Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  • DPN content and services are distributed and federated.

  • DPN supports succession and brightening of content.

  • Implementations are de-coupled implementations and architecturally distinct, as practicable, but the communication methodology is shared, resilient, and redundant.

  • DPN objects are preserved by all Nodes and support DPN preservation functions.

  • DPN has decoupled inter-node communication channels for

    • content transfer

    • process control.

  • Communication between nodes is not dependent upon other nodes.

Production Deliverables - Jan

...

2016

  • There is no expectation that we’ll have signed SLA’s before launch

    • DPN will have agreements to allow for Succession of content in the event an archive can no longer perform its function as an archive.

  • SLA’s with Ingest Nodes/Administrative Node

  • SLA documents between Ingest Nodes & Depositor shared and with legal staff

    • We will require in the SLA that one of the OTHER Administrative Nodes will become the Administrative node for the failed Node.

      • Act as the restorer of content for clients of the former administrative node

      • In the event of a Succession occurrence, all replicating nodes will recognize the new successor and act in accordance with prior agreements held by the former archive

    • Replication and update will take time - We are aware of this, might need to put something in the SLA recognizing this fact

    • SLA will state that: The communication layer is shared, but the repository layer is not

  • Depositor will be able to give us stuff and we can put it into storage

  •  There will be no expectation of global (DPN Level) fixity checking at launch

    • Initial fixity checking will occur at ingest

    • Each Node will check fixity according to their local policy

  • Some cursory reporting available to Depositor

    • Notification of ingest and replication will be provided by the Ingest Node to the depositor

  • Replication of content from the Administrative Node to all Replicating Nodes

    • Make sure we know how many Replicating Nodes will be storing the content

    • Make sure we know when the Ingest Node will store the ingested content and when they won’t

    • Make sure we have a clear idea of what kind of storage is available at each

  • The Ability to recover content to the depositor will be supported via the Ingest/Administrative node

  • Maintain a registry for objects, create transfer records for replicating nodes, update status of an existing object

    • Track stored status of replicated transfer

  • Agreement on the bag size limits that can be replicated across the Nodes

    • 250 Gig bags are the uppermost limit for this release

  • Bags will be validated upon receipt by the Ingesting Node and the Replicating Node

    • Validation means:

      • We will validate all of the files are present and the checksums match the manifest

      • The structure of the bag will be validated according to the DPN specs

  • In-Person Post-Mortem/Planning for Phase II - July 16th & 17th

...