4 The OAIS model is the common language of digital preservation. Libraries, archives, research data repositories and space agencies use the Open Archival Information System reference model to describe how a long-term archive should work, from accepting content to keeping it usable for future users. It is a conceptual model, not a product specification, yet it has direct consequences for storage. The OAIS Archival Storage function describes responsibilities, such as error checking, media replacement and disaster recovery, that the storage platform must support. This article explains the OAIS model in practical terms, how its information packages relate to storage and what the Archival Storage function asks of storage platforms. For the broader picture, see our hub on digital preservation storage. What OAIS is OAIS was developed by the Consultative Committee for Space Data Systems and is published as an international standard, ISO 14721. It defines an archive as an organization of people and systems that has accepted responsibility to preserve information and make it available for a designated community, meaning the users the archive serves, such as researchers, citizens or scientists. The model describes: Responsibilities an archive must meet, such as ensuring information is independently understandable to its designated community. Information packages that move into, through and out of the archive. Functional entities that perform the work of the archive. Because OAIS is conceptual, it can be implemented with many different systems. Most preservation platforms describe themselves in OAIS terms. Information packages OAIS describes three kinds of information package: Submission Information Package (SIP): what a producer sends to the archive, such as a set of files with descriptive information. Archival Information Package (AIP): what the archive stores for the long term, combining content with the preservation metadata needed to understand, verify and manage it. Dissemination Information Package (DIP): what the archive delivers to users, which may be a subset or transformed version of the AIP. For storage, the AIP is the unit that matters most. AIPs must be stored durably, verified regularly and moved to new storage when technology changes, all without loss. The functional entities OAIS defines six functional entities: Ingest receives SIPs, validates them and creates AIPs. Archival Storage stores, maintains and retrieves AIPs. Data Management maintains descriptive information and administrative data used to find and manage content. Administration manages day-to-day operations, agreements with producers and policies. Preservation Planning monitors technology and the designated community to keep content usable, recommending migrations or other actions. Access helps users find content and delivers DIPs. Storage teams work most closely with Archival Storage, but decisions in Preservation Planning and Administration, such as how many copies to keep and when to refresh technology, shape storage design. What Archival Storage asks of storage The OAIS model describes the Archival Storage function in terms of several sub-functions. In practical terms, these translate into storage requirements. Receive data Archival Storage accepts AIPs from Ingest and places them in permanent storage. Storage must accept large and small packages reliably, confirm successful storage and support the throughput of ingest campaigns such as digitization projects or web crawls. Manage storage hierarchy The archive places AIPs on appropriate storage according to policy, access needs and cost, and may use multiple tiers. Storage platforms should support policy-based placement and tiering, so frequently accessed content can sit on online storage while other copies use lower-cost media. See hot storage vs cold storage. Replace media Storage media degrade and become obsolete. The archive must copy AIPs to new media before old media fail, without altering content. Storage that allows hardware to be refreshed node by node, migrating data automatically in the background, makes this routine. The scality.com post on object storage hardware refresh describes the approach. Error checking The archive must detect and correct errors in stored AIPs. This is where fixity checking comes in: computing checksums and comparing them to known values, and repairing damaged copies from good ones. Storage platforms that store checksums and verify data in the background add an extra layer beneath the preservation system’s own checks. See how archives run fixity checks at petabyte scale. Disaster recovery The archive must be able to recover from disasters, which in practice means multiple copies in separate locations, documented recovery procedures and regular tests. See how many copies digital preservation needs, and where. Provide data Archival Storage retrieves AIPs for Access and for preservation actions. Storage must deliver content reliably and at reasonable speed, including for bulk operations such as format migrations. Storage characteristics that support OAIS Durability: erasure coding or replication within a site, plus copies across sites. See erasure coding vs replication. Integrity verification: checksums at multiple levels and regular verification. Immutability: protection against accidental or malicious changes to AIPs. See S3 object lock: immutability and WORM. Technology independence: open interfaces such as S3 and standard packaging formats, so AIPs can move between systems. Scalability: growth without disruptive migrations. Auditability: logs of storage operations for trustworthy repository certification. Location control: the ability to keep copies in specified places, important for national institutions. Mapping OAIS to a real deployment In a typical implementation, a preservation system or repository platform handles Ingest, Data Management, Access and much of Preservation Planning, while one or more storage platforms provide Archival Storage. For example, the preservation system receives a SIP from a digitization project, validates files, computes checksums, wraps the content and metadata into an AIP and writes it to two or three storage locations. It records where each copy lives in its own database. On a schedule, it asks storage to return each copy so checksums can be recomputed and compared. When a storage platform reaches end of life, the preservation system or the storage platform itself moves AIPs to the replacement and verifies them. The division of labor matters. If the storage platform verifies integrity in the background and refreshes hardware in place, the preservation system has less to do, and the risk of human error during migrations falls. If the storage platform is a simple file share, the preservation system must take on more of the Archival Storage responsibilities itself. Documenting who does what, and how it is evidenced, is essential for audits. Common gaps Archives assessing themselves against OAIS often find the same storage-related gaps: copies held in the same building, fixity checks that are scheduled but not completed, media refresh plans that exist only on paper, disaster recovery procedures never tested and storage operations without audit logs. Each of these is fixable with the right platform and processes, but they are worth checking before an external audit does. Packaging formats and storage Preservation systems package AIPs in various ways, often using standards such as BagIt for file packaging, METS for structural metadata and PREMIS for preservation metadata. Some store AIPs as single archive files, others as directories of files. Packaging affects storage in practical ways: many small files stress metadata performance and fixity checking, while very large packages make partial access slower. Choose packaging and storage together. OAIS and certification Trustworthy repository certification, such as CoreTrustSeal or audits based on ISO 16363, builds on OAIS concepts. Auditors look for evidence that storage meets preservation responsibilities: documented copies and locations, fixity results, media refresh plans, disaster recovery tests and access controls. Storage platforms that produce clear reports and logs make certification easier. Applying OAIS in different regions OAIS is used worldwide, from national archives in Europe and Asia to research data centers and space agencies. Regional requirements, such as legal deposit laws, public records acts and data protection rules, add obligations on top of the model, especially about where copies are kept and who may access sensitive records. National institutions commonly implement OAIS with storage located entirely within their own jurisdiction. Checklist: OAIS model storage Identify where AIPs are stored and how many copies exist. Support policy-based placement across storage tiers. Plan media and hardware refresh without content loss. Compute, store and verify checksums regularly. Repair damaged copies automatically from good ones. Keep copies in separate locations and test disaster recovery. Use open interfaces and packaging standards. Protect AIPs with immutability and access control. Keep audit logs and reports for certification. Putting it together The OAIS model asks storage to do what preservation requires: keep AIPs safe and verifiable, place them sensibly across tiers, replace media before it fails, detect and repair errors, recover from disasters and deliver content when needed. A storage platform that scales out, verifies integrity, supports immutability, refreshes in place and exposes open interfaces makes those responsibilities practical. Design storage with the Archival Storage function in mind, and it will support both daily preservation work and the scrutiny of certification. Frequently asked questions What does OAIS stand for? Open Archival Information System, a reference model for long-term digital archives published as ISO 14721. What is an AIP? An Archival Information Package, the unit the archive stores for the long term, containing content and preservation metadata. Is OAIS a software product? No. It is a conceptual model that many systems implement in different ways. What does OAIS require from storage? Its Archival Storage function covers receiving data, managing storage hierarchy, replacing media, error checking, disaster recovery and providing data. How does OAIS relate to certification? Certification schemes such as CoreTrustSeal and ISO 16363 audits build on OAIS concepts and look for evidence that storage meets preservation responsibilities. Further reading See digital preservation storage, digital preservation copies, fixity checks at scale, research data management storage and data archiving best practices.
