT_MAX_FILE_SIZE = 10 GB, with APP_RDM_DEPOSIT_FORM_QUOTA maxFiles = 100. Records are unlimited. This is the structural inverse of Dataverse: InvenioRDM ships a real quota and it is per-object, so total space is unbounded while each record is capped at 10 GB.",
"retry_when": null,
"first_recorded": "2026-10-21"
}
}Unknown or withheld values (5)
$.extras.dead_end.retry_whenprocess$.free_tier.notesprocess$.per_object_cap.notesnot-extracted$.retention.notesnot-extracted$.retention.statednot-extracted