View
6
Download
0
Category
Preview:
Citation preview
Finnish Digital Preservation Service for Cultural Heritage
Mikko Tiainen
IT-Architect Mar. 2015
CSC at a Glance
Owned by Ministry of Education and Culture of Finland
Operates on a non-profit principle
Short history:
– Founded in 1971 as a technical support unit
for Univac 1108
– Connected Finland to the Internet in 1988
– Reorganized as a company, CSC – Scientific
Computing Ltd. in 1993
– Facilities in Espoo, close to Otaniemi campus
(of 15,000 students and 16,000 technology
professionals) and Kajaani
– Staff 269 (March 2015)
– Turnover 2014 ~33 million euros
Enterprise Architecture for NDL
Dis
sem
inati
on p
ackage
Users
Metadata Metadata
DIGITAL
PRESERVATION
Obje
ct
request
and other 3rd
party services
PUBLIC
INTERFACE
Meta
data
Subm
issi
on p
ackage
SUPPORT
SERVICES
STANDARD
PORTFOLIO
External Services
Ontology services
Authentication Service
Integration Platform
Reachability Information
Geographical Information
Online Payment System
LIBRARY, ARCHIVE AND MUSEUM SYSTEMS
Preservation aspects and focus of
NDL’s Digital Preservation Service D
P s
erv
ice
Part
ner
org
an
iza
tio
ns
Long-term
utilization
• Storage device
• Storage media
• Materials & replication management
• Preservation actions
• File formats
• Preservation planning
• Descriptive metadata
• Content knowledge and semantics
Bit-level preservation
Logical preservation
Semantic preservation
• Administrative & technical metadata
NDL DP Specifications
Specifications available at: – http://www.kdk.fi/en/enterprise-architecture
– (mostly in Finnish)
Recommended
file formats
Acceptable file
formats for
transfer
Administrative
and structural
metadata
Descriptive
metadata
BACK-END SYSTEM
Standard portfolio
NDL METS profiles
SUBMISSION INFORMATION PACKAGES (SIP)DIGITAL
PRESERVATION
Digital Preservation – core principles
Multi vendor approach
3 copies with different media Hard disk, HP Proliant SL4540Gen8
Tape1, IBM TS1140 4TB/tape
Tape2, Oracle T10000D 8TB/tape
Dark Archive 2 different tape technologies
Fixity of AIP is verified twice per 5 years AIP checksum SHA-256
IBM Tapes are LBP verified once in a year
Open source based platform Software components are designed to be replaceable
Component life cycle estimations
Hardware
– Hard disk storage 5 years
– Tape drives & medias 5 years
– Tape libraries, 10 years
Software
– Commercial support at least for 5 years
– Open source, maintained and developed until
replaced
Platform overview
Software stack
Integration and control layer ~13000 lines of Python code ”in house”
Archivematica + Gearman
Several OS components e.g. for file format identification and validitation
Middleware Storage software GlusterFS 3.5.2
MongoDB, MySQL
Keepalived
Operating system CentOS 6.6
Configuration mgmt Spacewalk 2.2
Monitoring Opsview community release
– Traditional Nagios plugins + snmp polling & traps
VMWare
High-quality digital
preservation
• Support services
• Maintenance of specifications
• Management
• Cooperation with ATT DP
Development of NDL DP System
Future
Research data preservation
Sertifications ISO 27001 sertification in H1/2015
Data Seal of Approval later on 2015
ISO 16363 sertification on 2017 or later
Software Python 2.6 lifespan, what’s after this
Storage layer battle: etc. GlusterFS, Ceph,…
Hardware Kinetic hard drive techology (Seagate) intergation
into storage layer software
Data Integrity Feature for Oracle LTFS
Recommended