The IRC channel is #sakai on the freenode network, you can join using a web client at http://webchat.freenode.net/?channels=sakai and you can view the complete logs at http://sakai.iriscouch.com/irc/_design/viewer/index.html
00:05 <GitHub133> [nakamura] thecarlhall pushed 2 new commits to master: https://github.com/sakaiproject/nakamura/compare/bb2580e8ff3e...28fec0296969 00:05 <GitHub133> [nakamura/master] SOLR-75 Use solr-example-resources JAR for faster Solr builds - Ray Davis 00:05 <GitHub133> [nakamura/master] Merge pull request #958 from raydavis/solr-fast-build - Carl Hall 04:21 <GitHub101> [nakamura] zathomas pushed 3 new commits to master: https://github.com/sakaiproject/nakamura/compare/28fec0296969...d7962e703eab 04:21 <GitHub101> [nakamura/master] KERN-3070 Cucumber integration tests to cover files/impl; document pooled content data model - Chris Tweney 04:21 <GitHub101> [nakamura/master] KERN-3070 Make the default log level ERROR - Chris Tweney 04:21 <GitHub101> [nakamura/master] Merge pull request #957 from ctweney/kern-3070-squash - Zach A. Thomas 09:42 <SakaiGithub> [3akai-ux/master] SAKIII-5849 make content authoring more keyboard friendly - jsloane 15:19 <zacht> ping mrvisser 15:28 <MrVisser> hey zacht 15:29 <zacht> mrvisser do you have any objection to letting the OmniTI guys into our oae-performance environment? 15:30 <MrVisser> zacht, not particularly, what are their plans? 15:31 <zacht> mrvisser For starters, they can let us use an instance of their monitoring tool to hook up to it. 15:31 <zacht> mrvisser It's this: http://circonus.com/ 15:32 <zacht> mrvisser And it's my hope that they can offer us some pointers in the load testing arena. 15:33 <zacht> mrvisser To that end, I'd like to get you on the next call we have with them if that's alright. 15:33 <zacht> mrvisser We haven't scheduled it yet. 15:33 <MrVisser> zacht: yea, that would be cool. Keeping in mind right now the environment isn't prepared to run performance tests, I'm having troubles getting a large set of data loaded in. 15:35 <zacht> mrvisser It could be that we need the monitoring in place to tell us why we can't load data fast enough. 15:35 <zacht> mrvisser What ended up happening with the data load you had running for 15 or more hours? 15:41 <zacht> dgillman G'morning! 15:41 <dgillman> zacht: hey! 15:42 <zacht> dgillman Question about https://github.com/sakaiproject/nakamura/pull/930 15:42 <dgillman> ya? 15:43 <zacht> dgillman It's got merge conflicts. Is that a known issue? I know you said you had found the merge with raydavis to be not too bad. 15:43 <MrVisser> zacht: everything just froze right up. my machine had ~5M of ram available, I think postgres was trying to create processes but it couldn't, and the app server needed more memory, I think because solr was running embedded. a bunch of requests started failing and the attempt was basically toast. 15:44 <dgillman> zacht: has ray's stuff been merged already? 15:44 <raydavis> dgillman, yes indeed. 15:44 <zacht> dgillman Yeah. 15:45 <dgillman> zacht: ok, let me pull his changes and I'll resubmit the PR with the merge conflicts resolved ;) 15:45 <zacht> dgillman Here's the raydavis PR for reference: https://github.com/sakaiproject/nakamura/pull/948 15:45 <dgillman> zacht: thanks 15:46 <raydavis> mrvisser, as of 1.2 when running a data load on an embedded-Solr system we found that we needed to bump the app up to 2GB memory (!). 15:46 <raydavis> Don't know if that's a factor for you. 15:46 <zacht> mrvisser I think we still have big issues with event handlers making more work for the app server than is strictly necessary during a data load. 15:47 <MrVisser> raydavis: I had given it ~2.5G.. I was loading 5000 users (amounted to 86000 contacts), and 12500 worlds. 15:47 <raydavis> ACTION concurs with zacht. 15:47 <zacht> mrvisser Obviously, the content indexing listener is needed, 15:47 <zacht> mrvisser But anything having to do with counts or activities, etc. is just so much flailing. 15:48 <zacht> mrvisser I think the OSGi to JMS bridge can be disabled during that time 15:49 <MrVisser> zacht: wouldn't that disable creation of messages? 15:49 <raydavis> zacht, IIRC the content listener can also create contention with itself (child nodes smacking against the same target as parent nodes) 15:49 <raydavis> Although some of that was improved post-1.2. 15:49 <MrVisser> also, aren't activities important data to have in the system when analyzing performance? 15:50 <zacht> mrvisser yes, but there might be another way to write a big bag of activities than by making the system do the _actual_ activity. 15:52 <zacht> mrvisser The point of the data loader is to restore a lifelike state, right? We can take shortcuts to do that. In fact, after it has succeeded at least once, the shortcut should be something like tar -xzf my_awesome_data.tar.gz 15:53 <MrVisser> zacht: I have those scripts in place to restore from raw data. 15:53 <MrVisser> zacht: I just need to get data in now. 15:53 <zacht> mrvisser cool. 15:54 <zacht> mrvisser So do you have the same problems when you try to data load on our five-node environment? 15:55 <raydavis> mrvisser, I wonder if the final result would come faster if the individual updates happened more slowly. We've sometimes had to add delays to make load tests stop crashing early. 15:55 <MrVisser> zacht: not sure what I'm seeing yet on the 5-node environment. I just tried, and then all of a sudden the app server node didn't have access to the postgresql database anymore because some IP changed. Also, the Felix container seems to sporadically reboot itself sometimes. 15:56 <raydavis> That I have not seen. 15:56 <zacht> mrvisser That doesn't sound web scale. :-| 15:56 <MrVisser> I'm wondering if it's an amazon thing. I asked Kyle and Erik about it yesterday, waiting to hear back. Maybe they have process load throttles or something… ?? 15:57 <MrVisser> I had to reconfigure postgres to allow the new IP, but then I don't want to leave another data-load going and *boink*.. and not during load tests either, so it would be good to get to the bottom of those things. 15:59 <MrVisser> zacht raydavis: my next attempt will be to do it in chunks. e.g., just do the users, back things up. Then do a batch of the contacts, back it up. Then do the worlds, back it up, then group membership, done. 16:01 <MrVisser> so, right now I'm working on getting the system back in order after the IP change and to reload the data. 16:35 <denbuzze> stuartf: would it be ok if I put SAKIII-6035 back on pending merge? 16:35 <dgillman> zacht: FYI, I've fixed the merge conflicts but want to run the integration tests to ensure all is copacetic. I'll ping you when the PR is updated 16:36 <zacht> dgillman++ 16:38 <stuartf> denbuzze: sure, I just forgot to do it 16:39 <stuartf> there's a lot of excitement around here, since they're filming a movie in my building 16:39 <denbuzze> stuartf: are you in it ^^? 16:39 <stuartf> makes it easy to get distracted in the middle of updating a ticket 16:40 <denbuzze> stuartf: btw np at all, thanks for making this patch, I'm reviewing it right now. 16:40 <stuartf> I'm not in it, but I just walked through the set between takes to get my lunch, I walked right past Owen Wilson 16:41 <denbuzze> stuartf: sounds pretty awesome, next to our building they're during some construction, way less exciting 16:45 <PhysX> stuartf: any idea what movie it is? 16:46 <PhysX> if it's about the OAE I'd go and see it 16:46 <stuartf> it's The Internship http://www.imdb.com/title/tt2234155/ 16:46 <stuartf> it's about two guys working at Google 16:47 <stuartf> they've put all kinds of crazy stuff in the building to make it look like Google HQ 16:47 <PhysX> stuartf: I guess it's just not cool enough 16:50 <zacht> stuartf Did they give you a ball pit to jump in? 16:51 <stuartf> no, but there's a slide we're not allowed to use 16:51 <stuartf> and a full size replica of space ship one 16:52 <stuartf> and bikes http://i.imgur.com/XUdSR.jpg 16:54 <PhysX> stuartf you should go sit out there with your laptop and subtly display OAE logo's 16:54 <stuartf> lol 16:56 <dgillman> zacht: PR has been updated with merge conflicts resolved: https://github.com/sakaiproject/nakamura/pull/930 16:56 <zacht> dgillman awesome, thanks! Will review shortly. 17:00 <denbuzze> stuartf: one thing I just noticed is that the line numbers seem to be off by 3 - http://cl.ly/IXWK - any idea what might cause this? It only happens for lines that are put after the License declaration 17:02 <stuartf> denbuzze: I'm not sure exactly why that would happen, but the script keeps track of which line it's on, so it could be getting out of sync somewhere 17:09 <denbuzze> stuartf: makes sense, btw what should fail/pass for the double/single quotes? 17:12 <stuartf> denbuzze: for the quotes it just looks for double quotes on a line before a single quote 17:13 <stuartf> that one could probably be made a little better 17:15 <denbuzze> stuartf: that would be nice, my previous example - $('#newcreategroup_can_be_found_in option[value="' + defaultaccess + '"]', $rootel); is still failing, 17:17 <stuartf> denbuzze: ah, I see what I did wrong there 17:20 <stuartf> denbuzze: when the line numbers are wrong, are they too high or too low? 17:21 <denbuzze> stuartf: 3 too low - have a look at http://cl.ly/IXWK - it only happens after the license statement 17:22 <denbuzze> stuartf: btw I really like the test for the `,` at the end of a variable declaration 17:22 <denbuzze> stuartf++ 17:23 <denbuzze> stuartf: the prepare-message hook is now also working on Mac :) - I'm writing these things up in the jira ticket 17:25 <stuartf> denbuzze: yeah, pretty much everything from that list should be fixed 17:25 <stuartf> other than wehere I've messed up, like with the quotes check 17:29 <denbuzze> stuartf: posted a comment on SAKIII-6035 - seems like we're almost there 18:24 <stuartf> denbuzze: I just pushed a commit that should fix the stuff you commented on in the ticket 18:28 <denbuzze> stuartf: awesome, thanks - I'll look at it asap. 18:57 <thecarlhall> mrvisser1, did you put kern-3094 as a kern bug instead of sparse to keep it in the subtask hierarchy? 19:01 <mrvisser1> thecarlhall: that's basically it, also wasn't immediately sure where the fix would fall. 19:01 <thecarlhall> mrvisser1, it's in sparse. I'd like to move it to the sparse tracker but don't want to break up what you've got going on 19:01 <thecarlhall> mrvisser1, it's a pretty quick fix (patch ready) that we've done a fair number of times 19:02 <mrvisser1> thecarlhall: feel free to clone it, but please keep a copy and resolve it as dupe or something so I can track completion of the issue. 19:02 <zacht> mrvisser1 Is that going to be part of 1.3.7? 19:02 <thecarlhall> zacht, yeah 19:02 <mrvisser1> zacht: no.. 19:02 <mrvisser1> hehe 19:03 <thecarlhall> no? 19:03 <mrvisser1> thinking it might be best to control the scope of that fix. 19:11 <GitHub71> [sparsemapcontent] zathomas created 1.3.7-release (+1 new commit): https://github.com/sakaiproject/sparsemapcontent/commit/7a40b6a9ec4a 19:11 <GitHub71> [sparsemapcontent/1.3.7-release] [maven-release-plugin] prepare release org.sakaiproject.nakamura.core-1.3.7-RC1 - Zach A. Thomas 19:11 <GitHub14> [sparsemapcontent] zathomas pushed 1 new commit to 1.3.7-release: https://github.com/sakaiproject/sparsemapcontent/commit/af18b74d47ea873b215fe49144b2de93cad904e5 19:11 <GitHub14> [sparsemapcontent/1.3.7-release] [maven-release-plugin] prepare for next development iteration - Zach A. Thomas 19:14 <GitHub168> [nakamura] zathomas pushed 7 new commits to master: https://github.com/sakaiproject/nakamura/compare/d7962e703eab...0be8a5af19a5 19:14 <GitHub168> [nakamura/master] KERN-2880 deprecated calls using SlingHttpServletRequest in SolrSearchFactoryImpl and SolrResultFactoryImpl - Duffy Gillman 19:14 <GitHub168> [nakamura/master] KERN-2880 updated vestigal SparseResultSetFactory - Duffy Gillman 19:14 <GitHub168> [nakamura/master] KERN-2880 converted QueryHandler classes to respond to getQuery() and deprecated PropertyProvider and RequestProcessor methods - Duffy Gillman 19:19 <MrVisser> zacht thecarlhall: sorry, internet is in flux. 19:20 <MrVisser> thecarlhall: do you have objections to not include that fix in 1.3.7, and instead wait for a 1.3.8 release maybe with OAE 1.5? 19:21 <thecarlhall> mrvisser, I think I'm confused on release numbers. Is 1.3.7 not what is going to be in oae 1.5.0? 19:22 <thecarlhall> mrvisser, I honestly don't care when the fix gets packaged. 19:24 <MrVisser> thecarlhall: looks liek we'll be expediting a sparse release before OAE 1.5 pending some testing, so we'd like to control the scope of 1.3.7 19:25 <thecarlhall> right on. works for me 19:30 <zacht> thecarlhall For all we know, we'll be bundling 1.3.12 with OAE 1.5 :-| 19:30 <thecarlhall> um, sweet? 19:30 <zacht> Oh, yeah, it's gonna be hott. 19:35 <zacht> ping lancespeelmon 19:35 <lancespeelmon> pong 19:35 <zacht> lancespeelmon As a sakaiproject JIRA admin, have you ever had to perform a reindex? 19:36 <lancespeelmon> no I don't think so 19:36 <lancespeelmon> I have run one from time to time 19:36 <zacht> lancespeelmon I ask because there are some new Scrum capabilities that won't become available until the next reindex. 19:36 <lancespeelmon> but generally every jira instance I use always says a reindex is needed :) 19:37 <lancespeelmon> yeah just run it :) 19:37 <zacht> lancespeelmon I wonder how long it takes. 19:37 <lancespeelmon> not very long 19:37 <lancespeelmon> maybe 5 min 20:11 <zacht> ctweney I think I'm going to go here for my next insights into ruby: http://my.safaribooksonline.com/book/web-development/ruby/9780321700308 20:41 <thecarlhall> zacht, with the addition of cobertura for integration tests, where do you think the balance will be between integration and unit tests? 20:50 <zacht> thecarlhall ideally, we'd have high coverage by both. They each have their strengths. Unit tests are good for checking boundary conditions and exception handling. 20:50 <zacht> Unit tests are also better for testing while you code. 20:51 <thecarlhall> agreed but it seems like we'll have a lot of overlap 20:51 <thecarlhall> I think history shows that we tend to favor one type of testing in the long run 20:52 <zacht> Well if you were forcing me to choose, I would choose integration tests, because they're how you know your application is fully functional. 20:53 <thecarlhall> zacht, likewise especially now that we can see if the integration tests are flexing all areas of the application 20:55 <zacht> thecarlhall I think the integration tests are going to be much more useful in the coming months. I think it makes sense to organize them by functionality, rather than by jira ticket number. 20:56 <thecarlhall> zacht, +1 * a million to that. arranging them by jira # always irked me 22:36 <SakaiGithub> [sakai-widgetlibrary] christianv pushed 10 new commits to master: https://github.com/sakaiproject/sakai-widgetlibrary/compare/1ce485f40135...c9256509a483 22:38 <SakaiGithub> [sakai-widgetlibrary] christianv pushed 1 new commit to master: https://github.com/sakaiproject/sakai-widgetlibrary/commit/892033fb8b89b1dcdae1a3b1c18df446b96c1391 22:38 <SakaiGithub> [sakai-widgetlibrary/master] NOJIRA - remove the overflow:hidden and use a <hr instead - Christian Vuerings 22:57 <SakaiGithub> [sakai-widgetlibrary] christianv created OAEWDGT-188 (+4 new commits): https://github.com/sakaiproject/sakai-widgetlibrary/compare/a98a0c6d45b6^...eadcf5b7abc8 22:57 <SakaiGithub> [sakai-widgetlibrary/OAEWDGT-188] OAEWDGT-188 - Fix issue with the thirdparty lhnav - Christian Vuerings 22:57 <SakaiGithub> [sakai-widgetlibrary/OAEWDGT-188] OAEWDGT-188 - Add navigation items to the helloworld tutorials - Christian Vuerings 22:57 <SakaiGithub> [sakai-widgetlibrary/OAEWDGT-188] Merge remote-tracking branch 'sakaiproject/master' into OAEWDGT-188 - Christian Vuerings 22:58 <SakaiGithub> [sakai-widgetlibrary] christianv pushed 4 new commits to master: https://github.com/sakaiproject/sakai-widgetlibrary/compare/892033fb8b89...eadcf5b7abc8 22:58 <SakaiGithub> [sakai-widgetlibrary/master] OAEWDGT-188 - Fix issue with the thirdparty lhnav - Christian Vuerings 22:58 <SakaiGithub> [sakai-widgetlibrary/master] OAEWDGT-188 - Add navigation items to the helloworld tutorials - Christian Vuerings 22:58 <SakaiGithub> [sakai-widgetlibrary/master] Merge remote-tracking branch 'sakaiproject/master' into OAEWDGT-188 - Christian Vuerings 23:25 <SakaiGithub> [sakai-widgetlibrary] christianv pushed 6 new commits to master: https://github.com/sakaiproject/sakai-widgetlibrary/compare/eadcf5b7abc8...c26db6257989 23:25 <SakaiGithub> [sakai-widgetlibrary/master] OAEWDGT-187 Added page describing the widget submission and review process. - Dean Angerer _______________________________________________ oae-dev mailing list oae-dev@collab.sakaiproject.org http://collab.sakaiproject.org/mailman/listinfo/oae-dev
