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

Using timezone: Eastern Standard Time
* nbanks joins00:29
* nbanks leaves00:34
* github-ff joins00:38
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/6hIp8w
fcrepo4/master dff8f8b Edwin Shin: Removes SpringJunit4ClassRunner from SanityCheckIT in fcrepo-webapp since we're...
* github-ff leaves
<bljenkins>Project fcrepo4 build #352: UNSTABLE in 13 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/352/00:51
Edwin Shin: Removes SpringJunit4ClassRunner from SanityCheckIT in fcrepo-webapp since we're
* travis-ci joins00:57
[travis-ci] futures/fcrepo4#354 (master - dff8f8b : Edwin Shin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/b440bfd985f4...dff8f8b04ec7
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6271654
* travis-ci leaves
* nbanks joins01:32
* kaarefc joins02:19
* kaarefc leaves03:01
* nbanks leaves
* kaarefc joins03:48
* nbanks joins04:07
* nbanks leaves04:11
* nbanks joins05:08
* nbanks leaves05:13
* github-ff joins05:15
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/eT1Dbg
fcrepo4/master 0d9961d Edwin Shin: Uses build-helper-maven-plugin:reserve-network-ports to set test.port property...
* github-ff leaves
<pivotal-bot>Edwin Shin edited "HTTP/1.1 Range requests" https://www.pivotaltracker.com/story/show/4611349105:17
Edwin Shin edited "HTTP/1.1 Range requests" https://www.pivotaltracker.com/story/show/46113491
Edwin Shin edited "Fedora datastream content request should support asynchronous operations" https://www.pivotaltracker.com/story/show/4222737505:18
<bljenkins>Yippie, build fixed!05:25
Project fcrepo4 build #355: FIXED in 10 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/355/
Edwin Shin: Uses build-helper-maven-plugin:reserve-network-ports to set test.port property
* nbanks joins
<bljenkins>Project fcrepo-kitchen-sink build #162: SUCCESS in 2 min 49 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/162/05:29
Project fcrepo-fixity build #153: SUCCESS in 3 min 3 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/153/05:30
* travis-ci joins05:35
[travis-ci] futures/fcrepo4#355 (master - 0d9961d : Edwin Shin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/dff8f8b04ec7...0d9961d8ecde
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6275722
* travis-ci leaves
* bljenkins leaves05:51
* bljenkins joins05:53
<pivotal-bot>Edwin Shin added "Upgrade Jenkins from to 1.502 to 1.510" https://www.pivotaltracker.com/story/show/47959249
Edwin Shin started "Upgrade Jenkins from to 1.502 to 1.510" https://www.pivotaltracker.com/story/show/47959249
* bljenkins leaves05:55
* bljenkins joins05:56
* eddies leaves06:33
* eddies joins06:58
* eddies leaves
* eddies joins
<pivotal-bot>Edwin Shin added "Switch fcrepo-rss from legacy-api to http-api" https://www.pivotaltracker.com/story/show/4796244107:07
Edwin Shin edited "Enable fcrepo4 to run in a cluster" https://www.pivotaltracker.com/story/show/4440465307:08
Edwin Shin finished "Upgrade Jenkins from to 1.502 to 1.510" https://www.pivotaltracker.com/story/show/4795924907:09
* github-ff joins07:12
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/_0WSVQ
fcrepo4/master 2ded623 Edwin Shin: Cleans up imports. Switches from junit.framework.Assert.assertEquals (which is deprecated)...
* github-ff leaves
<pivotal-bot>Esme Cowles added comment: "Ingested a sample zip file on futures6, and the contents were successfully extracted and saved as a separate..." https://www.pivotaltracker.com/story/show/4783028107:18
* escowles joins
* travis-ci joins07:32
[travis-ci] futures/fcrepo4#356 (master - 2ded623 : Edwin Shin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/0d9961d8ecde...2ded62312b0c
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6278175
* travis-ci leaves
* github-ff joins07:33
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/zBVJBA
fcrepo4/master 4791b75 Edwin Shin: minor: cleans up imports, formatting
* github-ff leaves
<bljenkins>Project fcrepo-kitchen-sink build #163: SUCCESS in 6 min 40 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/163/07:37
* github-ff joins
[fcrepo-kitchen-sink] eddies pushed 1 new commit to master: http://git.io/zDTPmA
fcrepo-kitchen-sink/master 2611224 Edwin Shin: minor pom cleanup that defers to dependency/plugin management declarations in parent pom
* github-ff leaves
<bljenkins>Project fcrepo-fixity build #154: SUCCESS in 8 min 1 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/154/07:40
* eddies leaves07:43
* eddies joins07:54
* eddies leaves
* eddies joins
* travis-ci joins
[travis-ci] futures/fcrepo4#357 (master - 4791b75 : Edwin Shin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/2ded62312b0c...4791b7559755
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6278557
* travis-ci leaves
* VincentNG joins08:07
<bljenkins>Project fcrepo-kitchen-sink build #164: SUCCESS in 4 min 28 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/164/08:10
Project fcrepo-fixity build #155: SUCCESS in 5 min 8 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/155/08:12
* eddies leaves09:15
* nbanks leaves09:18
* eddies joins09:28
* eddies leaves
* eddies joins
* ap2972 joins09:40
* nbanks joins09:50
* barmintor joins10:25
* ap2972 leaves10:33
<VincentNG>barmintor: for the mock bagit store you have under /test/resource/test-objects. what's the projected path to an object?10:37
is it "/objects/BagItFed1/testDS"?10:38
<barmintor>VincentNG: /objects/BagItFed1
<VincentNG>would that be an nt:file or nt:folder?
<barmintor>testDS is realized as a child node of the object
BagItFed1 is nt:folder, fedora:object, and bagit:archive10:39
also, the bagit cnd needs to be moved into src/main/resources10:40
(note to self)
<VincentNG>is there a way to get a child node of the BagItFed1 object?
without iterating
<barmintor>do you know the name of the node?10:41
<VincentNG>let's say testDS
"assuming i do"
<barmintor>object.getNode('testDS') should work
now it sounded like a dumb question10:42
<barmintor>it wasn't dumb
the bagit archive gets realized as a fedora object, the files get realized a datastreams
but this will be weird if we put together a bagit archive with subdirs under data10:43
<VincentNG>how did you generate the *.modeshape.json properties files? just curious
<barmintor>I didn't: ajs6f added it 8 days ago10:44
I'm not sure what it's for- I think it's the infamous "sidecar file"10:45
ajs6f knows more about that stuff in MODE than I do
<VincentNG>ok, i guess it makes more sense now. those files are only generated for me when i have "extraPropertiesStorage": "json" (which you don't have in your test)10:47
in the repo.json
I guess we regen them every time
we need to suppress it in the bags, anyway10:48
* ajs6f joins10:53
Anybody have the URL for the hangout?10:54
<escowles>ajs6f: you are familiar with bookmarks, right?
<ajs6f>escowles: We aren't always using the same computer every day.10:55
<escowles>ok, then you're familiar with bookmark syncing systems? del.icio.us? etc.?
<VincentNG>Google Chrome syncs your bookmarks too, if you sign in and create a bookmark10:56
<ajs6f>Thanks, nbanks.10:59
<pivotal-bot>Chris Beer edited "Figure out the XA deadlock problem" https://www.pivotaltracker.com/story/show/4784207511:01
Chris Beer started "Figure out the XA deadlock problem" https://www.pivotaltracker.com/story/show/47842075
<eddies>cbeer, ajs6f: let's go11:02
<ajs6f>Having trouble loading the hangout.
Don't wait for me.
<VincentNG>ajs6f: if you want, i can invite you by phone11:03
<ajs6f>Okay— 434-243-8644
<pivotal-bot>Esme Cowles finished "Create BagIt bag-info.txt -based ExtraPropertiesStore" https://www.pivotaltracker.com/story/show/4738331511:04
A. "Erdocamic" Soroka delivered "Create BagIt bag-info.txt -based ExtraPropertiesStore" https://www.pivotaltracker.com/story/show/4738331511:05
A. "Vebelliant" Soroka started "Switch fcrepo-rss from legacy-api to http-api" https://www.pivotaltracker.com/story/show/4796244111:06
Chris Beer edited "Figure out how to demo bagit connector on futures6" https://www.pivotaltracker.com/story/show/4788989911:09
<pivotal-bot>feature: Figure out how to demo bagit connector on futures6 (unstarted) / owner: Benjamin Armintor
Edwin Shin edited "Measure time to ingest & generate tech md for resources via the REST API vs filesystem connector" https://www.pivotaltracker.com/story/show/4737603711:12
<barmintor>ajs6f: Are you thinking that this unassigned "create an analogue of the fcrepo3 externally managed content type" ticket is just going to be a thing we do at Salem? It sounds related to our brief talk about projections and what-not yesterday.
<pivotal-bot>Esme Cowles edited "Measure time to ingest files of varying sizes & run a sequencer comparing the REST API vs BagItFileSystemConnector" https://www.pivotaltracker.com/story/show/4737603711:13
<ajs6f>I don't know— it sounds like it might be.
We certainly need to discuss our use of projectins a little mroe before we can do that.11:14
<barmintor>ajs6f: I feel like that's a conversation that would be much easier to have in the same room
you knw?
<ajs6f>Well, I'm barely present even when I am present.11:15
But yeah, we probably need to wave our hands a lot.
<cbeer>i wonder if the fcrepo3 is configured to do checksums and things11:16
* ap2972 joins11:17
<barmintor>I'm kind of perplexed- are we saying that we've seen a performance degradation in FCR4 since Chicago vis-a-vis FCR3?
* github-ff joins11:18
[fcrepo4] ajs6f pushed 1 new commit to master: http://git.io/MGxJ8A
fcrepo4/master 53c586f ajs6f: Removed dependency on HTTP APIs for testing per: https://www.pivotaltracker.com/story/show/47962441
* github-ff leaves
<pivotal-bot>chore: Switch fcrepo-rss from legacy-api to http-api (started) / owner: A. "Vebelliant" Soroka
A. "Vebelliant" Soroka added comment: "https://github.com/futures/fcrepo4/commit/53c586f18b7e847d953decb603c59a337cce35c8" https://www.pivotaltracker.com/story/show/47962441
<ajs6f>Two candidates for a datastream construction:11:25
How can I see the running instance on futures6? (I.e., what's the URL of that instance?) I'm trying to check whether I've finished: https://www.pivotaltracker.com/story/show/4796244111:27
<pivotal-bot>chore: Switch fcrepo-rss from legacy-api to http-api (started) / owner: A. "Vebelliant" Soroka
<bljenkins>Project fcrepo-kitchen-sink build #165: SUCCESS in 2 min 39 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/165/11:31
<cbeer>ok, looks like there is something stupid happening in the junit tests11:35
turns out when you send an update request to 4 different hosts for the same node, you get trouble
<pivotal-bot>Benjamin Armintor added comment: "I think this ticket should be deferred until we sort out how we want to deal with projections." https://www.pivotaltracker.com/story/show/47374071
* ap2972 leaves11:37
<bljenkins>Project fcrepo-fixity build #156: FAILURE in 9 min 12 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/156/11:38
* github-ff joins11:39
[fcrepo4] ajs6f pushed 1 new commit to master: http://git.io/tgkbMg
fcrepo4/master 09fb9a8 ajs6f: Removed unused import
* github-ff leaves
<pivotal-bot>Nigel Banks started "Update jmeter test plan to support fcrepo4" https://www.pivotaltracker.com/story/show/47457523
* travis-ci joins
[travis-ci] futures/fcrepo4#358 (master - 53c586f : ajs6f): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/4791b7559755...53c586f18b7e
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6285058
* travis-ci leaves
<VincentNG>ajs6f: what's HSM an acronym for?11:40
<ajs6f>Hierarchical Storage Management?11:41
good old fcrepo 3
<eddies>ah it is still there11:42
nbanks: can you post that google doc link to irc?
<ajs6f>I thought that instance was administered by Fedo?
<eddies>ajs6f: you shush now ;-)11:43
<cbeer>eddies: no, i added it back.
<cbeer>at least until we remember why we took it away before
* github-ff joins
[fcrepo-kitchen-sink] ajs6f pushed 1 new commit to master: http://git.io/2I-x0A
fcrepo-kitchen-sink/master 17cadb5 ajs6f: Added RSS resource to JAX-RS profile to expose it
* github-ff leaves
<pivotal-bot>A. "Vebelliant" Soroka added comment: "https://github.com/futures/fcrepo-kitchen-sink/commit/17cadb57eebc9242d13890561a6f4b0ab920cbdb" https://www.pivotaltracker.com/story/show/4796244111:44
A. "Vebelliant" Soroka edited "Switch fcrepo-rss from legacy-api to http-api" https://www.pivotaltracker.com/story/show/47962441
<nbanks>So I can run jmeter on another server and have it hit futures6 for both 3 and 4?
which server should I use?
<bljenkins>Project fcrepo4 build #360: UNSTABLE in 6 min 48 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo4/360/
A. Soroka: Removed unused import
Project fcrepo4 build #361: FAILURE in 4 min 10 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo4/361/11:50
* github-ff joins11:51
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/OBcIxw
fcrepo4/master ebf0a6d Edwin Shin: Closes the InputStream after we're done with it.
* github-ff leaves
<ajs6f>eddies: why not use try-with-resource for that?11:52
<eddies>because i don't know that idiom =)
<ajs6f>Seriously? It's _rad_.11:55
ajs6f taught me about it, too :P
<ajs6f>If you're going to use C-style block syntax, _do it_!
All the way!
Get the context of the block all in front of it.11:56
<eddies>ajs6f: can you give me a gist that demonstrates how to use it on that snippet?
i'm in the middle of three other changes and can't really wrap my head around anything new right now ;-)
<ajs6f>How about I just do it, and you can look at it?11:57
<eddies>that works too
i was just trying to be nice ;-)
and not assume you'd do it all for me =)
<ajs6f>This is fcrepo. Nobody does their own issue.12:00
* travis-ci joins12:02
[travis-ci] futures/fcrepo4#359 (master - 09fb9a8 : ajs6f): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/53c586f18b7e...09fb9a8efd9c
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6285649
* travis-ci leaves
<cbeer>"you do all the work over there.. and then i'll copy it into the code and claim all the credit and glory!"
<bljenkins>Yippie, build fixed!
Project fcrepo4 build #362: FIXED in 11 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/362/
Edwin Shin: Closes the InputStream after we're done with it.
<ajs6f>Glory? Is there glory to be had?12:03
I hadn't seen any yet.
<cbeer>when you TRACE fcrepo, modeshape, infinispan, jgroups and arjuna you get a lot of output. go figure.12:05
<bljenkins>Project fcrepo-kitchen-sink build #166: SUCCESS in 2 min 35 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/166/12:06
* nbanks leaves
* nbanks joins12:07
<bljenkins>Yippie, build fixed!
Project fcrepo-fixity build #157: FIXED in 3 min 4 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/157/
* github-ff joins12:08
[fcrepo4] ajs6f pushed 1 new commit to master: http://git.io/zu9gtw
fcrepo4/master 9ae0ae9 ajs6f: Demonstrated try-with-resources
* github-ff leaves
<cbeer>ajs6f: we're no longer maintaining repository size and whatnot on /objects, right?12:09
<ajs6f>I understand that we don't _need_ to. We may still be doing it without realizing it.
i think someone is locking /objects.. maybe it's just when adding a new child node?
i wonder if we need to shard under /objects somehow
* travis-ci joins12:12
[travis-ci] futures/fcrepo4#360 (master - ebf0a6d : Edwin Shin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/09fb9a8efd9c...ebf0a6dfc4b4
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6286055
* travis-ci leaves
<eddies>ajs6f: do you know what to make of the following?12:13
findbugs seems to be complaining (marking as a "critical" violation) this override of apply
<ajs6f>I click on that link and it doesn't show me any code. Where is this actually occuring?
<eddies>oh, click on FedoraTypesUtils in the third column12:14
<ajs6f>Okay, see this now.12:15
com.google.common.base.Function<F, T>
has Nullable annotations on the parameter of the apply() method.
In theory, to fulfill the complete contract of Function,
we should do a null-check in apply().
That's what it's bitching about.12:16
But Sonar is right. That would be the right thing to do.
We could use Guava's machinery for that, if I could remember what it is.12:17
Oh, its:12:18
com.google.common.base.Preconditions.checkNotNull(T, Object)
We could use liberally slather that around and it would make those faux pas disappear.12:19
<bljenkins>Project fcrepo-kitchen-sink build #167: SUCCESS in 2 min 18 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/167/12:20
Project fcrepo-fixity build #158: SUCCESS in 2 min 28 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/158/12:22
<eddies>ajs6f: i'm not sure i follow. does the null check need to happen in the map method or in each of the apply() methods?12:23
Look: here's the contract:
@Nullable T apply(@Nullable F input);12:25
(Which is kind of bullshitty. I mean, the return is sensibly @Nullable because
a function may not be surjective, but the parameter being null?
Either they've conflated the contracts of a generator and a function,
or they're arguing that it's possibly the set-union of a12:26
zeroary function and a unary function.
Which is kind of weak.)
Anyway, input has been declared to be possibly null.
But in our impls, we act as though it can't be.12:27
We dereference it without checking it.
That's what Sonar is telling us.
<eddies>right. i find the meaning of @Nullable a bit odd given the javadoc for apply: throws NullPointerException if {@code input} is null and this predicate does not accept null
* arguments
I don't think the contract is right.
<eddies>not really what i would have guessed Nullable to mean :P
<ajs6f>BUt Google isn't paying me to either write code or bitch about theirs.
* travis-ci joins
[travis-ci] futures/fcrepo4#361 (master - 9ae0ae9 : ajs6f): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/ebf0a6dfc4b4...9ae0ae9018df
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6286606
* travis-ci leaves
<ajs6f>Anyway, I think we have to check for nulls if we want Sonar to go away.12:29
Or possibly, we might be able to reannotate w/ a not-nullable annotation.
I don't know if that would work.12:30
<eddies>so *we* do want to throw an NPE if we're given a null Node here, right?
ack. have a call. bbl
<ajs6f>Yeah, I think we do.12:31
Which means that Guava's static checker would do fine.12:32
cbeer: did we ever kill the permgen problem?12:35
<cbeer>no. we just restart tomcat periodically12:36
* nbanks leaves
<ajs6f>Well, it must be barmintor's fault, since he's at lunch.12:37
I assume we're running a HotSpot JVM?
* github-ff joins12:39
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/eH5lug
fcrepo4/master 1053fd5 Edwin Shin: Adds null checks for apply(F), throwing an NPE if parameter is null as evidently...
* github-ff leaves
<eddies>ajs6f: futures6 is running12:42
java version "1.7.0_15"
OpenJDK Runtime Environment (IcedTea7 2.3.7) (7u15-2.3.7-0ubuntu1~12.04.1)
OpenJDK 64-Bit Server VM (build 23.7-b01, mixed mode)
<ajs6f>Yeah, that's what I would have thought.12:43
<eddies>isn't hotspot part of the openjdk vms now?12:44
<ajs6f>Frankly, I think the only way we will get to the bottom of this permgen thing is flat-out- profiling.
Which is not going to be pleasant.
<eddies>i thought fasseg mentioned that hibernate was known to be (at least partly) responsible
i don't know that anyone's tried passing in some different java opts to mitigate12:45
<ajs6f>Yes, that has been tried, by cbeer.12:47
The usual suspects around managing permegen and GC.
Maybe there are Hibernate adjustments we can make. I know < 0 about runtime HIbernate.12:48
out for errands bbl12:57
* ajs6f leaves
* travis-ci joins13:02
[travis-ci] futures/fcrepo4#362 (master - 1053fd5 : Edwin Shin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/9ae0ae9018df...1053fd5c50ce
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6287353
* travis-ci leaves
* nbanks joins13:07
* github-ff joins14:28
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/NV-sPw
fcrepo4/master 34175e0 Chris Beer: bump lock timeout back up, seems to make matters worse
* github-ff leaves
<bljenkins>Project fcrepo-fixity build #159: FAILURE in 1 hr 50 min: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/159/14:40
Project fcrepo-kitchen-sink build #168: FAILURE in 1 hr 51 min: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/168/
<bljenkins>Yippie, build fixed!14:46
Project fcrepo-kitchen-sink build #169: FIXED in 2 min 21 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/169/
Yippie, build fixed!14:47
Project fcrepo-fixity build #160: FIXED in 2 min 38 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/160/
* travis-ci joins14:49
[travis-ci] futures/fcrepo4#363 (master - 34175e0 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/1053fd5c50ce...34175e0ff1b3
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/6290115
* travis-ci leaves
* nbanks leaves14:57
<pivotal-bot>Esme Cowles edited "Measure time to ingest files of varying sizes & run a sequencer comparing the REST API vs BagItFileSystemConnector" https://www.pivotaltracker.com/story/show/4737603714:59
* bljenkins leaves15:05
* bljenkins joins15:07