Time/Place

This meeting is a hybrid teleconference and slack chat. Anyone is welcome to join...here's the info:

Attendees 

  1. Danny Bernstein
  2. Andrew Woods   
  3. Peter Winckles 
  4. Ben Pennell 
  5. Ben Cail 
  6. David Wilcox 
  7. Jared Whiklo (star) 
  8.  Calvin Xu
  9. Thomas Bernhart 
  10. Daniel Lamb 

Agenda

  1. Announcements/Updates
  2. Planned Breaking Changes (standing topic)
  3. Sprint Issues that could benefit from discussion

  4. Fedora 6 Testing

  5. Your topic here

    type key summary assignee reporter priority status resolution created updated due

    Unable to locate Jira server for this macro. It may be due to Application Link configuration.

Tickets

  1. In Review

    type key summary assignee reporter priority status resolution created updated due

    Unable to locate Jira server for this macro. It may be due to Application Link configuration.

  2. Please squash a bug!

    key summary type created updated due assignee reporter priority status resolution

    Unable to locate Jira server for this macro. It may be due to Application Link configuration.

  3. Tickets resolved this week:

    key summary type created updated due assignee reporter priority status resolution

    Unable to locate Jira server for this macro. It may be due to Application Link configuration.

  4. Tickets created this week:

    key summary type created updated due assignee reporter priority status resolution

    Unable to locate Jira server for this macro. It may be due to Application Link configuration.

Notes

  1. Updates
    1. From Steering Committee. Revisiting membership model for Fedora. Funding has been an issue, but the volunteer "pay what you want" model has not been working. Trying to improve funding model, diversify streams and find other possible benefits for membership.
    2. Leaders/Committers discussion on Alpha/Beta release requirements. Functionality is there, but the documentation is definitely lacking in someways.
  2. Alpha needs.
    1. There are 4 documentation tickets to be done. Peter Winckles and Jared Whiklo will move to work some of those tickets.
    2. Upgrade process is largely done, but requires releases.
    3. Would it be good to define a code freeze point where we stop adding new features and concentrate on cleaning up for alpha release?
    4. How do we release the alpha, are they actually version tagged with alpha. Also good to have an specific alpha docker tag to allow people to access it.
  3. Beta needs
    1. feature completeness as defined by API compliance (at least as compliant as 5.x was), and the features we promised (primarily the search).
    2. Beta release of the migration tooling.
    3. Clear idea of OCFL extensions requirements from the OCFL spec editors to complete the extensions already in use.
    4. Documentation of breaking changes.
    5. Defining the performance testing criteria.
    6. Validation tooling to verify migration completeness.
  4. Performance testing, what types of tests would be good for a next pass? Complex objects, different interactions.
    1. A test where we create an Archival Group with 10,000 members. Autoversioning is turned on and each member is created individually. - from Danny Bernstein
    2. Many members type test, we were able to return approximately 4 million direct children.
    3. Repository has 1,000,000 Archival Groups with binaries in them. How quickly can that content be served. Possible a comparison against a repository with a single AG. - from Ben Cail
    4. Some possible tests - from Ben Pennell
      1. impact of number of versions
        1. Incremental additions to an AG
        2. Touch updates to atomic resource
      2. Containment tests, many children
      3. Membership, many members
        1. Members generated by Direct vs Indirect container
        2. isMemberOf scenarios
      4. Update times for adding a member to container with many members, Direct vs Indirect
      5. Update times for adding a child to container with many child, is there degradation?
      6. Update types for object with many properties
        1. Retrieval times for object with many properties (streaming vs non-streaming serialize)
      7. comparison of commit time for autoversioning vs on demand
      8. Performance impact of commiting transactions with many changes
      9. Ingest of large files, ensure performance doesn’t degrade as files grow or cause OOM
      10. Throughput of many small files
  5. Code freeze
    1. for Fedora as soon as https://github.com/fcrepo/fcrepo/pull/1786 lands
    2. for migration-utils as soon as https://github.com/fcrepo-exts/migration-utils/pull/124 lands
    3. for fcrepo-upgrade-utils as soon as https://github.com/fcrepo-exts/fcrepo-upgrade-utils/pull/24 lands.

Actions

  • Danny Bernstein to ensure that a JIRA exists for scenario of transaction does not succeed completely, but rollback does succeed (see: 2020-08-13- Fedora Tech Meeting)
  • Peter Winckles to create a JIRA to investigate support for optimistic locking and mutable head (see: 2020-08-13- Fedora Tech Meeting)
  • Who:  Clarify in documentation that multiple simultaneous writes to OCFL are not supported
  • migration-utils - translating RELS-EXT/INT into RDF in Fedora 6: is the functionality there? If not, add it.
  • No labels