# Farbe anpassen, sobald Background gefixt

Archiving

##

jan.leendertse@rz.uni-freiburg.de

Time horizon

Ten year archiving mostly acknowledged

Usual storage systems have shorter life spans

Stability of once selected solutions should be preferred

Provider selection

Duration of archiving should be bound to organisation (not technical systems)

Dedicated responsibility for archiving and service maintenance

Service level requirements

• data integrity
• access control (→ confidentiality)
• authenticity

Authenticity

Keep tracebility of data provenance

Context documentation

Explicit identification of every piece of data

Data sets should work alone

Hashing as technique

As prove that data are unaltered, called integrity

Condition for trustworthiness, that data is not changed, either unintended or intended

Detects data degradation going back to technical issues

• incomplete copies
• bit error on storage level

Example for Windows: Get-FileHash -Algorithm SHA256 "C:\Pfad\zur\Datei" (Supports MD5, SHA1, SHA256, SHA384, SHA512, MACTripleDES, RIPEMD160)

Hashing of research data sets

https://ocfl.io/ (or BagIt)

Data sets should be hashed completely

Process can be automated (via scripting)