Descriptions of terms used in the Netarchive software. The terms are listed in alphabetic order

Active Bit Preservation

• Active bit preservation means that the bits are actively monitored. Depending on the Quality of Service level this will give different requirements for the configuration of the archive in question. In the Danish installation of NetarchiveSuite we have the following configuration:

  • Two bitarchive replicas representing copies of the data objects to be preserved (in order to be able to recover a bit error in a data object).
  • The two bitarchive replicas are placed on separate physical locations (in order to have a copy in case of fire, floods, earth quake etc.).
  • One checksum replica giving the number of three voting replicas (which is minimum number for the voting to make sense).
  To have a higher quality level for the active bit preservation more requirements will be needed.

• Please also refer to term Bitarchive Replica.

• Please also refer to term Checksum Replica.

• Please also refer to term Quality of Service Level.

• Please also refer to figure under term Repository.

ArcRepository

• Please refer to term Repository.

Bitarchive

• Very broad term that should be use with care since it can be used on many different levels. In princip a bitarchive is an archive of bits stored on some media.

• If possible it is preferable to use the terms like repository, bitarchive replica and bitarchive instance.

Bitarchive Application

• Please refer to term Bitarchive Instance.

Bitarchive Instance

• A number of bitarchive instances are the representation of a bitarchive replica. A bitarchive instance is an instance of the bitarchive software which can upload, run batch-jobs, correct or get-file/-record information for the represented part of the bitarchive replica. In technical terms it is the instance that the NetarchiveSuite bitarchive-application is running on. Note that this is not bound to be on one machine.

• Please also refer to figure under term Repository.

• Please also refer to term Bitarchive Replica.

Bitarchive Replca

• A bitarchive replica is a replica in a repository containing the repository files as data objects.

• Please also refer to figure under term Repository.

• Please also refer to term Replica.

Checksum Instance

• A checksum instance is the representation of a checksum replica. A checksum instance is an instance of the checksum software which can upload, get checksums, get file list or correct checksums on the checksum replica. In technical terms it is the instance that the NetarchiveSuite checksum-application is running on (according to assignement B2 described under ["Development"].

• Please also refer to figure under term Repository.

• Please also refer to term Checksum Replica.

Checksum Replica

• A checksum replica is an replica in a repository containing checksums of the repository files.

• Please also refer to figure under term Repository.

• Please also refer to term Replica.

Development release

• Internal development release. Can be used for experiments.

Event harvest

• A selective harvest made at special events.

Full harvest

• See Snapshot harvest.

Replica

• A replica is either a bitarchive replica or a checksum replica, i.e. either a replica with a copy of repository data objects or a replica with a copy of checksums of the Repository data objects. A replica will have one vote in voting on which replica is having the correct copy, as part of the active bit preservation.

• Please also refer to figure under term Repository.

• Please also refer to term Bitarchive Replica.

• Please also refer to term Checksum Replica.

Partial harvest

• See Selective harvest.

Physical Location

• A physical location is the geographical address or abbreviation indicating the geographical placement of maschines for NetarchiveSuite applications.

Quality of Service Level (of Active Bit Preservation)

• In the NetarchiveSuite the Quality of Service level can (or can in future) be regulated via different parameters, for instance:

  • Number of checksum algorithms used (future)
  • Distance between physical locations
  • Number of checksum replicas (future)
  • Number of bitarchive replicas
  • Time interval between checksum checks (future)
  • Quality of hardware

Release

• See Stable Release or Development release

Repository

• The term Repository (in NetarchiveSuite also called ArcRepository) is used for the full repository of archived data objects where the repository is the administration of replicas in the repository. The archive repository can consist of one or many object replicas, where each replica either holds the data objects themselves or the checksum of the data object.

• The below figure shows a repository with three replicas.

• attachment:LocationDefinition.gif

• Please also refer to term Bitarchive Replica.

• Please also refer to term Bitarchive Instance.

• Please also refer to term Checksum Replica.

• Please also refer to term Checksum Instance.

• Please also refer to term Replica.

Repository with Active Bit Preservation

• A repository that fulfills the requirements listed in the explanation of the term Active Bit Preservation.

• Please also refer to term Repository.

Segment

• Abstract concept that covers both mount on Linux and drive on Windows

Selective harvest

• A harvest of a few, selected sites that can be done repeatedly.

Snapshot harvest

• A limited-depth harvest of all domains known in the system. A snapshot harvest can only be run once, but can be based on an earlier snapshot harvest, harvesting only those domains that were not fully harvested in the previous harvest.

Stable Release

• Stable release of NetarchiveSuite SW.