You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 11 Current »

Table of Contents

Time/Place

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

  • Time: 11:00am Eastern Daylight Time US (UTC-4)
  • ReadyTalk:
    • U.S.A/Canada toll free: 866-740-1260, participant code: 2257295
    • International toll free:
      • http://www.readytalk.com/intl
      • Use the above link and input 2257295 and the country you are calling from to get your  country's toll-free dial-in number
      • Once on the call, enter participant code 2257295
  • IRC:

Attendees

Agenda

Previous Actions

Minutes

Getting a Triple store implemented (Esme)

  • Generic identifier to fixity nodes, now have public URI.
  • Jena and sesame working.
  • We have a SPARQL indexer working, including removes.
  • Soon will be configurable rather than hard coded.
  • Documentation on the way.
Open ended questions/statements
  • Blank nodes (where the RDF doesn't have a proper identifier), causes problems when deleting or selecting.
  • Our API allows for the creation of blank nodes
  • Fixity nodes cause blank nodes.
    • Added generic id to fixity nodes (addresses issue of blank nodes)
    • Fixity nodes will likely not be in the index anyways
  • Attaching metadata directly to an object properties rather than data-stream?
  • Should we disallow blank nodes?
    • Some standards require blank nodes to be constructed.
  • Could use uuid to give them an identifier?
  • If attached to object rather than data stream than blank nodes will be hard to avoid.
  • Possible solution custom indexer class?
How is it working now? Anything remain
  • Repository created/deleted/modified events via JMS.
  • Indexer listens to messages and indexes them in either JENA/SESAME triple store.
  • Deployment is not done but most likely will be a deployable WAR file.
  • Currently a JAR
  • To test now use the included integration tests.
  • Maybe this afternoon, make it load some data from a fixture file?
  • Action: Create ticket for adding configuration element of SPARQL endpoint to be exposed in repository:describe
  • Action: Create a simple deployment wrapper for new message-consumer/indexer project

Large file support federation and events (Nigel).

  • SHA-1 is generated for every file
    • Takes a long time to compute checksum for large files
    • Up to an hour for ~60-gb files
  • It would be nice if modeshape made the alg pluggable
    • We need to look into the ability to swap checksum alg
  • Outcomes: Talk with Mode-shape about why SHA1 was chosen and if something else could suffice?
  • Do documentation discussing how to enable/use the file-connector.
  • Do write a how to guide about what one with large files should do.
  • Will document process/config for of storing checksum in sidecar

Wrap up work on policy driven storage (Osman)

  • Backup currently configuration issues are causing problems
  • Filesystem binary store working
  • Post a question on Modeshape forum to clarify? Need to know if we can back up a repository if it's using a inifinispan binary store.
  • Run time storage policy creation now works with the Rest API, and spring bean file.
  • Wants to add more functionality, maybe next week.
    • Before expanding the policy rule definitions we will wait for specific use cases.

Fedora 3.7 (Ben)

  • Needs Windows Testing
  • JUnit tests assumed the tests would be run in a predictable order (Not the case on Windows). Patch on the way.
  • Java 6 tests should be done today.
  • Go over manual tests, see if we can get someone from 3.6 to do 3.7 manual tests.
  • Lots of tickets around performance. XML processing in particular.
  • Do a 3.7.1 soon after? (To sort the performance issues).
  • Does anyone run windows?

Actions
#trackbackRdf ($trackbackUtils.getContentIdentifier($page) $page.title $trackbackUtils.getPingUrl($page))
  • No labels