fshy

fshy / Storage

National Research Platform

nrp.ai · Storage · 1 record · JSON

National Research Platform (NRP / Nautilus, US) — free Ceph S3 to any namespace holder, multi-site Storage

caveatinstitutionno cardRper-object-cap

nrp.ai/documentation/userdocs/storage/ceph-s3/ · academic · storage/national-research-platform-free-ceph-s3-to-any-namespace · JSON · compare

Free tierNo published byte quota. The docs are explicit that the storage is donated: 'The Nautilus Ceph storage cluster can be accessed via S3 protocol. It uses our own storage, which is free for our users and is not part of Amazon or any commercial cloud.' Quota is whatever the namespace is granted by admins; CephFS carries a documented 16 TB per-file cap.
unit TB; metered on stored-bytes
Pricenone stated
Per-object capper-object-cap: CephFS has 'a per-file size limit of 16 TB (16000000000000 bytes)
RetentionRetention: 'Any storage volume that was not accessed for 6 months can be purged without notification' — this is the hard floor on it as an archive (180 days)
EgressFree and unmetered ('free for our users'), served from university-hosted Ceph RGW pods across west/central/east.
Identityinstitution derived from the record's prose, not from an identity field, and no signup was executed: CILogon account, then namespace admin status via 'eligible faculty members and researchers can request to become namespace admin' (at LBNL this is an email). academic
Limits
  • retention-clamped Retention: 'Any storage volume that was not accessed for 6 months source
  • per-object-cap per-object-cap: CephFS has 'a per-file size limit of 16 TB source
per-object-cap: CephFS has 'a per-file size limit of 16 TB'; S3 per-object cap not stated. Retention: 'Any storage volume that was not accessed for 6 months can be purged without notification' — this is the hard floor on it as an archive. request-rate: not published; pods are policed instead ('a user can not submit more than 4 pods' outside the utilisation rules; interactive pods 'destroyed in 6 hours'). Gate: CILogon account, then namespace admin status via 'eligible faculty members and researchers can request to become namespace admin' (at LBNL this is an email).
Bindingper-object-cap the per-object cap sets the shard size and therefore the object count; it bounds the total more tightly than any headline quota
Interfaces3, rclone, cli s3
Toolingrclone/aws cli/restic against https://s3-west|s3-central|s3-east.nrp-nautilus.io. Credentials from the User Portal S3 Tokens page (nrp.ai/s3token/) for public Ceph pools. Inside-cluster and outside-cluster endpoints documented.
Composes withSix independent Ceph pools (west, central, east, haosu, tide, ha) each with their own S3 endpoint. 'public Ceph pools' give self-service tokens — so any CILogon-authenticated researcher can mint S3 credentials without an admin. That is the closest thing to self-service in this whole niche.
Reliability6-month untouched-data purge is documented and unannounced; namespaces under-utilising requests are closed. Treat as hot storage only, never archive.
NotesNo published byte quota. The docs are explicit that the storage is donated: 'The Nautilus Ceph storage cluster can be accessed via S3 protocol. It uses our own storage, which is free for our users and is not part of Amazon or any commercial cloud.' Quota is whatever the namespace is granted by admins; CephFS carries a documented 16 TB per-file cap.
Evidence rungreported One page read moves it to documented.
Checked2026-10-21 (inherited); stale on 2026-11-04
Parsed numbers
  • $.limits[1].value limit:per-object-cap 16.0 TB
Evidence (4)
datesourcewhat it proved
2026-10-21nrp.ai/documentation/userdocs/storage/ceph-s3/'It uses our own storage, which is free for our users and is not part of Amazon or any commercial cloud'; six pool endpoints; 'You can get your credentials (key and secret) in the User Portal ... for public Ceph pools'
2026-10-21nrp.ai/documentation/userdocs/start/policies/'Any storage volume that was not accessed for 6 months can be purged without notification'; pod/utilisation enforcement rules
2026-10-21nrp.ai/documentation/userdocs/storage/intro/'There is a per-file size limit of 16 TB in CephFS'
2026-10-21s3-west.nrp-nautilus.io/probe: curl -> 200 (187 bytes Ceph RGW ListAllMyBucketsResult); s3-central and s3-east identical; s3-haosu/s3-tide/s3-ha documented as further endpoints. Endpoints live and answering S3.
Unknown or withheld values (3)
  • $.free_tier.notes process
  • $.per_object_cap.notes not-extracted
  • $.retention.notes not-extracted