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

Using timezone: Eastern Standard Time
* ksclarke leaves01:11
* kaarefc joins02:31
* nbanks joins04:10
* nbanks leaves05:03
* nbanks joins06:04
* fcrepo-bot joins06:34
* fcrepo-bot leaves07:41
* jonathangee joins08:03
* kaarefc leaves09:13
* gregjansen joins09:31
re: scalability and large numbers of children, this is from the Modeshape 3.2 release notes: "Larger content. ModeShape 3 has been designed to store and access the content so that a node can have hundreds of thousands (or more!) of child nodes (even with same-name-siblings) yet still be incredibly fast." Presumably we haven't run into problems on the Mode level?09:33
<awoods>gregjansen: We have run into performance slowdowns with only a couple of thousand direct children of a node. Therefore, we have created a hierarchical structure of children. But I am glad ModeShape's design at least thinks about the issue. We are on ModeShape 3.409:39
<gregjansen>awoods: yep, not if these release notes necessarily apply to projected nodes. Scalability there necessarily depends on non-Modeshape systems and this claim may not even apply to their connector SPI.. However, I am putting together some fedora talking points for upcoming AP trust meeting and came across this.. Love to have your help crafting a message for this meeting (next tuesday I have 15 mins)09:45
* ksclarke joins09:46
<awoods>gregjansen: We can organize sometime this week for a chat.
<gregjansen>awoods: great! I am out this Friday09:47
<awoods>gregjansen: I will jump on the stand-up hangout a few minutes early today if you want to organize.09:48
<gregjansen>awoods: great, let's do that
say 10 mins early?
<awoods>sure
<pivotal-bot____>Esme Cowles added comment: "FedoraBackupFSIT was the culprit. Removing this IT fixes the build on MacOSX, so I've added that to the pul..." https://www.pivotaltracker.com/story/show/5758941410:05
Andrew Woods added "FedoraBackupFSIT failure on MacOSX" https://www.pivotaltracker.com/story/show/5799756410:20
Andrew Woods added comment: "Thanks for tracking this down to a specific test. Although removing the test allows the build to pass, it d..." https://www.pivotaltracker.com/story/show/5758941410:21
Andrew Woods finished "fcrepo-http-api build failures on MacOSX" https://www.pivotaltracker.com/story/show/57589414
<mikeAtUVa>About the scalability: could our performance slowdowns be because of the really inefficient step or populating a node with all the properties from its children?10:31
<pivotal-bot____>Esme Cowles added comment: "The test that causes the failures just sets a System property, runs an IT that succeeds, and then unsets the..." https://www.pivotaltracker.com/story/show/5799756410:33
<awoods>mikeAtUVa: In the next couple of sprints we need to put heavy focus on performance and scalability. I can see such experiments as not loading child properties as a point of investigation.10:42
<gregjansen>awoods: hangout real slow today10:51
<pivotal-bot____>Gregory Jansen added "Setup an auth filter in test container for integration tests" https://www.pivotaltracker.com/story/show/5800340411:22
Gregory Jansen edited "Setup an auth filter in test container for integration tests" https://www.pivotaltracker.com/story/show/5800340411:23
Gregory Jansen estimated "Setup an auth filter in test container for integration tests" as 2 points https://www.pivotaltracker.com/story/show/58003404
Gregory Jansen started "Setup an auth filter in test container for integration tests" https://www.pivotaltracker.com/story/show/58003404
Gregory Jansen added comment: "The filter is set up, but only partially working. Somewhere the security context is lost and I am trackin..." https://www.pivotaltracker.com/story/show/5800340411:24
* ermadmix joins11:49
* jonathangee leaves12:41
* ermadmix leaves13:06
* jonathangee joins13:16
* jonathangee leaves
* jonathangee joins13:20
<cbeer>awoods: is https://wiki.duraspace.org/download/attachments/34660715/dc-architecture-2013-09.pdf?version=1&modificationDate=1379105230853 sharable?13:24
<awoods>cbeer: on a call
cbeer: I do not think it is ready for prime-time13:29
<cbeer>ok. do we have any other stack diagram floating around?
<awoods>cbeer: check email13:30
<cbeer>thx
* jonathangee leaves13:43
<barmintor>awoods: me too, please? Giving a "state of FF" preso this afternoon.13:53
<cbeer>awoods: i put it on the wiki. hopefully by sending it, he meant that that slide was sharable.13:54
if not, i guess it's easy enough to purge.13:55
* ermadmix joins13:57
<awoods>barmintor: Yes, this side is fine for use: https://wiki.duraspace.org/pages/viewpageattachments.action?pageId=33237722&metadataLink=true14:00
<cbeer>awoods: slide #4 in the deck would also be pretty useful, but I agree that it probably needs more work/context/etc/etc.14:03
<awoods>cbeer: on a call
<cbeer>are you ever not on a call?14:04
<awoods>sometimes14:05
cbeer: Feel free to extract slide 5
or 4
<cbeer>thanks14:06
* jonathangee joins14:11
<gregjansen>If your LOC schemas do not resolve, the shutdown has apparently impacted loc.gov14:22
* nbanks leaves14:32
<barmintor>awoods: Are there any shceduled developers that are not represented on the beta schedule?14:44
<awoods>barmintor: no
<barmintor>awoods: So 13 developers part time over course of beta?
<awoods>sounds right14:45
<barmintor>thanks
<awoods>barmintor: please share your presentation when you have a good draft.
* jonathangee leaves14:46
<barmintor>awoods: it's really just some slides summarizing the Beta sprint reports, and some resource allocation.14:48
* jonathangee joins
<awoods>barmintor: ok
<barmintor>By my count, FCR4 has 52/9 =~ 6 FTEs of developers scheduled in addition to awoods
<awoods>no need to share, if you do not think it is worthwhile.
barmintor: have you done a similar calculation for F3?14:49
<barmintor>awoods: Did we ever schedule devs for F3?
(not being silly, serious question)
<awoods>barmintor: no, but there were names associated with releases.14:50
<barmintor>Hmm.
It would be hard to go back, although I was more-or-less an FTE on F3 for a while
I'll see what I can do, that would be a useful comparison14:51
<awoods>barmintor: I see there being a significant shift in how Fedora development has shifted from F3 to F4, namely, F3 consisted of developers engaging out of self-interest. F4 consists of developers that are supported by their institutions and management. I feel like this change is a reflection of the heightened awareness within the Fedora community of how we can sustainably ensure Fedora continues to meet the collective needs.15:07
<barmintor>+1
even if you just called me self-interested15:08
:)
<awoods>you get my meaning
<barmintor>of course
I'm just a little punchy
<awoods>even if you are self-interested
* pivotal-bot_____ joins15:21
* cbeer leaves15:27
* pivotal-bot____ leaves
* cbeer joins15:32
* jonathangee leaves15:48
* jonathangee joins15:49
* jonathangee leaves15:54
* awoods leaves16:08
* nbanks joins16:33
<pivotal-bot_____>Mike Durbin edited "document and implement fedora 3 properties stretegy" https://www.pivotaltracker.com/story/show/5654801816:39
* github-ff joins16:43
[fcrepo-fedora3-federation-connector] mikedurbin opened pull request #7: Added fedora 3 object and datastream properties. (master...56548018-fedora-3-content-modeling) http://git.io/6pBCTQ
* github-ff leaves
<pivotal-bot_____>Mike Durbin added comment: "Pull request: https://github.com/futures/fcrepo-fedora3-federation-connector/pull/7" https://www.pivotaltracker.com/story/show/5654801816:44
Mike Durbin finished "document and implement fedora 3 properties stretegy" https://www.pivotaltracker.com/story/show/5654801816:45
Mike Durbin added "Implement special handling for DC, RELS-EXT and RELS-INT datastreams for federated fedora 3 content." https://www.pivotaltracker.com/story/show/5803426616:47
Mike Durbin added "Implement versioning for fedora 3 federated content." https://www.pivotaltracker.com/story/show/5803443616:49
Mike Durbin started "Present entire fedora 3 repository in a manageable hierarchy." https://www.pivotaltracker.com/story/show/5776232616:50
* gregjansen leaves16:59
<pivotal-bot_____>Eric James delivered "Identify bottlenecks" https://www.pivotaltracker.com/story/show/5571982217:24
Andrew Woods added comment: "Can you describe your outcomes (at least in summary) here? What warrants completion of this ticket?" https://www.pivotaltracker.com/story/show/5571982217:34
Eric James added comment: "I described it on the confluence page: https://wiki.duraspace.org/display/FF/Design+-+Large+Files in the sect..." https://www.pivotaltracker.com/story/show/5571982217:50
* ksclarke leaves18:24
* ermadmix leaves18:25
* pivotal-bot_____ leaves19:47
* pivotal-bot_____ joins19:48
* nbanks leaves20:27
* nbanks joins21:33
* nbanks leaves21:38
* ksclarke joins21:39
* jonathangee joins22:31
* jonathangee leaves22:38