Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Multisite capabilities: separate IR from digital collections sites #76

Closed
elizoller opened this issue Nov 27, 2019 · 5 comments
Closed

Comments

@elizoller
Copy link
Contributor

No description provided.

@elizoller elizoller added this to the MVP Launch milestone Nov 27, 2019
@elizoller elizoller modified the milestones: Devops and Preservation, MVP Beta Jan 23, 2020
@wgilling
Copy link
Contributor

I would like to understand the different options for separating the data and objects into different areas or if those can be kept in a single fedora, and blazegraph (assuming that it would be achieved with Solr filtering) -- as to how each site would only serve up the objects intended for that site.

@elizoller
Copy link
Contributor Author

  • Fedora - will have multiple "roots" for partitioning different sections of the repository (all in a single fedora installation). a set of PRs has been merged to accomplish the "root" piece but it isn't quite fully there because you have to establish the root in fedora first ( at least in my testing). there are also issues to be resolved regarding WebAC (in terms of separating permissions for each of the roots - Meta-Issue: Multi-tenancy in Fedora Islandora/documentation#926)
  • Solr - approach is to have multiple cores in the same solr instance (one for each subsite, and one that is a combined/aggregate index). each site would be configured to index to 2 cores (the individual and the aggregate). then if someone wanted a "parent" site to search all the subsites they would just point their views to the aggregate core (here is the "cookbook" that discusses this approach: Cookbook: Multi-Tenancy with Aggregated Search & Browse Islandora/documentation#1300)
  • Blazegraph - don't know much about how this would work, I assume it'd be something like fedora with separate "roots" in a tree

@elizoller
Copy link
Contributor Author

Basic testing steps:

  • spin up a islandora vagrant (site 1)
  • create 2 fedora roots
  • create an additional solr core for the combined index
  • create an additional solr core for the second site
  • change the existing islandora site (site 1) to point to one of those roots
  • spin up a second drupal with islandora i
  • configure both sites to have to index to two cores - the combined one, and their own separate ones
  • spin up a 3rd drupal with islandora install that doesn't have any content and configure solr to look just at the combined core

@elizoller
Copy link
Contributor Author

Related notes on proof of concept presented to the Islandora Multitenancy Interest Group on 2/11/2020 https://docs.google.com/document/d/1qmCz2VB2g0al0nFA1i1B2rkZYVGHw3UUAft2iEV6r3g/edit

@elizoller elizoller modified the milestones: MVP Beta, Digital Collections Launch Aug 30, 2020
@elizoller
Copy link
Contributor Author

we have chosen to use a true drupal multisite approach for keep and prism. it required deployment of a second RDS database. we are also using the same s3 bucket but with a sub-folder for prism.
deployment for the prism multisite occurred with: https://github.com/asulibraries/claw-playbook/blob/develop/build_prism.yml and asulibraries/islandora-playbook@8b46232

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

2 participants