Preservation & sustainability

What's preserved, and where

Metadata for every record that reaches the Continental Hub's aggregated catalog — species, provenance, citable ids, access decisions — lives in the Hub's own database. Actual file bytes the Hub has taken custody of (mirrored genomes, predicted structures, and anything moved via the node<->Hub transfer feature) live in the Hub's own object storage. Records a node holds but hasn't synced or transferred upward remain that node's own responsibility to preserve — the Hub cannot preserve what it was never given.

Retention

There is no automatic deletion anywhere on this platform — a record, once ingested, persists until an operator explicitly removes it. The Continental Hub operator's own institutional records-retention policy governs in practice.

Backup

Every node's database and object storage are backed up on a schedule matching how much data loss is acceptable — daily is a reasonable starting point for most Spokes. Content checksums, computed for every file the Hub takes custody of, are one way to later verify a restored copy matches what was originally captured.

If the Continental Hub operator ever ceases operations

This is a real risk for any single-institution-hosted infrastructure, stated plainly:

Sustainability model

The platform's published roadmap names a funding mix co-funded between African institutions and governments, the private sector, and international partners. In practice, infrastructure cost (storage, compute) scales with real usage and is not the largest risk — the single largest sustainability risk today is engineering and operational capacity, since the platform is built and maintained by a very small team. A funder conversation about sustainability should center that, not storage bills. Real cost figures should be built from an actual infrastructure quote once a deployment's scale is known, rather than assumed from a general estimate.

See also: Governance, Node operator handbook.