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

Using timezone: Eastern Standard Time
* nbanks joins00:54
<bljenkins>Project fcrepo4-jmeter build #2: UNSTABLE in 26 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4-jmeter/2/01:25
* kaarefc joins01:42
* leeata joins01:48
* leeata leaves01:53
* kaarefc leaves02:09
<bljenkins>Project fcrepo4-jmeter build #4: FAILURE in 17 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4-jmeter/4/03:05
* nbanks_ joins03:13
* nbanks leaves03:16
* kaarefc joins03:39
* github-ff joins03:45
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/hmxkMQ
fcrepo4/master c5464e8 Edwin Shin: Adds some more logging to this sporadically failing test
* github-ff leaves
* bljenkins leaves03:48
* bljenkins joins03:49
* nbanks joins03:56
* nbanks_ leaves03:59
<bljenkins>Project fcrepo4 build #808: UNSTABLE in 15 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/808/04:04
Edwin Shin: Adds some more logging to this sporadically failing test
* travis-ci joins04:07
[travis-ci] futures/fcrepo4#688 (master - c5464e8 : Edwin Shin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/d3ec2098bc6c...c5464e8d94f4
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8043210
* travis-ci leaves
<bljenkins>Project fcrepo4-jmeter build #6: NOW UNSTABLE in 13 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4-jmeter/6/04:09
Edwin Shin: Adds some more logging to this sporadically failing test
Project fcrepo4-jmeter build #7: ABORTED in 11 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo4-jmeter/7/04:19
* eddies leaves04:29
* eddies joins04:31
* eddies leaves
* eddies joins
<bljenkins>Project fcrepo4-jmeter build #8: STILL UNSTABLE in 12 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4-jmeter/8/04:32
* eddies leaves04:41
* nbanks_ joins04:51
* nbanks leaves04:54
* nbanks_ joins04:58
* nbanks leaves04:59
* eddies joins05:01
* eddies leaves
* eddies joins
* github-ff joins06:33
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/h-ZM5Q
fcrepo4/master 90f0eda Edwin Shin: AtomJMSIT.testDatastreamTerm() assumed messages were arriving in a particular order,...
* github-ff leaves
<bljenkins>Yippie, build fixed!06:56
Project fcrepo4 build #809: FIXED in 22 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4/809/
Edwin Shin: AtomJMSIT.testDatastreamTerm() assumed messages were arriving in a particular order,
Yippie, build fixed!06:57
Project fcrepo-fixity-corrupter build #94: FIXED in 54 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/94/
* fasseg joins06:58
<bljenkins>Project fcrepo-fixity build #279: NOW UNSTABLE in 5 min 20 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/279/07:01
Project fcrepo-kitchen-sink build #378: STILL UNSTABLE in 6 min 35 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/378/07:04
<pivotal-bot>Edwin Shin added "AtomJMSIT.testDatastreamTerm fails intermittently" https://www.pivotaltracker.com/story/show/5163236307:08
Edwin Shin started "AtomJMSIT.testDatastreamTerm fails intermittently" https://www.pivotaltracker.com/story/show/51632363
Edwin Shin added comment: "https://github.com/futures/fcrepo4/commit/90f0eda7025cd3421c892be5d39786110aee6a57" https://www.pivotaltracker.com/story/show/51632363
Edwin Shin delivered "AtomJMSIT.testDatastreamTerm fails intermittently" https://www.pivotaltracker.com/story/show/5163236307:09
* travis-ci joins07:13
[travis-ci] futures/fcrepo4#689 (master - 90f0eda : Edwin Shin): The build has errored.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/c5464e8d94f4...90f0eda7025c
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8046746
* travis-ci leaves
<bljenkins>Yippie, build fixed!07:28
Project fcrepo4-jmeter build #9: FIXED in 14 min: http://ci.projectblacklight.org/jenkins/job/fcrepo4-jmeter/9/
Edwin Shin: AtomJMSIT.testDatastreamTerm() assumed messages were arriving in a particular order,
* eddies leaves07:33
* eddies joins07:45
* eddies leaves
* eddies joins
* eddies leaves07:46
* eddies joins07:47
* eddies leaves
* eddies joins
* ff-github joins07:51
[fcrepo-fixity] fasseg pushed 1 new commit to master: http://git.io/gbhmlA
fcrepo-fixity/master c2bce0e fasseg: added bean for FedoraTransactions
* ff-github leaves
<bljenkins>Yippie, build fixed!07:55
Project fcrepo-fixity build #280: FIXED in 3 min 46 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/280/
* bljenkins leaves07:59
* bljenkins joins08:09
* nbanks_ leaves08:29
* jcoyne joins08:32
* ajs6f joins
<jcoyne>barmintor: have you seen Jeff Minelli's email to the FC list?08:33
<ajs6f>fasseg: I see that you stepped up on the transactions endpoint— thanks! Did you put the session-injection stuff in there, too, or should I do that.
?
<fasseg>ajs6f: cbeer already merged my pull request for that
so the change is in master and you should be able to rebase on that i guess
<ajs6f>fasseg: I know, and I'm looking at master, and it doesn't have injected sessions for FedoraTransactions.08:34
cbeer already merged my every-other-endpoint PR.
<fasseg>oh sry now i understand the question
<ajs6f>Do you want o do it so you're sure you know how, or should I do it? It's like three lines of code.
<fasseg>o the session injection is not in there
*so
<ajs6f>right.
<fasseg>and if you could do that I'd be happy to answer any questions regarding the impl or change it when it doesnt fit your needs08:35
<ajs6f>Okay, I'll try to do it right against master, since we're really confident at this point, and it is literally a few lines.
<fasseg>but I guess it's just adding the session to the method signature or sth?08:36
so I could do that as well
<ajs6f>not to the sig, as a filed.
and adding a mutator, and maybe changing the u-test to use it. Thats all.
If you look at any other endpoint, you'll see it.
It looks like:08:37
@InjectedSession
protected Session session;
and then you just don't need to do final Session session = … in the methods, because session is automagiclly there.
* nbanks joins
<fasseg>ok ill do that in a branch, which you will review?
<ajs6f>Sure thing.
<fasseg>kk08:38
<ajs6f>cbeer: since you merged my branch (thanks!) I'm going to make those clean-up changes to the test helpers some other way… maybe a branch that just contains cleanups for u-tests?08:42
<jcoyne>barmintor++08:53
<ajs6f>cbeer: I take it back. The cleanup you suggested for TestHelpers can't be done because the session isn't being injected into AbstractResource ( I wish it was!) because of Jersey's weakness on inheritance. So we'll just have to leave .setSession(mockSession) in the individual tests, for now.08:55
* github-ff joins08:57
[fcrepo4] fasseg created tx-session-injection (+1 new commit): http://git.io/iuCSOA
fcrepo4/tx-session-injection 2c768c9 fasseg: added @InjectedSession and @Scope(prototype) to FedoraTransactions for Session injection
* github-ff leaves
<ajs6f>fasseg: You could see how to do it? Not hard, right?08:58
<fasseg>ajs6f: that's rigtht then: https://github.com/futures/fcrepo4/commit/2c768c9f5d8a1514229be57bd5717f9da619aae4 ?
yeah didnt had the scop in there at forst and the Transaction tests were fialing promptly08:59
<ajs6f>right, right. I'm surprised you didn't have to create a setter… how does the unit test still work?
<fasseg>the mocksession gets set via reflection in the unit tests09:00
<ajs6f>Really?
It didn't for me...
<fasseg>I so hope this is true for my own sake: http://www.profkrg.com/wp-content/uploads/2013/03/letters.jpg
hmm lemme check
<ajs6f>well, if tests are passing they are passing...
<fasseg>heh...there's no unit test for FedoraTransactions.....just an IT09:01
<ajs6f>problem… solved? :)
<fasseg>ill create a ticket :)
<ajs6f>quickly, before barmintor find out!
<fasseg>yeah theres only unit test for the TransactionService09:03
<pivotal-bot>Frank Asseg added "Add Unit test for FedoraTransactions in fcrepo-http-api" https://www.pivotaltracker.com/story/show/5163765109:04
<ajs6f>okay, that maks sense. I had to add a setter for session to each JaX-RS resource so that I could set a mockSession into them for their unnit tests. I imagine you will too.
<fasseg>I see
Just out of curiousity: what was the trouble when using reflection?09:06
<ajs6f>I don't really know what you mean… are the unit test helpers capable of that?09:07
<fasseg>sec...
<ajs6f>They seem to just build a bunch of mocks and set them on the test resources.
<fasseg>I got this one in the fixity app which does nothing else: https://github.com/futures/fcrepo-fixity/blob/master/fcrepo-fixity-core/src/test/java/org/fcrepo/fixity/TestHelper.java09:08
and it used here: https://github.com/futures/fcrepo-fixity/blob/master/fcrepo-fixity-core/src/test/java/org/fcrepo/fixity/FixityServiceTest.java#L3009:09
<ajs6f>Ah, that's very different looking than the TestHelper for the http stuff.09:13
* kaarefc leaves
<ajs6f>Maybe we should be using reflection btoh places?
barmintor: wtdy?09:14
wdyt?
<fasseg>Yeah I would suggest adding this method the the fcrepo TestHelper as well
<bljenkins>Project fcrepo-fixity-corrupter build #95: SUCCESS in 1 min 23 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/95/09:16
* bljenkins leaves09:30
* kaarefc joins09:33
* gregjansen joins09:37
* bljenkins joins09:39
Project fcrepo-kitchen-sink build #379: STILL UNSTABLE in 23 min: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/379/09:40
<ajs6f>fasseg/barmintor: we could take the pattern from fixity/TestHelper.java and09:41
wrap it in a dynamic proxy so that you could just do:
myTestObj.setFieldForTest(field, value);09:42
<bljenkins>Project fcrepo-fixity build #281: FAILURE in 26 min: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/281/
<ajs6f>where myTestObj is actually a proxy for the real testable.
* kaarefc leaves09:43
* kaarefc joins
<pivotal-bot>A. "Torbulater" Soroka accepted "Document our minimal OAuth impl and create tickets for follow on work" https://www.pivotaltracker.com/story/show/5156567309:48
* github-ff joins09:49
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/CGyPig
fcrepo4/master 002b086 Edwin Shin: 51517753: Publish JMeter reports to Sonar via Maven...
* github-ff leaves
<ajs6f>I'm pretty clueless about AWS. Can I use the authn info that eddies sent to sign into some kind of management console? I poked around but could only find a password-based login that accepted a password I set, but then tells me "You have not yet signed up for AWS. To log in, https://portal.aws.amazon.com/gp/aws/developer/subscription/index.html?ie=UTF8&packageCode=AWSUnifiedRegistration&redirectOnSignUp=1&redirectURL=https%3A%2F%2Fsig09:54
ooh, bad cut andpaste.
<eddies>ajs6f: sign in at https://duraspace.signin.aws.amazon.com/console09:58
the credentials i made for you are underneath the duraspace account
<ajs6f>eddies: Thanks, but I get:
There was a problem with your request
The AWS Account Alias you entered is invalid. Please correct and try again.
maybe I need to complete some additional step?09:59
<eddies>just hitting the url or w/ your credentials?
<ajs6f>w/ cred (my email add'y and the password that I just made)
<eddies>oh actually, i don't think the credentials i sent out let you hit the web console by default10:01
it just gives you api access
so you can use the command-line tools or spin stuff up programmatically10:02
<ajs6f>Oh, okay. Can we turn that on (the web console) or will I have to learn the SDK to build up a new instance?
Oh, there are cmdline tools? That's sounds easy enough.
* ff-github joins
[fcrepo-fixity] fasseg pushed 1 new commit to master: http://git.io/Xn7Nmw
fcrepo-fixity/master e3d357e fasseg: fixed query param handling for single fixity checks
* ff-github leaves
<ajs6f>installing now...10:03
* kaarefc leaves10:04
* kaarefc joins
<eddies>ajs6f: i generated a password for you too, which should let you use the web console
sent via email10:05
<pivotal-bot>Edwin Shin started "Generate AWS IAM accounts for Frank" https://www.pivotaltracker.com/story/show/51567371
Edwin Shin accepted "Generate AWS IAM accounts for Frank" https://www.pivotaltracker.com/story/show/51567371
<ajs6f>Got it, tried it, but got:
There was a problem with your request
There was an error with your User Name/Password combination. Please try again or contact your system administrator.
maybe it just hasn't updated yet?
It's a long way from Singapore.
<eddies>you're using the original username i sent you, plus the new password i just sent and hitting the duraspace url?10:08
<ajs6f>aha! no, that was it. I was using my email add'y. I'm in now! thanks!10:09
now to spin up a jillion giant instances and mint bitcoins.
* travis-ci joins10:10
[travis-ci] futures/fcrepo4#690 (master - 002b086 : Edwin Shin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/90f0eda7025c...002b08649649
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8051492
* travis-ci leaves
<aawoods>ajs6f: are you able to run the puppet script?
<ajs6f>I haven't gotten nearly that far. I need to create an instance first, right?
<aawoods>no
the script does that
<ajs6f>So where do I run it? Locally?
<aawoods>yes10:11
https://github.com/futures/fcrepo-aws-puppet
<ajs6f>Okay. We'll see how long it takes to install puppet and everything on a mac
* kaarefc leaves10:13
<aawoods>ajs6f: It would probably be useful for other mackies if you documented your install steps in the readme.10:18
* bljenkins leaves
<ajs6f>I'm not sure what to do with:10:21
Update the EC2 configuration elements found in:
fcrepo-aws-puppet/cloud-init/create-ec2-instance.conf
Update to what? Different size of instance or something?
<aawoods>ajs6f: you can use all of the default values listed there...10:22
<ajs6f>cool.
<aawoods>except...
the two empty fields.
<ajs6f>aws_group=[...]10:23
aws_key=[...]
<aawoods>one second, I will create a group and key for you.
<ajs6f>is that duraspace and my non-secret key?
<aawoods>no
<ajs6f>okay.
<aawoods>the group is a "firewall" group which defines the open ports
the key is the ssh key to accessing the server10:24
<ajs6f>okay, cool.
* bljenkins joins
Project fcrepo-kitchen-sink build #380: STILL UNSTABLE in 2 min 59 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/380/10:27
<aawoods>ajs6f: I sent you an email. The notes.doc is actually you pem file.10:30
<bljenkins>Project fcrepo-fixity build #283: SUCCESS in 8 min 1 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/283/10:32
<cbeer>eddies: can you make a console password for me too?10:36
fasseg: did you hear anything back from barmintor about those fixity failures?
<barmintor>send me an email, idk what you mean10:37
<fasseg>nah wanted to ping him in the committer's call if hes on there
ah hey!
<barmintor>debugging this fcr 3 upload thing for PSU
<fasseg>barmintor: For every fixity request I request form fcrepo on a datastream I get BAD_CHECKSUM errors
<eddies>cbeer: sent10:38
<cbeer>eddies: thanks
<fasseg>so e.g. at barmintor: e.g.: this response: http://fcrepo4.fcrepo.org/fcrepo/rest/objects/18dc9083-0674-4ee3-801a-9676c52b8a9c/license-1/fcr:fixity10:39
<eddies>ok. for number of hours of effort, it's a bit underwhelming, but jmeter results are now on sonar: http://sonar.fcrepo.org/dashboard/index/1?did=1
(this for just a tiny test plan, not the full test plan in ff-jmeter-madness)10:40
<ajs6f>aawoods: WHat shall I do with the pem file?
<fasseg>hmm wait these datastreams are actually null lemme check for another example on the webserver
<aawoods>ajs6f: nothing, unless you want to ssh into the instance.
<barmintor>fasseg: the job is not getting a bytestream to checksum. That's the empty string checksum, and length == 010:41
<ajs6f>Okay, cool.
<barmintor>you should figure out why the job can't find the bytestream
<fasseg>kk
<ajs6f>aawoods: getting "./create-user-data.sh: line 12: write-mime-multipart: command not found" Is that inlcuded in an AWS SDK or the like?10:42
<pivotal-bot>Edwin Shin added comment: "ended up resorting to 0.3-SNAPSHOT build of the the Sonar JMeter plugin: http://snapshots.repository.codehaus..." https://www.pivotaltracker.com/story/show/51517753
Edwin Shin added comment: "and the corresponding snapshot build of the report server: http://snapshots.repository.codehaus.org/org/codeh..." https://www.pivotaltracker.com/story/show/51517753
<aawoods>write-mime-mulitpart is included in the "cloud-utils" package10:43
<pivotal-bot>Edwin Shin added comment: "See preliminary results at http://sonar.fcrepo.org/dashboard/index/1?did=1" https://www.pivotaltracker.com/story/show/51517753
<aawoods>that you presumably installed locally
as mentioned in the readme
<pivotal-bot>Edwin Shin added "Daemonize jmeter-report-server on gluck" https://www.pivotaltracker.com/story/show/5164498510:44
Edwin Shin edited "Daemonize jmeter-report-server on gluck" https://www.pivotaltracker.com/story/show/51644985
<ajs6f>no, because I can't find a mac version. The readme only discusses Ubuntu.
it is this: https://launchpad.net/ubuntu/+source/cloud-utils ?
<eddies>standup time10:45
<ajs6f>maybe I will have to build it.
<cbeer>eddies: i'll take that ticket, unless you have it in progress?
<eddies>cbeer: that would be great
cbeer++
<pivotal-bot>Frank Asseg added "Investigate the large number of events that are created by RepositoryChangeBus$ChangeSetDispatcher.call()" https://www.pivotaltracker.com/story/show/5164543510:49
Frank Asseg started "Create new, specific performance enhancement tickets based on the profiling investigation" https://www.pivotaltracker.com/story/show/51567861
Frank Asseg added comment: "Added this ticket: https://www.pivotaltracker.com/story/show/51645435" https://www.pivotaltracker.com/story/show/51567861
<ajs6f>Has anyone installed cloud-utils on a Mac?10:51
I can't seem to find much info...
fasseg: I'm supposed to be looking at it? Is there a PR?
<fasseg>ajs6f: will create...10:52
<ajs6f>fasseg: cool
<fasseg>still in the middle of managing
<ajs6f>no prob10:53
<pivotal-bot>Frank Asseg added "Fixity checks always return BAD_CHECKSUM when retrieving fixity results" https://www.pivotaltracker.com/story/show/51645801
Chris Beer started "Daemonize jmeter-report-server on gluck" https://www.pivotaltracker.com/story/show/51644985
Frank Asseg edited "fcrepo4 always return BAD_CHECKSUM when retrieving fixity results " https://www.pivotaltracker.com/story/show/5164580110:54
Frank Asseg edited "fcrepo4 always returns BAD_CHECKSUM when retrieving fixity results " https://www.pivotaltracker.com/story/show/51645801
A. "Torbulater" Soroka started "Run fcrepo3 performance tests on AWS" https://www.pivotaltracker.com/story/show/5151598710:58
Frank Asseg added "Inject Session in FedoraTransactions" https://www.pivotaltracker.com/story/show/5164646111:01
Chris Beer accepted "Daemonize jmeter-report-server on gluck" https://www.pivotaltracker.com/story/show/5164498511:02
* github-ff joins
[fcrepo4] fasseg opened pull request #81: added @InjectedSession and @Scope(prototype) to FedoraTransactions (master...tx-session-injection) http://git.io/sfNvPg
* github-ff leaves
<pivotal-bot>Frank Asseg added comment: "created pull request: https://github.com/futures/fcrepo4/pull/81" https://www.pivotaltracker.com/story/show/51646461
Frank Asseg finished "Inject Session in FedoraTransactions" https://www.pivotaltracker.com/story/show/51646461
<fasseg>ajs6f: It's all your's: https://www.pivotaltracker.com/story/show/5164646111:03
<pivotal-bot>feature: Inject Session in FedoraTransactions (finished) / owner: Frank Asseg
* github-ff joins
[fcrepo4] ajs6f pushed 1 new commit to master: http://git.io/KG3JHw
fcrepo4/master b916f69 ajs6f: Merge pull request #81 from futures/tx-session-injection...
* github-ff leaves
<ajs6f>merged
<fasseg>https://www.pivotaltracker.com/story/show/5164543511:05
<pivotal-bot>feature: Investigate the large number of events that are created by RepositoryChangeBus$ChangeSetDispatcher.call() (unscheduled) / owner:
Chris Beer delivered "Inject Session in FedoraTransactions" https://www.pivotaltracker.com/story/show/5164646111:08
Chris Beer delivered "alter FedoraTransactions so that it no longer maintains the state of the transaction store inside the endpoint itself" https://www.pivotaltracker.com/story/show/5160251311:09
* nbanks_ joins11:11
<pivotal-bot>Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:12
* nbanks leaves
<pivotal-bot>Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/51519825
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:14
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:15
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:16
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:18
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:19
Andrew Woods edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:20
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/51519825
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/51519825
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/51519825
Andrew Woods edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:21
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/51519825
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:22
<bljenkins>Project fcrepo-fixity-corrupter build #97: SUCCESS in 44 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/97/11:24
<pivotal-bot>Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:25
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:27
* travis-ci joins
[travis-ci] futures/fcrepo4#692 (master - b916f69 : ajs6f): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/002b08649649...b916f69e4575
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8054142
* travis-ci leaves
<bljenkins>Project fcrepo-kitchen-sink build #381: STILL UNSTABLE in 4 min 12 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/381/11:28
Project fcrepo-fixity build #284: SUCCESS in 5 min 52 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/284/11:29
<pivotal-bot>Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:30
Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:33
Chris Beer edited "fcrepo4 always returns BAD_CHECKSUM when retrieving fixity results " https://www.pivotaltracker.com/story/show/5164580111:34
Chris Beer started "fcrepo4 always returns BAD_CHECKSUM when retrieving fixity results " https://www.pivotaltracker.com/story/show/51645801
<cbeer>fasseg: ^^ i'm going to do some poking around
<fasseg>kk
<pivotal-bot>Edwin Shin edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982511:35
<cbeer>barmintor-- # i might have to throw this fixity thing back at you11:39
<barmintor>oh ho ho
now I'm all --
@karma barmintor
<pivotal-bot>Karma for barmintor has been increased 18 times and decreased 3 times for a total karma of 15.
<barmintor>cbeer: go ahead, I'll see what I can do
<cbeer>k. i'll write up what i know in that ticket11:40
<barmintor>probably means there's a hole in the IT, as well
<cbeer>barmintor: i think it's a binary store config issue. i'm trying to figure out if you're pulling the wrong store or what.11:43
barmintor: ok. i probably found the issue. https://github.com/futures/fcrepo4/blob/master/fcrepo-kernel/src/main/java/org/fcrepo/services/ServiceHelpers.java#L122 is just grabbing the first cache11:49
not necessarily the data store
<barmintor>Hmm
so probably an alt method that cycles through with a key, and sess if any respond?11:50
<cbeer>the test config happens to use the same cache for both metadata and binaries
barmintor: i think we can pull out the binary cache through some package abuse.11:51
* barmintor shifts uncomfortably
<cbeer>before we were doing the DEF, i think we did iterate and call it good
actually..11:52
barmintor: ArrayLists guarentee order, right?
<barmintor>cbeer: yes
<cbeer>because:11:53
protected Cache<String, Metadata> metadataCache;
private Cache<String, byte[]> blobCache;
:/
hm. i guess i could see what type of cache it is
<barmintor>interesting
can you inspect the runtime generic decls?11:54
ajs6f^^
<ajs6f>Not after type-erasure.
Jave generics--
Scala generics++
A List<String>, at runtime, is just a List.
<barmintor>we should have done this whole project in Scala
<ajs6f>Damn right.11:55
<cbeer>hm.
<ajs6f>Dev challenge: rewrite everything in Scala!
<barmintor>god, the fcr3 test suite11:57
<ajs6f>^^^ Yep.
<cbeer>ok. i give up. i'm just taking the last cache store and leaving a big note about how we're making bad assumptions11:58
<ajs6f>Fedora: We're lying to ourselves.11:59
<barmintor>what if I pulled all the grizzly2 crap I just did in fcr4 into fcr3?
could I even start fcr3 that way?12:00
<ajs6f>In theory, but have fun getting FEDORA_HOME into shape.
<barmintor>screw FEDORA_HOME
<ajs6f>eddies and I have done it by just including a pre-packed F_H.
Well, let me know how it goes. It would be a vast improvement over carrying Tomcat around everywhere.12:01
Has anyone run the AWS test install Puppet script on a Mac? I can't seem to find any way to get at this cloud-utils package that is needed.12:02
<aawoods>eddies: It is probably reasonable to go ahead and submit that full list of Dev Challenge Proposals in one click to the OR2013 website.
I will hit it unless someone else is on it.
<cbeer>oh, and eddies: i assume the jmeter server thing works as long as i see the jmeter process running and the pane in sonar12:03
<eddies>barmintor: it's not as nice as using grizzly, but i did make that cargo plugin for fedora12:04
<ajs6f>Be sure to add: "Challenge: Find a beer that makes Eddie forget about the Rodenbach 2008".
<eddies>cbeer: oh i still have the shell script running in a terminal. let me kill that
<ajs6f>barmintor: I was able to use plain Cargo for i-tests with fcr3
<cbeer>CLOSED WONTFIX
<barmintor>I'm going to spend some time burning the fcr3 test suite to the ground
<cbeer>eddies: i killed it
<eddies>nope. someone else killed it
=)
<ajs6f>barmintor: Seriously?
it's pretty huge...
<barmintor>I'm not going to finish today or anything, but this is fucking ridiculous12:05
<ajs6f>Yes, it is, but isn't 3.7 3.Final anway?
<eddies>oh there's a beer fest in singapore for the next three days. and i just found a bottle shop (the only one, apparently) that has some stock of rodenbach grand cru =)
<barmintor>eddies++
<eddies>$9 for 330ml i think
i'll take if it's real
* github-ff joins
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/mhr8xQ
fcrepo4/master c62ed3c Chris Beer: be sure to pluck the blob cache out of the infinispanbinarystore to run binary checks against.
* github-ff leaves
<ajs6f>Do people sell fake beer?12:06
<pivotal-bot>Chris Beer finished "fcrepo4 always returns BAD_CHECKSUM when retrieving fixity results " https://www.pivotaltracker.com/story/show/51645801
<eddies>ajs6f: i just mean if they really have it
<pivotal-bot>Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/c62ed3cec1446199a59bee4e0afddb54126a499b" https://www.pivotaltracker.com/story/show/51645801
<ajs6f>oh, right.
<eddies>although fake liquor's everywhere
<pivotal-bot>Chris Beer added comment: "That just fixes the problem and changes a test configuration to hopefully catch it in the futures (when the i..." https://www.pivotaltracker.com/story/show/51645801
<eddies>wasn't there just a scandal at like a dozen applebees in NJ that were serving fake liquor?12:07
<ajs6f>I could not say. [insert either NJ or Applebees joke here, as you prefer]12:08
<eddies>barmintor: the fedora-client project uses that fedora cargo plugin, if that's helpful as an example12:09
<barmintor>eddies thanks
<pivotal-bot>Chris Beer delivered "Document the size of an AWS instance we should run performance testing against" https://www.pivotaltracker.com/story/show/51517025
<cbeer>ajs6f: can you check and deliver https://www.pivotaltracker.com/story/show/51571173 ?12:10
<pivotal-bot>feature: Stop recreating a test namespace every time a JAX-RS resource is created. (finished) / owner: Chris Beer
<eddies>actually. we *should* release fedora 3.7 by OR
<cbeer>and maybe tell me how to know if https://www.pivotaltracker.com/story/show/51077645 is working so it can be delivered?
<pivotal-bot>feature: Use Apache Oltu to provide simple OAuth2 authentication (finished) / owner: A. "Torbulater" Soroka
<cbeer>oh, and deliver https://www.pivotaltracker.com/story/show/51502565 (which i think is done?)
<pivotal-bot>feature: Refactor http-commons test rig to use grizzly2 and parse web.xml (finished) / owner: Benjamin Armintor
<cbeer>eddies: is https://www.pivotaltracker.com/story/show/51517753 finished, or did you want to pull the upstream test first?12:11
<pivotal-bot>feature: Publish JMeter reports to Sonar via Maven (started) / owner: Edwin Shin
<eddies>cbeer (at least git blame says it was you): is modeshape-web-jcr necessary for anything? it's a dependency of fcrepo-webapp and there are corresponding web.xml entries. but i don't think we're using it and when i nuked it locally, everything still builds/seems to run fine12:12
sorry modeshape-web-jcr-rest
<cbeer>eddies: that's the modeshape provided rest endpoint that someone (you?) wired in12:13
<eddies>cbeer: it's finished. i'll make a new ticket for the follow-on
<cbeer>oh, look. i wired it in.
<eddies>hmm. if it was me, i don't remember
=)
<cbeer>i'm not attached to it.
<eddies>ok, i'll push my updates that nuke it in a sec
and do we know why all the nice new ITs you added to kitchen-sink are failing?12:14
<ajs6f>cbeer: Is https://www.pivotaltracker.com/story/show/51571173  done in master?12:15
<pivotal-bot>feature: Stop recreating a test namespace every time a JAX-RS resource is created. (finished) / owner: Chris Beer
<cbeer>eddies: on it. must be some lingering spring junk
<pivotal-bot>A. "Torbulater" Soroka delivered "Refactor http-commons test rig to use grizzly2 and parse web.xml" https://www.pivotaltracker.com/story/show/51502565
<cbeer>ajs6f: sorry, i'll find the commit
<ajs6f>cool
as for https://www.pivotaltracker.com/story/show/5107764512:16
<pivotal-bot>feature: Use Apache Oltu to provide simple OAuth2 authentication (finished) / owner: A. "Torbulater" Soroka
Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/5f087c1067ebd6515c28d3703d6f9c91977005d8" https://www.pivotaltracker.com/story/show/51571173
<ajs6f>the Authn and Authx page in our wiki shows the recipe for trying it..
<pivotal-bot>A. "Torbulater" Soroka delivered "Stop recreating a test namespace every time a JAX-RS resource is created." https://www.pivotaltracker.com/story/show/5157117312:17
<bljenkins>Project fcrepo-fixity-corrupter build #98: SUCCESS in 43 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/98/12:19
<aawoods>ajs6f: These may be helpful:
https://github.com/lovelysystems/cloud-init/blob/master/tools/write-mime-multipart12:20
https://help.ubuntu.com/community/CloudInit
<ajs6f>ah, the source code. thanks!
I could not find it.
<aawoods>I do not think it is the official source code... so use at your own risk
<ajs6f>I have no choice, do it?12:21
do I
<aawoods>it always has a choice
<ajs6f>That's not what Martin Buber thought. (Yes! Finally, a chance for a Martin Buber joke!)
<barmintor>https://twitter.com/Justin_Buber12:22
<bljenkins>Project fcrepo-kitchen-sink build #382: STILL UNSTABLE in 3 min 15 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/382/
<barmintor>"I bear within me the sense of Self, that is not included in the world. I could be your Buzz Lightyear, fly across the globe."
<ajs6f>Ouch. I do not want to be in a relationship of dialog with Justin Bieber.
<barmintor>"f I was your boyfriend, I'd never let you go, I'd never let you go. The self-willed man knows only his feverish desire to use the world."12:23
"Real faith means holding ourselves open to the mystery in every sphere of our life. Ooh pretty lady, don't you think it's crazy?"
<bljenkins>Project fcrepo-fixity build #285: SUCCESS in 5 min 8 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/285/
<aawoods>that is beautiful, barmintor.
<ajs6f>There is a whole world of these streams: https://twitter.com/KimKierkegaard12:25
Is this what the Web has come to? Yes. It was always the plan.
* travis-ci joins12:28
[travis-ci] futures/fcrepo4#693 (master - c62ed3c : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/b916f69e4575...c62ed3cec144
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8056266
* travis-ci leaves
<barmintor>ajs6f: do you remember how to get eclipse to stop barking about the XSD generated classes in fcr3? (APIAMTOM and all that)12:35
<ajs6f>Oh, jeez.
<barmintor>that's what I said12:36
<ajs6f>It's got something to do with doing a build with cli Maven first, I think.
If you do that, you generate the code from the WSDL and that helps.
But you may still have to tell Eclipse where the generated code is (in COnfigure Build Path...)12:37
WSDL first. Yay.
Actually WSDL-first can be cool, but we never did it right.
<barmintor>I take issue with the first half of that sentence, but that's neither here nor there12:38
<ajs6f>Nor is fcrepo3.
<barmintor>mmm
<cbeer>anyone know what the prefixes are about in the bagit object serialization?12:39
<ajs6f>The ones that get used to filter?
<barmintor>ajs6f: add fcrepo-common/target/generated-sources/cxf to source path in project12:40
<cbeer>ajs6f: yeah.. looks like.
<ajs6f>barmintor: No. No, I will not.
<cbeer>ajs6f: i'm trying to figure out of i can hard code them, or just not filter at all
or if our scan-all-the-things model is already falling apart.12:41
<ajs6f>cbeer: You can hard code them, but I thought it would be nice to leave them adjustable, because some people may want to include metadata from namespaces that aren't dc:
You probably don't want to leave them out, because you could get a lot of cruft.
You can do it while still scanning.
<cbeer>oh?12:42
<barmintor>File hasStarted = new File(FEDORA_HOME, "server/fedora-internal-use/has-started.txt");
<ajs6f>Put them in a util:map, and call them config.
* barmintor shakes his head
<ajs6f>or util:set, or something.
barmintor: I thikn that's a marker for "has first-run setup occurred?".
barmintor: like loading default XACML policies.12:43
<barmintor>yea
still: No.
<ajs6f>To fcrepo3, we say a great No.
<barmintor>classesthatare2000lineslong--12:50
<ajs6f>aawoods: fyi, got an instance up. yay!12:56
now for ssh'ing in.
<aawoods>ajs6f: I am glad. Would you please update the scripts and/or the readme to help other mac-lings?12:57
<ajs6f>SUre.
<aawoods>The YourKit agent tie-in is in /etc/default/tomcat712:58
It is a standard debian tomcat, running at /var/lib/tomcat7
<pivotal-bot>Chris Beer added comment: "see https://wiki.duraspace.org/display/FF/Authentication+and+Authorization" https://www.pivotaltracker.com/story/show/5107764512:59
<cbeer>ajs6f: should auth-oauth get wired into kitchen sink?
<ajs6f>cbeer: I don't know. It wouldn't hurt anything, but it really should have more functionality added before we feature it as a feaute.13:00
* github-ff joins13:01
[fcrepo4] eddies pushed 1 new commit to master: http://git.io/2v3Pyw
fcrepo4/master 172e0c8 Edwin Shin: Drops unused modeshape-web-jcr-rest dependency and corresponding web.xml config
* github-ff leaves
* github-ff joins13:02
[fcrepo-bagit-object-serialization] cbeer pushed 1 new commit to master: http://git.io/zPetww
fcrepo-bagit-object-serialization/master ac6b189 Chris Beer: update bagit serialization for the new spring and IT configurations
* github-ff leaves
<bljenkins>Project fcrepo-bagit-object-serialization build #14: UNSTABLE in 1 min 51 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-bagit-object-serialization/14/13:04
<cbeer>ajs6f: as long as it doesn't hurt, i'm going to go ahead and do it.
* github-ff joins
[fcrepo-kitchen-sink] eddies pushed 1 new commit to master: http://git.io/v6lL0A
fcrepo-kitchen-sink/master a2476be Edwin Shin: Drops unused modeshape-web-jcr-rest config from web.xml...
* github-ff leaves
<bljenkins>Project fcrepo-kitchen-sink build #383: STILL UNSTABLE in 2 min 30 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/383/
<ajs6f>cbeer: I wasn't making a guarantee so much as a weak predication.
<bljenkins>Yippie, build fixed!13:08
Project fcrepo-kitchen-sink build #384: FIXED in 3 min 16 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/384/
* github-ff joins
[fcrepo-bagit-object-serialization] cbeer pushed 1 new commit to master: http://git.io/59YeQg
fcrepo-bagit-object-serialization/master a42a27d Chris Beer: scan fcrepo services.
* github-ff leaves
<cbeer>eddies: caught you not running your tests!13:12
wait, now i'm confused.13:13
<eddies>?
<cbeer>eddies: i'm wrong. my ide helpfully forgot to refresh after pulling your code
<eddies>=)
<cbeer>why did you move metrics into webapp, though?13:14
<bljenkins>Yippie, build fixed!13:15
Project fcrepo-bagit-object-serialization build #15: FIXED in 7 min 14 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-bagit-object-serialization/15/
Project fcrepo-fixity-corrupter build #99: SUCCESS in 1 min 5 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/99/13:19
<ajs6f>Anyone have a good WSYIWYG Markdown editor for Mac they like?
<eddies>cbeer: you mean the MetricsConfig in repo.xml?13:20
i wanted to go through swap out the spring xml config in general for annotations, but didn't have the heart.13:21
* travis-ci joins
[travis-ci] futures/fcrepo4#694 (master - 172e0c8 : Edwin Shin): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/c62ed3cec144...172e0c8ab098
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8058140
* travis-ci leaves
<cbeer>eddies: https://github.com/futures/fcrepo-kitchen-sink/commit/69fb903bb885abeb8047fc86426f7a6eea66342e
<aawoods>ajs6f: here is a cheatsheet: https://github.com/adam-p/markdown-here/wiki/Markdown-Cheatsheet13:22
<eddies>cbeer: you mean, why did i *remove* that metrics-reporting.xml from kitchen-sink?
<aawoods>Also editing directly in github allows for "preview".
<cbeer>eddies: ok, sure. i thought i saw a commit that added it to webapp.13:23
<ajs6f>aawoods: I'm working from a cheat sheet. It's like going back in time to 1985. I'll try the Github editor.
<eddies>cbeer: i dropped that xml config in kitchen-sink in favor of https://github.com/futures/fcrepo4/blob/master/fcrepo-metrics/src/main/java/org/fcrepo/metrics/MetricsConfig.java
<cbeer>oh, hm.13:24
ajs6f: i'm having trouble getting the authn stuff to work.. i guess i'll push it to a branch and you can tell me what i'm missing
<eddies>ajs6f: i've used Mou (http://mouapp.com/). There's also markdown editor support for Eclipse (don't have link handy). But I just use TextMate with the Markdown bundle for the most part13:25
<ajs6f>What are you trying to do? Just get a token?
<bljenkins>Project fcrepo-fixity build #286: SUCCESS in 7 min 9 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/286/
<ajs6f>eddies: yeah, Mou requires 10.7.
* github-ff joins
[fcrepo-kitchen-sink] cbeer created auth-oauth (+1 new commit): http://git.io/uCXcVQ
fcrepo-kitchen-sink/auth-oauth 6ee0899 Chris Beer: wire in fcrepo-auth-oauth
* github-ff leaves
<ajs6f>I tried the Eclipse editor, but it dosn't offer a preview that I can find.
<eddies>ajs6f: what are you running?
<ajs6f>This Github editor is fine.
10.6
<bljenkins>Project fcrepo-kitchen-sink build #385: SUCCESS in 7 min 37 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/385/13:27
Project fcrepo-kitchen-sink build #386: FAILURE in 1 min 1 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/386/13:28
<cbeer>aawoods: what do you think about moving the stories on https://wiki.duraspace.org/display/FF/Sponsor+Use+Cases
into sub-pages per story?
(i'm also not sure why we distinguish between the use cases, but..13:29
<aawoods>cbeer: Yes, the current page is too busy. I was thinking about keeping the existing page, but moving each institution into its own sub-page.13:30
<cbeer>sure, that works (but I guess gets a little weird if multiple institutions have the same story?)
maybe that's ok though.
<aawoods>We can clean that up when it happens
Are you on that gardening?13:31
<cbeer>sure
* github-ff joins13:42
[fcrepo-aws-puppet] ajs6f pushed 1 new commit to master: http://git.io/S48Diw
fcrepo-aws-puppet/master b807c44 ajs6f: Update README.md
* github-ff leaves
<bljenkins>Project fcrepo-auth-oauth build #1: UNSTABLE in 5 min 0 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-auth-oauth/1/13:44
* github-ff joins13:45
[fcrepo-aws-puppet] ajs6f pushed 1 new commit to master: http://git.io/-6wwXw
fcrepo-aws-puppet/master 5533f6b ajs6f: Update README.md
* github-ff leaves
* github-ff joins13:46
[fcrepo-aws-puppet] ajs6f pushed 1 new commit to master: http://git.io/2wfavw
fcrepo-aws-puppet/master a5f838c ajs6f: Update README.md
* github-ff leaves
<ajs6f>aawoods: You sent me a private key for SSH login to the AWS intsances. Is the public key already there?13:49
<aawoods>yes, but I would have to see where AWS stashes it13:50
what are you trying to do?
<ajs6f>log in
I haven't tried yet, just setting up
<aawoods>ssh -i <your pem> ubuntu@<aws-dns-name>13:51
<ajs6f>Thanks!13:52
Just to be clear (although I don't think it would make much difference) we're going to run tests against 3.6.2, not trunk, right?13:55
And we're using the 'quick' (default) install, right?13:58
<aawoods>sounds reasonable14:00
<ajs6f>ok
Oh, yeah. Got to put FEDORA_HOME in somewhere. {sigh}14:06
Can we still do that in web.xml? Or anyway, is there a standard place to put environment variables for Tomcat in Ubuntu?14:09
<pivotal-bot>Chris Beer added "Finish wiring auth-oauth into kitchen sink" https://www.pivotaltracker.com/story/show/5166194314:13
<aawoods>ajs6f: /etc/default/tomcat7
sudo service tomcat7 restart
<ajs6f>aawoods: thanks, just did it with a context-param
http://ec2-174-129-72-48.compute-1.amazonaws.com:8080/fedora/describe14:16
Yay.
<cbeer>aawoods: would you say we've accomplished https://www.pivotaltracker.com/story/show/45692317?14:20
<pivotal-bot>feature: Verify ease of deployment of fcrepo-kitchen-sink to AWS (unstarted) / owner: Jonathan Green
<aawoods>cbeer: yes, but fcrepo-kitchen-sink will soon be swapped with fcrepo-webapp14:21
<pivotal-bot>Chris Beer accepted "fcrepo-jms integration test fails intermittently" https://www.pivotaltracker.com/story/show/48578517
Chris Beer started "Verify ease of deployment of fcrepo-kitchen-sink to AWS" https://www.pivotaltracker.com/story/show/45692317
Chris Beer delivered "Verify ease of deployment of fcrepo-kitchen-sink to AWS" https://www.pivotaltracker.com/story/show/45692317
Chris Beer deleted "Automate modeshape snapshot builds." https://www.pivotaltracker.com/story/show/4818401914:22
<ajs6f>I suppose we should run the JMeter scripts against fcrepo3 from the same mcahine, to eliminate network oddities?
Also, have we settled on a standard config for loops/file size/std dev for these tests?14:28
<aawoods>ajs6f: No, I believe you will be pioneering the detailing of those parameters.14:29
<ajs6f>okay, I'll just make something up
<aawoods>we have faith in your creative abilities
<ajs6f>I can definitely invent 3 numbers,14:30
but they may not be very creative.
<pivotal-bot>Andrew Woods accepted "Verify ease of deployment of fcrepo-kitchen-sink to AWS" https://www.pivotaltracker.com/story/show/4569231714:32
<cbeer>(free points!)14:33
<aawoods>karma14:34
off to an appt.14:36
* kaarefc joins14:50
<ajs6f>Took me a while to notice this, but am I right in thinking that the ff-jmeter test plan has been rewritten for fcrepo4 and now no longer works with fcrepo3's API at all?15:01
I guess it'll have to be altered..
hm, looks like just datastreams was changed.15:03
* bljenkins leaves15:15
* bljenkins joins15:22
* nbanks leaves15:26
* nbanks joins15:27
<cbeer>i need to stop being so surprised when, after making a change: a) everything compiles, b) the tests pass, c) the feature actaully works15:31
* fasseg leaves
<ajs6f>It still surprises me.
<cbeer>i mean, writing more tests certainly isn't a problem, but..15:32
* gregjansen leaves15:33
<pivotal-bot>Chris Beer finished "Add pagination for child node listings in rdf responses" https://www.pivotaltracker.com/story/show/5118392715:36
* github-ff joins15:37
[fcrepo4] cbeer pushed 2 new commits to master: http://git.io/sHqLcg
fcrepo4/master a811dc6 Chris Beer: add logging to serializer map builder
fcrepo4/master 4be6008 Chris Beer: update FedoraResource#getPropertiesDataset to include limit/offset parameters to control which (if any) child object graphs are returned in-full. by default, all child graphs are returned in full.
* github-ff leaves
<pivotal-bot>Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/4be6008c3225960fc92f49238f95f5bf268153cf" https://www.pivotaltracker.com/story/show/51183927
Chris Beer edited "Add pagination for child node listings in rdf responses" https://www.pivotaltracker.com/story/show/5118392715:38
Chris Beer delivered "Add pagination for child node listings in rdf responses" https://www.pivotaltracker.com/story/show/51183927
Chris Beer started "Make fcrepo-object-serialization handle mime-types" https://www.pivotaltracker.com/story/show/4683958715:46
<bljenkins>Project fcrepo-fixity-corrupter build #100: SUCCESS in 41 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/100/15:50
* github-ff joins
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/RBVUHw
fcrepo4/master 1578d77 Chris Beer: add a media type hint to serializers so they can provide the correct content type header to http endpoints
* github-ff leaves
<pivotal-bot>Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/1578d77c2a77d584d5fa979b30da601bf32e0012" https://www.pivotaltracker.com/story/show/46839587
Chris Beer finished "Make fcrepo-object-serialization handle mime-types" https://www.pivotaltracker.com/story/show/46839587
<bljenkins>Project fcrepo-bagit-object-serialization build #16: FAILURE in 1 min 5 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-bagit-object-serialization/16/15:51
<cbeer>bljenkins: next time, don't build that project so fast!15:52
<bljenkins>cbeer did you mean me? Unknown command 'next'
Use '!jenkins help' to get help!
<ajs6f>:)
!jenkins help
<bljenkins>Available commands:
abort <job> - specify which job to abort
alias [<alias> [<command>]] - defines a new alias, deletes one or lists all existing aliases
botsnack [<snack>] - om nom nom
build <job> [now|<delay>[s|m|h]] [<parameterkey>=<value>]* - schedule a job build, with standard, custom or no quiet period
cb - list jobs which are currently in progress
comment <job> <build-#> <comment> - adds a description to a build
currentlyBuilding - list jobs which are currently in progress
h [<job>|-v <view>] - show the health of a specific job, jobs in a view or all jobs15:53
health [<job>|-v <view>] - show the health of a specific job, jobs in a view or all jobs
jobs [<job>|-v <view>] - show the status of a specific job, jobs in a view or all jobs
q - show the state of the build queue
queue - show the state of the build queue
s [<job>|-v <view>] - show the status of a specific job, jobs in a view or all jobs
schedule <job> [now|<delay>[s|m|h]] [<parameterkey>=<value>]* - schedule a job build, with standard, custom or no quiet period
status [<job>|-v <view>] - show the status of a specific job, jobs in a view or all jobs
testresult [<job>|-v <view>] - show the test results of a specific job, jobs in a view or all jobs
userstat <username> - prints information about a Jenkins user
<ajs6f>And now we know.
<cbeer>ajs6f: i'm poking around org.fcrepo.session to see about injecting a transaction-generated session15:57
<ajs6f>aye
<cbeer>ajs6f: i think i'm supposed to add logic to InjectableSession?
* kaarefc leaves
<cbeer>or SessionFactory.getSession, maybe
<ajs6f>Hm.
I say SessionFactory.
<cbeer>k15:58
<ajs6f>InjectableSession shold just be concerned with injectiion.
SessionFactory shold keep responsiblity for constrcuting all sessions.
<cbeer>works for me
<ajs6f>That was barmintor original thinking, I think.
Cool.
<pivotal-bot>Chris Beer started "Make existing REST API transactions-aware" https://www.pivotaltracker.com/story/show/4901057116:00
<ajs6f>cbeer: you're not going to wrap every op in a tsaction, right?
<pivotal-bot>Chris Beer edited "Ensure REST API supports workspace identifier prefixes by adding tests" https://www.pivotaltracker.com/story/show/49012389
<cbeer>ajs6f: nope.
<ajs6f>cool.
<cbeer>ajs6f: i'm going to inspect the request path for a tx:* prefix16:01
and if there is one, go fetch it
<ajs6f>right on, brother!
<cbeer>and while i'm at it, i'm going to finish nbanks' ticket
<ajs6f>FINISH HIM!
"2. Sign-up for the Hackfest (taking place on Monday, July 8) and Dev Challenge [6] so we can make sure we have enough beer and pizza to help get you started."16:07
Looks like snacks are taken care of.
http://or2013.net/content/fedora-futures-workshop-starts-100016:09
<bljenkins>Project fcrepo-fixity-corrupter build #101: SUCCESS in 1 min 13 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/101/
<ajs6f>Wow. I hope no one comes expecting to see:
• “Drop-in” JAX-RS resources
• Via dynamic language support for native Ruby, Python and Scala
or
• “Sequencers” that enable services such as metadata extraction, file characterization, generating derivatives (e.g. thumbnails, PDFs), etc.
cause we're going to ask _them_ to build those thing.
* jcoyne wants language support for Clojure16:10
<ajs6f>Hey, as long as it runs on the JVM.
* travis-ci joins16:16
[travis-ci] futures/fcrepo4#696 (master - 1578d77 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/4be6008c3225...1578d77c2a77
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8063865
* travis-ci leaves
<cbeer>ajs6f: what claims are we making about authn and transactions?16:18
<ajs6f>none yet.
havent' even thought about it.
<cbeer>well.. should anyone with the txid be able to act as a proxy to the user that made the original session
<ajs6f>the authn stuff contemplated the things being authn'd as resources with relatvely static names,16:19
<cbeer>or should the agent who made the initial session be the exclusive user to a transaction?
<ajs6f>and txactions have fast-moving shortlived names.
I think the txaction is a resrouce in and of itself.
you have to authn to get at it.
then you have to authn to get at any protected resources16:20
you want to use inside it (the txaction)
does that make sense?
<cbeer>hm. sure.16:22
(and I just found Session#impersonate.. which makes this maybe easier.)
<ajs6f>is that JCR, or MODE?
I'm not sure we want to associate a txaction with its agent.16:23
I think we're taking more responsbility than we want to.
(if we do that)
<cbeer>Tx contain sessions, which are associated with agents.16:24
but impersonate makes all that fine, i think
<ajs6f>I don't think of sessions being associated with agents at all.
they're just containers of jointly valid operations.
dang it. I have really grown to hate JMeter16:37
<cbeer>wow. this might be twice in one day where i've written code that passes its tests16:40
must be time to quit while i'm head16:43
<ajs6f>No, no. You keep working. Maybe you'll make up for all my crap code.
<cbeer>ajs6f: in JCR, workspaces have to exist before you can use them, right?16:52
i can't just repo.login("some random workspace")
<ajs6f>cbeer: yes, to my knowledge
That's why n the JSON config
<cbeer>right
<ajs6f>we set a parameter to say 'you're allowed to make new workspaces'16:53
you do it from the Workspace object, I think?
maybe
<cbeer>yeah. not sure i feel like wiring in some JAX-RS resource to make them though.
guess that's a new ticket anyway
<pivotal-bot>Chris Beer added "Wire in a JAX-RS resource to manipulate JCR workspaces" https://www.pivotaltracker.com/story/show/5167523516:54
<ajs6f>You mean a "workspace management endpoint"? No, that sounds like a major leak from JCR to Fedora.
Never mind.
Well, if we're going to do that
we should figure out what the semantics of a Fedora workspace are.
<cbeer>yeah.16:55
<ajs6f>We can start with the JCR semantics, but I think we should maintain a jot or iota of independence.
* github-ff joins16:56
[fcrepo4] cbeer created adv-session-injection (+1 new commit): http://git.io/J9x2Rw
fcrepo4/adv-session-injection c0c122f Chris Beer: allow injected sessions to be an existing transaction session, or get created within a workspace
* github-ff leaves
<pivotal-bot>Chris Beer added comment: "https://github.com/futures/fcrepo4/pull/82" https://www.pivotaltracker.com/story/show/4901057116:57
* github-ff joins
[fcrepo4] cbeer opened pull request #82: allow injected sessions to be an existing transaction session, or get cr... (master...adv-session-injection) http://git.io/mPWlSw
* github-ff leaves
<pivotal-bot>Chris Beer finished "Make existing REST API transactions-aware" https://www.pivotaltracker.com/story/show/49010571
Chris Beer added comment: "https://github.com/futures/fcrepo4/pull/82" https://www.pivotaltracker.com/story/show/49012389
Chris Beer finished "Ensure REST API supports workspace identifier prefixes by adding tests" https://www.pivotaltracker.com/story/show/49012389
<cbeer>i guess i should prove that transactions work like that
<ajs6f>cbeer: do you want me to look at #82?16:59
<cbeer>ajs6f: please. i'm going to add a transactions IT soon to prove it actually does something useful17:00
<ajs6f>you want me to wait for that? (Doesn't sound like I need to)
<cbeer>ajs6f: probably should wait to merge, but you can probably review it for being sensible now17:01
<ajs6f>k
<cbeer>if it worked, that injectable session stuff is awesome17:02
<ajs6f>cbeer: you made a change at:17:03
The Lawbook Exchange
{sigh{
at https://github.com/futures/fcrepo4/commit/c0c122f793d861d2cf5c400b19af01179183b148#L1R4917:04
that makes sense. Is there not a corresponding change in the authenticated version,
or is there enough info in the HTTP context already there to find the txaction?
<cbeer>ajs6f: i think the authn version always had the servlet request in it
i'll double check
<ajs6f>Oh, cool.
It wouldn't suprise me if it did. Just checking.17:05
<cbeer>yes, it does have it
<ajs6f>Groovy
I hope fasseg made the TransactionService scanned. {just a side thought}17:06
<cbeer>there's a bit of near-duplication in the factory, but i didn't see a good way to deal with it
yes, it's scanned.
<ajs6f>good on him
or you, whoever did it.
I'm still uncertain about this .impersonate thing.17:08
<bljenkins>Project fcrepo-fixity-corrupter build #102: SUCCESS in 32 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/102/
<ajs6f>It implies that only one agent can contribute to a transaction (on the record).
Kind of misses out on some of the power of exposing txactions RESTfully.17:09
Unless maybe I'm misunderstanding the semantics of that method.
Better check.
<cbeer>ajs6f: i understand it to mean the opposite
a session is always associated with a single agent17:10
<ajs6f>I think I may have gotten this thing backwards.
<cbeer>but impersonate allows other agents to contribute to the session
<ajs6f>Wait, no. I do have a genuine disagreement.
I don't think a session is associated with a single agent. It's just an object that collects actions against the repository.
You get an authenticated session
if you want to engage with JCR's authZ.17:11
(Which we don't)
<barmintor>I shudder to think of the authZ situation down the pike if we allow multiple agents to share a transactional session17:12
<ajs6f>What authZ situation?
When we do authz,
it will be at the HTTP layer.
<barmintor>"is this agent allowed to join this transaction"
<ajs6f>Not in the JCR repo.
Just a matter of URLs instead of JCR objects.17:13
Look, I'm not going to push this one. If y'all are really sure that a Session has to be associated with an agent, then let's just do that and move on.
I've got a G&T that is callign to me.17:14
* jcoyne leaves
<cbeer>i'm not sure there's a hurry to resolve this, either. maybe we should hammer out our authz strategy at OR17:15
<ajs6f>+!17:16
+1
<barmintor>+1
<ajs6f>Plenty of drink at PEI.
<cbeer>oh, bah. it was going so well until our API methods saved and closed my transaction for me17:17
<ajs6f>Cause they save and logout the Sesssion?
<cbeer>ajs6f: yeah. i thought we said we'd proxy a transaction session and just null out save and logout
but no
<ajs6f>"Oh, poopy!" — TV's Frank17:18
But wait— in a non-txactional context, how do we then know when to save/logout?
<cbeer>ajs6f: so, here's what i thought we said was a good idea forever ago:17:19
<ajs6f>(I'm sure this turns out to be something I said I would do.)
<cbeer>when you get a session from the tx service, you get a session (wrapped by a dynamic proxy? something else? just a sub-class of Session?) that has a no-op implementation of save() and logout()
(and other methods with some impact outside the scope of a session)17:20
<ajs6f>Ohhhh, yeah.
<cbeer>so we don't need to touch our REST API at all
but we have to be careful with these transaction-aware sessions
<ajs6f>So either we change InjectableSession or we introduce a new type/pseudo-type (dynamic proxy)
Yeah, when do the txaction-aware sessions actually save?17:21
<cbeer>POST /fcr:tx/fcr:commit
<ajs6f>Do they expose a special "Save. No, I mean really, save." method?
supersave()
<cbeer>the tx endpoint knows about the transaction service and can do a run-around on it17:22
<ajs6f>So TxactionService can either supply a savable or unsaveable session?
<cbeer>sure
<ajs6f>(and only the txaction endpoint asks for the saveable version)
<cbeer>getRegularJcrSession()17:23
<ajs6f>but its really the same session, with diffferent behavior enabled.
the same SessionCache in MODE.
<cbeer>yes
<ajs6f>k
<cbeer>so.. is dynamic proxy the right hammer/
<ajs6f>Well, you know how much MODE likes final classes.
I'm willing to be loonies to doughnuts that JcrSession is final.17:24
bet
A dynproxy that just null-opts save and logout is trivial.
I guess the type should live in kernel, near TxactionService.17:25
<cbeer>ScaryTxAwareSessionProxy
<ajs6f>ScaryJavaMagicThatSurprisesMeEveryTimeItWorks
I suggest that the getSession()17:26
method from TxactionService should get the proxy.
<cbeer>agreed.
<ajs6f>And getRegularJcrSession() should be something… I dunno. I guess that's pretty much what we mean. :)17:27
k
You want to make a branch?
Keep in mind that the proxy isn't inherently a type.
So we can just alter TxactionService to produce a proxy on demand.17:28
cbeer: You, uh… you want to swap? I do this txaction stuff, you make jmeter run against fcrepo3 on AWS...?17:31
<cbeer>nooo
<ajs6f>There's a beer in it. Whatever bottle we can find on PEI...17:32
{sigh}
Back to the coal mine.
* jcoyne joins17:33
<ajs6f>Has anyone in the room right now run the current JMeter script against fcrepo3 successfully (i.e. with results of any kind at all)?17:36
* nbanks leaves17:37
<cbeer>whoa! i think it works17:40
<ajs6f>I hate that. It just raises expectations.
<cbeer>but it's the last sprint! if ever there was a time to raise expections.. . it's for the next poor saps who get stuck on this project17:45
<ajs6f>Hello, Next Poor Sap!17:46
<cbeer>(oh yeah -- of course we were doing 100 point sprints!)
hell, we had people working around the clock!
<ajs6f>Singapore time.
I'm going to use that phrase again. "Singapore time." It has a nice, "orientalist", "exotic" flavor.17:47
JMETER! &*^&*^& URGGGRGGR *&*&)!17:48
You are _willfully_ ignoring my honest and sincere intentions.17:49
* github-ff joins17:50
[fcrepo4] cbeer force-pushed adv-session-injection from c0c122f to ea45052: http://git.io/T3BcAQ
fcrepo4/adv-session-injection 6f73036 Chris Beer: allow injected sessions to be an existing transaction session, or get created within a workspace
fcrepo4/adv-session-injection ea45052 Chris Beer: make every resource that uses injectable sessions able to participate in transactions
* github-ff leaves
<cbeer>ajs6f++
ajs6f++
ajs6f++
<ajs6f>WTF? Just my ability to excoriate?
<pivotal-bot>Chris Beer delivered "Make existing REST API transactions-aware" https://www.pivotaltracker.com/story/show/4901057117:51
Chris Beer delivered "Ensure REST API supports workspace identifier prefixes by adding tests" https://www.pivotaltracker.com/story/show/49012389
<cbeer>ajs6f: i'm still thrilled about these injectable sessions.
<ajs6f>Can I inject you into the JMeter committers' group?17:52
<cbeer>you want to do that? nah, just write some beanshell to do it yourself
what do you think this is, a project designed to make it easy to run performance tests on your webapp?17:53
ha.
we're a framework, not an application.
<ajs6f>"BBBEEEEAAANNSSSHHEEEELLLLL!" — Kirk: Star Trek II
<cbeer>someone once asked us for an admin GUI. we made a swing app.
and they liked it! they really liked it!
<ajs6f>Then an Air app.
What now? An iPhone app so my wife can do my work for me?17:54
<cbeer>none of the committers know how it works, though. but we keep it around
<ajs6f>None of the committers know _whether_ it works.
<cbeer>@decide jmeter-users or fcrepo-users
<ajs6f>I cannot imagine the insanity that come across jmetere-users.17:55
I'm two steps away from adding to it
<cbeer>"so, we were going to build fcrepo4. we got really close too.. but in the last 2 weeks, we threw it all away to work on a jmeter test plan instead"17:56
<ajs6f>We had enough time and love for one or the other.
XOR17:57
<aawoods>ajs6f: back... so what is the issue with jmeter?17:58
you need a hand?
<ajs6f>aawoods: Same as ever. It does nothing. Literally nothing.
I need an octopus.
<aawoods>are you using the gui, or just the cmdline?
<ajs6f>Using the gui locally to edit, then pushing (scp) up to the instance to run via cmdline, cause I cannot get X to work from this instance.17:59
It's seems right.
But it ain't.18:00
ec2-174-129-72-48.compute-1.amazonaws.com
The JMeter stuff is in "ubuntu"'s home dir
plans/fedora3.jmx
<aawoods>alright, I am in18:01
<ajs6f>The last time I tried this (running Jmeter against fcrpo4) it worked fine via cmdline, wouldn't work via GUI.
?!?!
So I'm taking no chance.s
ubuntu@ip-10-68-51-141:~$ users18:02
ubuntu ubuntu ubuntu
:)
<aawoods>to you want me to put the x-server on that instance?
<ajs6f>I thought I had, but the usual ssh -X accomplished nothing18:03
If you know it, do it!
<aawoods>looks like the x-server is there18:04
<ajs6f>Right.
<bljenkins>Project fcrepo-fixity-corrupter build #103: SUCCESS in 33 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/103/
<ajs6f>Even more, I think my time-consuming back-and-forth
was fine— I have up there what I thikn I have.
I think this is about JMeter.
I've turned on the thread groups that make sense.18:05
I've altered the URLs in use in the working module to fcrepo3 spec.
And no requests fire at all.18:06
That's what is making me bitter, like the kind of really nice house-made bitters that barmintor expects at a hip NYC bar.
<aawoods>for what it is worth, remote x is working for me on that instance18:07
<ajs6f>I hate you.
But only a little
<barmintor>ajs6f http://www.deathandcompany.com/lounge/
<aawoods>I logged in as:
<ajs6f>"Death and Company". Oh, my goodness.
<aawoods>ssh -i <your-pem> -Y ubuntu@<host>
<ajs6f>Trying...18:08
<barmintor>http://www.timeout.com/newyork/bars/little-branch18:09
<ajs6f>ubuntu@ip-10-68-51-141:~/ff-jmeter-madness$ jmeter &
[1] 5024
ubuntu@ip-10-68-51-141:~/ff-jmeter-madness$ An error occurred: null
null
"null"?
<barmintor>ajs6f: http://brandylibrary.com/sections2007/home.htm
<ajs6f>"Little Branch". That's a nice name for a bar. \
<aawoods>ajs6f: can you put your jmeter command in a text file in ubuntu's home dir?18:10
<ajs6f>Doing it.
<aawoods>not only can you, you are.18:11
<ajs6f>jmeter.sh, and that's to be run from the ff-jmeter-madness dir.
(because of the relative path for the test plan)18:12
<pivotal-bot>Chris Beer added "Make the TransactionService more RESTful" https://www.pivotaltracker.com/story/show/5168054518:13
<ajs6f>"As befits cocktail progenitor Sasha Petraske’s liquid legacy, the drinks at this clubby, low-ceilinged Village rathskeller are nigh perfect." Now that's canonical TimeOut prose.
<pivotal-bot>Chris Beer added comment: "documented on https://wiki.duraspace.org/display/FF/How+to+Use+the+REST+API" https://www.pivotaltracker.com/story/show/49010571
* jcoyne leaves18:15
<ajs6f>barmintor: Does Brandy Library use LCC? Dewey Decimal? Just accession numbers?18:16
Can I get a drink poured in my lap there for asking?
<barmintor>ajs6f: Sasha Petraske's principal bar is a private club with a $400 per annum membership, and rules specifically barring discussing their clientele's celebrity18:17
<ajs6f>Thank goodness! I am _so_ tired of being WATCHED.18:18
<barmintor>Brandy Library actually does use catalog numbers for its bottles
<ajs6f>Good practice, until ALA finally finishes its Spiritous Liquors Standard.
Something seems to be preventing that Working Group from getting work done.18:19
aawoods: Are you seeing what I saw?18:21
<aawoods>yes, nothing in the log except the header
<ajs6f>Yay!
Wait. Nay?
<aawoods>that usually means there is something wrong
<ajs6f>Well, we're not getting numbers,
and that's what we want.18:22
And I can't for the life of me find out why.
If you don't see anything, I will start trying to log JMeter and see what it thinks it is doing.18:23
<aawoods>have you tried to run jmeter locally against the remote aws instance?18:24
<ajs6f>No.
<aawoods>I will try it
I updated your jmeter.sh script
to add debug logging
but to no/little avail
<ajs6f>++aawoods
Yeah, it tells you it isn't doing anything.
* github-ff joins18:25
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/WxoLOA
fcrepo4/master 69476da Chris Beer: Merge pull request #82 from futures/adv-session-injection...
* github-ff leaves
<ajs6f>I could tell you that, just by standing on this street corner and watching.
<aawoods>2013/06/13 18:29:31 INFO - org.apache.commons.httpclient.HttpMethodDirector: I/O exception (java.net.ConnectException) caught when processing request: Connection refused18:30
<ajs6f>Firewall?18:31
<aawoods>no
<ajs6f>http://ec2-174-129-72-48.compute-1.amazonaws.com:8080/fedora/describe
<aawoods>right
<ajs6f>is available to me?
<aawoods>have you tried the create command?
<ajs6f>OHHHHH *(^(^*(!
Wait, no.18:32
I was thinking of the
"no one but localhost" XACML policies that are fcrepo3's default.
But that wouldnt matter whn we're running the jmetere scripts on the same box.
I'll try an object create now.18:33
Locally, and on the box.
Good on the box.18:34
Good locally.
(but I should have checked these things already)18:35
Anyway, that shouldn't stop the script. I set the
"fail in case of error?" setting to "continue".
<bljenkins>Project fcrepo-fixity-corrupter build #104: SUCCESS in 34 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity-corrupter/104/
<ajs6f>Say, that's odd. A default fcrepo3 install _shouldn't_ allow object creation (or any APi-M call) from outside the box.18:36
* travis-ci joins18:44
[travis-ci] futures/fcrepo4#699 (master - 69476da : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/1578d77c2a77...69476dabe6a2
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8069112
* travis-ci leaves
<ajs6f>Does anybody remember where the "load-on-repo-start" policies are in fcrepo3? Is it FEDORA_HOME/server/fedora-internal-use/config?18:49
I don't understand why I was just able to create an object from my local machine. That's not supposed to be possible.
No, pdp/policies.18:50
Steve Bayliss was good at answering these questions.18:51
<aawoods>ajs6f: looks like it is working: local-jmeter-ui --> aws fedora319:03
<ajs6f>aawoods++
<aawoods>I will send you my jmx file
<ajs6f>Can we just go with that? I guess we want to do it on aws to avoid network latency?
anyway, I'll start with that and go from there.
What did you do to the script?
<aawoods>I suspect we can get it to work on the aws instance
the real kicker was disabling: "Fedora Profiling":"Authenticate":"Fedora 4 Describe"19:04
and using the following command to launch jmeter: ./jmeter/bin/jmeter.sh -Ljmeter=DEBUG19:05
<ajs6f>I only see Authenticate as a thread group. The others are request sections?
<aawoods>to see the jmeter logging, select the yellow warning symbol in the top-right of the ui
<ajs6f>Yeah, I would have done that, if I could have used the GUI on the instance.19:06
Oh, wait. You did it locally.
I was too doctrinaire about running it on the box.
<cbeer>ajs6f: looks like jax-rs 2.0 was just released
<ajs6f>Santa Clause is here: CDI FOR EVERYBODY!
That's a clause as in XML clause, of course.19:07
We'll see how fast Jersey catches up. They've had 2.x in the works for months...19:08
<cbeer>ajs6f: it looked like 2.0 was out
<ajs6f>Jersey?
<cbeer>yeah
<aawoods>or "claws"
<ajs6f>Satan Claws should be a Spinal Tap mockumentary about a terrible black metal band.19:09
I thought Jersey had milestones out for 2, but not solid releases.
<bljenkins>Project fcrepo-fixity build #287: FAILURE in 3 hr 20 min: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/287/
Project fcrepo-kitchen-sink build #387: STILL FAILING in 3 hr 19 min: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/387/
<cbeer>ajs6f: 2.0 as of 5 hours ago
<ajs6f>WHY aren't we using it now!?!
<cbeer>i know!?19:10
<ajs6f>We _are_ the bleeding edge.
Actually, the Oauth stuff _is_ using snapshots.
I didn't want to say that while I thought eddies was awake.
<aawoods>ajs6f: running the script: /home/ubuntu/jmeter.sh now works19:11
<ajs6f>G-d bless you.
Okay, now for the R scripts.19:12
<aawoods>I am going to switch gears... back to looking at gregjansen's bags
<bljenkins>Yippie, build fixed!
Project fcrepo-fixity build #288: FIXED in 2 min 56 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-fixity/288/
Yippie, build fixed!19:13
Project fcrepo-kitchen-sink build #388: FIXED in 3 min 10 sec: http://ci.projectblacklight.org/jenkins/job/fcrepo-kitchen-sink/388/
<ajs6f>It makes you wonder… what do other people use for load testing? Do they just suffer through JMeter's unintelligible XML?
<aawoods>you should never need to look at the xml
<ajs6f>No, you shouldn't.
<aawoods>do you look at it?
<ajs6f>I often cannot understand the visual idioms of the GUI.19:14
<aawoods>I only looked at it once, to set an "enabled" from true to false.
<ajs6f>But then, my wife often asks me to tell her what I think of a particular color of paint, and I tend to scream and run away.
<pivotal-bot>Chris Beer edited "Make FedoraTransactions more RESTful" https://www.pivotaltracker.com/story/show/51680545
Chris Beer estimated "Make FedoraTransactions more RESTful" as 2 points https://www.pivotaltracker.com/story/show/51680545
Chris Beer started "Make FedoraTransactions more RESTful" https://www.pivotaltracker.com/story/show/51680545
<ajs6f>I'm just not a visual person.
Now I'm going to go engage with something I understand: food. I'm going to go cook supper. See y'all tomorrow.19:15
<pivotal-bot>Chris Beer unstarted "Make FedoraTransactions more RESTful" https://www.pivotaltracker.com/story/show/5168054519:16
* ajs6f leaves
<aawoods>cheers19:17
<pivotal-bot>Andrew Woods edited "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/5151982519:46
Andrew Woods edited "Run fcrepo4 single-node configuration performance tests" https://www.pivotaltracker.com/story/show/51516137
Andrew Woods edited "Run fcrepo4 single-node configuration performance tests" https://www.pivotaltracker.com/story/show/51516137
Andrew Woods added comment: "Dev Challenge ideas submitted 2013-06-13" https://www.pivotaltracker.com/story/show/5151982519:47
Andrew Woods accepted "OR13 Dev Challenge Proposals" https://www.pivotaltracker.com/story/show/51519825
* nbanks joins20:38
* nbanks leaves20:42
* aawoods leaves21:44
<pivotal-bot>Andrew Woods accepted "Create an integration test that copies projected bagit nodes to internal nodes via Workspace.copy() and conversio..." https://www.pivotaltracker.com/story/show/5141617722:13
* jcoyne joins22:19
<pivotal-bot>Chris Beer edited "Create an integration test that copies projected bagit nodes to internal nodes via Workspace.copy() and conversion of..." https://www.pivotaltracker.com/story/show/5141617722:26
* nbanks joins22:39
* nbanks leaves22:43
* jcoyne leaves22:50
* jcoyne joins22:57
* jcoyne leaves23:07
* jcoyne joins23:35
* jcoyne leaves23:52
* nbanks joins00:27
* kaarefc joins00:56
* kaarefc leaves01:43
* kaarefc joins02:13
* kaarefc leaves02:16
* kaarefc joins02:18

Generated by Sualtam