Difference between revisions of "Options for Repository Software"
From CURATEcamp
Doug Reside (talk | contribs) |
Chris Adams (talk | contribs) (Tidied openstack section) |
||
Line 25: | Line 25: | ||
(Lots of interest in combination of OpenStack SWIFT with solr) | (Lots of interest in combination of OpenStack SWIFT with solr) | ||
− | SWIFT [http://openstack.org] provides | + | SWIFT [http://swift.openstack.org] provides |
− | + | * Open source object storage | |
− | + | * Amazon S3, simple semantics REST based? | |
− | + | * clustering reliability | |
+ | * Large scale production deployments: Rackspace, [https://cloud.sdsc.edu/ San Diego Supercomputing Center] |
Latest revision as of 18:28, 26 July 2012
Tools we are using
- Fedora
- DSpace
- Archivematica
- OpenStack / SWIFT
- LOCKSS
- Ex Libris
- Omeka
- E.R.A.
What is a Repository
Provisional definition: A file system with transactional ingest (let's you know that the file was copied properly) with regular backups and abstracted links to actual file stores managed by a metadata system that may also include other metadata information.
Microservices needed by repository developers
- File storage
- Integrity checking
- Metadata linking
- Public access
- Backups
- Transaction based ingest
(Lots of interest in combination of OpenStack SWIFT with solr)
SWIFT [1] provides
- Open source object storage
- Amazon S3, simple semantics REST based?
- clustering reliability
- Large scale production deployments: Rackspace, San Diego Supercomputing Center