Architecture straw-man document must be drafted and circulated to WP3 partners. Main architecture concepts must be discussed and clarified.
Infrastructure partners (GWDG, CERN, PSNC, GRNET) are going to meet and converge viewpoints on the basic architecture design.
It was agreed to deploy a central Up2U infrastructure ie. "playground" including:
- storage and File Sync&Share platform with OCM
- VM infrastructure for LMS platform and apps with LTI
- possibly storage and repository for metadata/content/learning-path connected to eduOER.
The central Up2U infrastructure should be able to interconnect/federate with national infrastructure. Software may be packaged and ported to national infrastructure (design goal).
Taking into account the rapid prototyping approach, off-the-shelf, open source software platforms are recommended in the first phase:
- possibly DSpace repository
The central Up2U infrastructure playground MUST have:
- eduGAIN supported federated access, SAML2 based SSO
- local LDAP sign-on for homeless
- pluggable group management function
- multi-tenancy on LMS (accommodate multiple schools)
- ThinCan and/or LTI support on LMS
- OCM support for ownCloud federations
- all repositories connected via eduOER (gateway)
Infrastructure partners should discuss ASAP what central components can be hosted where, who will be responsible for what, and how those components can seamlessly interoperate.
The central infrastructure must have at least a pilot and a production instance. Preferably: testing, pilot and production instances.
Global CDN networks (Google, Akamai, etc) already have cashing points inside the NREN networks. Usually, there is no problem with accessing global content.
A CDN might make sense in case of content sitting in distributed smaller repositories across institutions. PSNC will "pilot and test" a CDN infrastructure connecting pilot countries and evaluate network/streaming performance in different scenarios. Partners willing to participate in the initial CDN pilot should contact PSNC.
The Deliverable D3.1 in M6 will be the analysis of network services requirements of schools. This is an on-line document, perhaps the survey results about schools' connectivity and edoroam availability in the initial pilot countries (how schools are connected, by NREN, by ISP, if eduroam is available at schools, libraries, public places????)
Our project is not for rolling out eduroam to schools however, we have to investigate and prioritize those schools where eduroam is already available during piloting.
The same applies to federated access. Our project is not to roll out IdPs to schools however, those schools with their own IdPs (or centralized IdP provided by the NREN of the country) should be prioritized.