...
- Added tests and made changes to validation
- Suggestion to create fresh fork due to recent, broad f4 package updates
- Andrew Woods to review Osman's latest code Action:
...
- Finished velocity template
- May need triplestore for navigation
- Question: what functionality is needed for tuque?
- Currently Tuque relies on itql
- count query
- walk query
- Which sqarql endpoint would be used? 1.0? 1.1?
- Answer: 1.1
- Federation: suggestion to start by using modeshape filesystem connector
- Nigel, Yuqing, and Emse to schedule an itql/sqarql meeting
Wednesday 2013-08-14
Esme
- Posted email to ff-tech re:rdf refactoring
- Andrew Woods to extend email to fedora-dev list
- Updating URIs in code
Jay
- Basic filesystem federation working
- Working on detecting new content in filesystem
- Note that Nigel is working on same issue
Yuqing
- Able to see some items in islandora
- getObject is presenting some issues
- Nigel offered to help, maybe pair program?
Nigel
- Filesystem connector does not handle large files
- Issue may be with what f4 is doing
- Working on events in modeshape filesystem connector
- Need profiler
Osman
- Updating policy-driven-storage code based comments
- Should be ready for commit today
Thursday 2013-08-15
Friday 2013-08-16
Esme
- Looking into various triplestores
- Sesame makes good sense
- Allows for bundling the app
- Looking into JMS/Solr/Triplestore/indexing project
Yuqing
- Can read Fedora object in Islandora
- Still working on encoding path
- Andrew to pass along uuid request details
Osman
- Getting up to speed on clustering
- Will have a workstation cluster next week
- Updating policy-driven documentation
- Will begin working on investigating backup
Andrew
- Upgraded to Modeshape 3.4
Monday 2013-08-19
Nigel
- Working with modeshape filesystem
- How are properties managed by mode?
- Ingesting large objects through curl is quite fast
- Reading takes significant time
Osman
- Working backup task
- Will be adding documentation shortly
- Added documentation for storage policy
Yuqing
- Working on generating DC
- Collection content model is not working, obsolete
- May want to create a hasModel property
- Encoding Fedora path for Drupal is not working
Jay
- Working on creating a node via REST API that is associated with federated content
- Wanting to create federated node via API
Esme
- Working on getting SPARQL endpoint
- Issue getting integration tests working
- fcrepo-jms-indexer-puggable
- mvn verify shows the errors
- Content needs to be created for ontology
Tuesday 2013-08-20
Esme
- Ontology prepared and published
- Getting back to SPARQL
- Working through integration test issues
- Integration server is not getting started
- Andrew Woods to investigate IT issue
Yuqing
- Dublin Core creation works
- Will add commit links to "Finished" tickets
- Working add/get relationships
Nigel
- Working with Yuqing on Islandora debugging
- Largefiles slowdown confirmed to be due to hash-generation
- Look into buffering hashing stream
- Need to document need for stress test
Osman
- Working backup
- Running into issues
Wednesday 2013-08-21
Osman
- Ran backup tests, seems to work
- From backup branch
- Using infinispan cache (filesystem)
- Using strings for data
- Will try to use Modeshape BinaryStore
Nigel
- Large files
- Conclusion, not much we can do to help performance
- 6-GB file takes ~1.5 min
- 64-GB ran more than 45min
- SHA-1 needs to be recalculated on reboot/cache-invalidation
Yuqing
- Tickets that are "finished" but have not code references
- Nigel to help Yuqing with Git workflow
Jay
- Working on relationship between objects
Esme
- Integration tests working
- Moving towards Fuseki indexer
- Working through issues related to updating the triplestore based on repository updates