Log of the #duraspace-ff channel on chat.freenode.net

Using timezone: Eastern Standard Time
* ksclarke leaves01:24
* longshou leaves03:57
<pivotal-bot>Esme Cowles started "Investigate whether or not metadata records can be updated under read-only file system federation" https://www.pivotaltracker.com/story/show/7455373807:33
Esme Cowles started "Verify that creating metadata records using writable file system federation works" https://www.pivotaltracker.com/story/show/74553610
* dwilcox joins09:23
* gregjansen joins09:26
* mohamed joins09:28
* ksclarke joins09:46
* escowles joins09:56
<pivotal-bot>Esme Cowles added comment: "Metadata updates on an object in a read-only federated fileystem fail with a 500 Internal Server Error and t..." https://www.pivotaltracker.com/story/show/7455373809:59
Esme Cowles finished "Investigate whether or not metadata records can be updated under read-only file system federation" https://www.pivotaltracker.com/story/show/7455373810:00
Andrew Woods added comment: "Retesting this, I am not seeing any errors. The issue may be with the ':' in your filename: "aic:ds_master"..." https://www.pivotaltracker.com/story/show/6984886010:05
* gregjansen leaves
* gregjansen joins10:10
<pivotal-bot>Andrew Woods added comment: "I am not seeing any new commits on PR-394. Am I missing sum" https://www.pivotaltracker.com/story/show/7194631410:12
Esme Cowles added comment: "After enabling read/write operation, I can perform metadata updates (e.g., 10K SPARQL Update requests) witho..." https://www.pivotaltracker.com/story/show/74553610
Andrew Woods accepted "Investigate whether or not metadata records can be updated under read-only file system federation" https://www.pivotaltracker.com/story/show/74553738
Esme Cowles finished "Verify that creating metadata records using writable file system federation works" https://www.pivotaltracker.com/story/show/7455361010:13
<awoods>escowles: For https://www.pivotaltracker.com/story/show/74553610 ...10:16
<pivotal-bot>feature: Verify that creating metadata records using writable file system federation works (finished) / owner: Esme Cowles
<awoods>escowles: did you have to update the FedoraFileSystemConnector to allow writes?10:17
<escowles>awoods: yes, it currently throws an exception if it's not read-only
<awoods>escowles: yes, exactly.
<escowles>awoods: i have a branch with that one change (plus removing the test that checks for that exception being thrown)10:18
<awoods>escowles: so the above ticket demonstrated that you were not limited to 22 metadata updates?
escowles: and that indeed the errors you saw months before relate only to adding new files.10:19
<escowles>awoods: adding files or creating objects/directories -- and i think the 22 datastreams or objects is per-directory -- i'm writing a script right now to try a three-level hierarchy to be sure
<pivotal-bot>Andrew Woods accepted "Verify that creating metadata records using writable file system federation works" https://www.pivotaltracker.com/story/show/7455361010:20
<escowles>awoods: i can create a 21/21/21 object hierarchy with no problems -- but trying to add the 22nd object at any level fails with a NPE at session.save()10:28
<awoods>escowles: and no transactions are involved, correct?10:29
<escowles>awoods: correct -- no transactions
<awoods>escowles: that is odd, but should be a very trackable bug (likely in the modeshape layer)10:30
escowles: you may also want to try the straight modeshape filesystemconnector to see if the FedoraFileSystemConnector is the source of the issue.10:31
<escowles>yes, the filesystem connector isn't even in the stacktrace -- just FedoraNodes and LockReleasingSession -- so I think this is a problem with Modeshape -- so making sure we see the same error in the Modeshape filesystem connector should confirm this
<pivotal-bot>Esme Cowles started "Adding 22nd file to federated filesystem with REST API fails" https://www.pivotaltracker.com/story/show/7095453210:48
* longshou joins10:49
* gregjansen leaves10:52
<pivotal-bot>Esme Cowles added comment: "The error looks like it's happening when Modeshape is trying to load a document to append a new child entry ..." https://www.pivotaltracker.com/story/show/7095453210:56
Longshou Situ added comment: "The commit that fixed the conflicts is https://github.com/lsitu/fcrepo4/commit/d1821def16d3e739de6663d7eda..." https://www.pivotaltracker.com/story/show/7194631410:57
* gregjansen joins11:04
* scossu joins11:12
* longshou leaves11:22
* longshou joins11:24
* nikhiltri joins11:28
* scossu leaves
* scossu joins
* longshou_ joins
* longshou leaves11:29
<pivotal-bot>Longshou Situ started "NPE when submitting SPARQL query with FILTER" https://www.pivotaltracker.com/story/show/71073904
Stefano Cossu added comment: "The provided example was with a non-federated node.11:39
Actually, the problem was with my CND where I had defi..." https://www.pivotaltracker.com/story/show/69848860
* github-ff joins12:34
[fcrepo-module-auth-xacml] mohideen opened pull request #19: Xacmlworkspaceinitializertest (master...xacmlworkspaceinitializertest) http://git.io/Ir2NYQ
* github-ff leaves
<pivotal-bot>Mohamed Mohideen Abdul Rasheed added comment: "Created pull request https://github.com/fcrepo4/fcrepo-module-auth-xacml/pull/19" https://www.pivotaltracker.com/story/show/7358080412:35
Mohamed Mohideen Abdul Rasheed finished "Create XACMLWorkspaceInitializer Unit Test" https://www.pivotaltracker.com/story/show/73580804
Mohamed Mohideen Abdul Rasheed started "Cluster: Read load balancing" https://www.pivotaltracker.com/story/show/7455981012:42
Esme Cowles added comment: "I've filed a Modeshape bug for this: https://issues.jboss.org/browse/MODE-2252" https://www.pivotaltracker.com/story/show/7095453212:47
* dwilcox leaves12:49
* dwilcox joins12:51
* ksclarke1 joins12:54
* ksclarke leaves12:55
* ksclarke joins13:03
* ksclarke1 leaves13:04
<escowles>awoods: do you know how to configure fcrepo to use my branch of modeshape? i tried a simple fix and it makes the my test pass, but i don't know how to plug that in to fcrepo and try it out13:17
<awoods>escowles: you can probably change the dependency in the master pom... on a call.13:18
* escowles leaves13:45
* escowles joins13:47
* ksclarke1 joins
* ksclarke leaves13:49
* nikhiltri leaves13:54
<awoods>escowles: did you get it worked out?13:57
<escowles>awoods: nope, but i'm not sure the test is really working so i'm working on that now13:58
but i also remembered that i haven't been able to get all of modeshape to build (even without tests) -- so i'm not sure how i could make a jar file to depend on13:59
<awoods>escowles: I have the same issue14:00
escowles: I think I am going to create a ticket with my build output for the ModeShape team.
escowles: It might be good to get your output as well.14:01
escowles: would you mind running a clean master build of ModeShape piping the stdout to a file?
* gregjansen leaves
<escowles>awoods: sure, with tests disabled or enabled?14:02
<awoods>escowles: maybe one of each
* gregjansen joins14:06
<pivotal-bot>Andrew Woods added "Seamless external content" https://www.pivotaltracker.com/story/show/7462775014:07
Andrew Woods added comment: "I have created a new ticket for that feature: #74627750" https://www.pivotaltracker.com/story/show/69848860
Andrew Woods edited "Seamless external content" https://www.pivotaltracker.com/story/show/7462775014:08
Longshou Situ added comment: "One issue with Modeshape jcr/xml is that export works for objects with giant binary content more than 4GB ..." https://www.pivotaltracker.com/story/show/7455811814:09
Andrew Woods edited "Enable creation of nt:linkedFile nodes via REST API" https://www.pivotaltracker.com/story/show/69848860
Longshou Situ finished "Analyze jcr/xml export" https://www.pivotaltracker.com/story/show/74558118
Andrew Woods accepted "Enable creation of nt:linkedFile nodes via REST API" https://www.pivotaltracker.com/story/show/6984886014:10
Longshou Situ estimated "Transparent filesystem persistence" as 5 points https://www.pivotaltracker.com/story/show/7134708614:11
Longshou Situ started "Transparent filesystem persistence" https://www.pivotaltracker.com/story/show/71347086
* nikhiltri joins14:12
* gregjansen leaves14:23
* escowles leaves14:40
* escowles joins14:43
* gregjansen joins14:46
<escowles>awoods: modeshape builds for me if i use the options "-DskipTests -s settings.xml" -- apparently the settings.xml adds some JBoss repo where the missing artifacts can be found. if i keep tests enabled, i get a failure which i've put in this gist: https://gist.github.com/escowles/c8c35c17125e9a32b5b314:58
* gregjansen leaves15:05
* gregjansen joins15:14
<awoods>escowles: thanks. Let me compare that with my run.15:17
<escowles>awoods: now that i have a working mode build, i'm rebuilding fcrepo4 to use 3.8-SNAPSHOT and will test if it really fixes the problem15:18
<awoods>escowles: I take it you were able to find an obvious bug?
<escowles>awoods: i just put in a check for null before the NPE to see what would happen, and it fixed my test15:20
awoods: so that fixed creating objects and datastreams using the modeshape connector, but i'm still seeing an issue with the fedora connector15:29
<awoods>escowles: what is the nature of the fedora connector issue?
<escowles>it's a NPE when the FedoraFileSystemConnector tries to lookup the primary type15:31
so there may still be a problem serializing properties
<awoods>escowles: does the fedora-connector hit an NPE on every 22nd look-up?15:34
<escowles>awoods: no -- the fedora connector is hitting the NPE for every document -- no objects are being created at all15:40
<awoods>escowles: sounds like an integration test is needed15:41
<escowles>awoods: sure -- i can add an IT to demonstrate this problem15:42
<awoods>escowles: p.s. modeshape -DskipTests ran successfully15:44
<escowles>awoods: ok, so at least we can now build modeshape15:45
* github-ff joins16:04
[fcrepo4] escowles created writable-federation (+1 new commit): http://git.io/nNel_w
fcrepo4/writable-federation 2de88d6 Esmé Cowles: Enabling read-write operations on federated filesystem, depends on fix for MODE-2252 (https://github.com/escowles/modeshape/tree/mode-2252)
* github-ff leaves
<escowles>awoods: i found an ignored IT for read/write filesystem federation tests -- i enabled it and they all passed16:05
so i added a test in the writable-federation branch, to create nodes and it fails (errors)
* travis-ci joins16:06
[travis-ci] fcrepo4/fcrepo4#2133 (writable-federation - 2de88d6 : Esmé Cowles): The build has errored.
[travis-ci] Change view : https://github.com/fcrepo4/fcrepo4/commit/2de88d6b6352
[travis-ci] Build details : http://travis-ci.org/fcrepo4/fcrepo4/builds/29448654
* travis-ci leaves
<awoods>escowles: no you are working on getting that test to pass?16:08
escowles: now you are working on getting that test to pass?
<escowles>awoods: i've got to go pick up my kids from camp shortly, so i won't be working on it right now, but i will get back to it this evening or tomorrow am16:09
* github-ff joins
[fcrepo4] lsitu opened pull request #408: Fixed the NPE bug to support the SPARQL str function, and throw NotImplementedException for comparing subject variable value (master...feature/sparql_function) http://git.io/Bqkl0A
* github-ff leaves
<awoods>escowles: sure.
<escowles>i want to try writing properties, etc. and fully test the modeshape filesytem connector to see if it's working for all operations or not, since that will tell us if there is still a mode bug, or if we just need to fix the fedora connector16:10
<awoods>escowles: sounds like a good plan16:11
<pivotal-bot>Esme Cowles added comment: "I tried a simple fix for this (just checking for null right before the NPE and it fixed my modeshape tests: ..." https://www.pivotaltracker.com/story/show/7095453216:12
* escowles leaves16:15
<pivotal-bot>Longshou Situ added comment: "I think the NPE bug is fixed and the str() function in enabled for FILTER query support. However, the abov..." https://www.pivotaltracker.com/story/show/7107390416:16
Longshou Situ finished "NPE when submitting SPARQL query with FILTER" https://www.pivotaltracker.com/story/show/71073904
Longshou Situ added comment: "https://github.com/fcrepo4/fcrepo4/pull/408" https://www.pivotaltracker.com/story/show/7107390416:17
* nikhiltri leaves16:20
* scossu leaves
* dwilcox leaves16:24
* scossu joins16:29
* nikhiltri joins16:35
<pivotal-bot>Andrew Woods accepted "Analyze jcr/xml export" https://www.pivotaltracker.com/story/show/7455811816:54
* escowles joins17:00
* nikhiltri leaves17:09
* gregjansen leaves17:25
* edInCo joins18:05
<pivotal-bot>Andrew Woods delivered "NPE when submitting SPARQL query with FILTER" https://www.pivotaltracker.com/story/show/7107390418:18
* github-ff joins
[fcrepo4] awoods pushed 2 new commits to master: http://git.io/rFyxyA
fcrepo4/master 86ccf2c lsitu: Fixed the NPE bug to support the SPARQL str function, and throw NotImplementedException for comparing subject variable value.
fcrepo4/master e12bc00 Andrew Woods: Merge pull request #408 from lsitu/feature/sparql_function...
* github-ff leaves
<pivotal-bot>Andrew Woods added comment: "@scottprater, would you please verify this?
" https://www.pivotaltracker.com/story/show/71073904
* escowles leaves18:19
* mohamed leaves18:21
* ksclarke1 leaves18:29
<pivotal-bot>Scott Prater added comment: "@awoods @longshous : will do" https://www.pivotaltracker.com/story/show/7107390418:33
* scossu leaves18:36
* github-ff joins18:59
[fcrepo4] awoods pushed 1 new commit to hierarchy: http://git.io/3SJH4Q
fcrepo4/hierarchy 4c9625e lsitu: Enable transparent hierarchy support for versions, locks, import, export, batch, and fixity etc....
* github-ff leaves
<pivotal-bot>Andrew Woods added comment: "Resolved with: https://github.com/fcrepo4/fcrepo4/commit/4c9625e100469d12e9f3c9684ffa79bc25d987ba" https://www.pivotaltracker.com/story/show/7194631419:00
Andrew Woods delivered "Enabled transparent hierarchy feature for versions and other components, including refactor the codes to fix the..." https://www.pivotaltracker.com/story/show/71946314
Andrew Woods delivered "getChildren() on FedoraResource" https://www.pivotaltracker.com/story/show/71611298
* travis-ci joins19:12
[travis-ci] fcrepo4/fcrepo4#2136 (hierarchy - 4c9625e : lsitu): The build passed.
[travis-ci] Change view : https://github.com/fcrepo4/fcrepo4/compare/ab5139cfc26a...4c9625e10046
[travis-ci] Build details : http://travis-ci.org/fcrepo4/fcrepo4/builds/29460345
* travis-ci leaves
* edInCo leaves20:29
<pivotal-bot>Andrew Woods added comment: "See PR: https://github.com/umd-lib/fcrepo-module-auth-xacml/pull/1" https://www.pivotaltracker.com/story/show/7358080421:15
Andrew Woods rejected "Create XACMLWorkspaceInitializer Unit Test" https://www.pivotaltracker.com/story/show/73580804
* longshou leaves21:25