16:00:15 #startmeeting MD-SAL Interest Call 16:00:15 Meeting started Tue May 26 16:00:15 2015 UTC. The chair is colindixon. Information about MeetBot at http://ci.openstack.org/meetbot.html. 16:00:15 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:15 The meeting name has been set to 'md_sal_interest_call' 16:00:18 #topic agenda bashing 16:00:30 #chair catohornet 16:00:30 Current chairs: catohornet colindixon 16:01:36 #info Moiz says so far he’s seen very little assigning of bugs during this meeting, mostly just discussing bugs 16:02:10 #info catohornet says she agrees, but she doesn’t have a good idea how to get more participants 16:03:50 #info ttkacik_mobile says that his approach is to do some hand-holding, but we’re not getting the attendance here 16:06:36 #action colindixon to work with phrobb to try to get newcomers to attend this meeting to get involved 16:06:56 #link https://wiki.opendaylight.org/view/Simultaneous_Release:Lithium_Release_Plan the lithium release plan for dates 16:07:59 #info colindixon says that in theory, anything can be fixed for the next 4 weeks, the formal release is on 6/25 16:08:48 #info pantelis asks if we should start to push some of these bugs to “not imporant enough to be fixed in Lithium” 16:09:50 #info catohornet says she’d love to see that prioritization, but doesn’t have any great ideas there, it’s the right question though 16:11:28 #info it seems like this bug scrub is sort of the catch-all place because the critical bugs are usually already triaged and being worked on by the “core” developers 16:14:41 #info bugs that make it to this meeting are (1) low-priority bugs that the “core” team hasn’t assigned yet, (2) possibly high-priority bugs that have slipped through the cracks 16:17:21 #info there are other bugs that don’t have a target-milestone which need to be scrubbed and give one so that our current queries, e.g., 3160 and 3155 16:20:27 #info there is a suggestion to attack things in this order: (1) unassiged bugs that are in our query, (2) critical bugs in the query just to make sure, (3) the list of bugs that don’t have a milestones set 16:20:32 #topic bug scrub 16:20:58 #info it appears as the only non-clustering, uassigned bug is the upgrade bug (3160) 16:21:27 #link https://bugs.opendaylight.org/show_bug.cgi?id=3160 this is the preserving user-config upgrade bug 16:23:04 #info pantelis has been working on part of this and testing Maros’s patch for bug 2976 16:23:18 #link https://bugs.opendaylight.org/show_bug.cgi?id=2976 the bug specific to dealing with netconf connector config 16:25:03 #info pantelis gives a lot of detail about what he’s done 16:30:38 https://github.com/nilok/odlutils 16:32:42 #info pantelis: can project team leads if they have changed their yang model 16:34:34 #topc data migration 16:34:37 #info colindixon: Ask projects to provide Data compatibility between beryllium and lithium 16:34:47 #info we don’t have a good story about migrating data back and forth 16:36:29 #info ttkacik1 says that RESTCONF can’t handle revisions, it always usese the new revisions which has some issues meaning that you can’t migrate dara via RESTCONF 16:38:14 #action colindixon to produce a script to catalog how models have changed since helium 16:39:22 #info moiz asks if the migration code should be part of the app or separate, ttkacik1 argues separate because we should avoid using it at all for clean Ltihium installs 16:40:01 #info moiz’s idea was to launch the controller in a “migration” mode, which upgrades all the data, then shuts down and relaunches with clean Lithium 16:51:48 #topic more config migration discussion 16:52:11 #info ttkacik1 notes that we tried to make changes to config yang backward compatible, but they might load the old versions of services even if they are “compatible" 16:52:50 #info pantelis agrees and says we probably want to enoucrage most people to blow away most of the config so that they get the new services and wiring 16:53:07 #info pantelis argues that we should separate config from wiring as we move forward 16:56:13 #info ttkacik says we should discuss mdsal plans for the beryllium release 16:58:31 #info there’s a Beryllium target milestone to start opening bugs 16:58:38 #info colindixon lists topic we need to think about for beryllium. Example Tooling. We should start a list of Beryllium focus areas and start discussing them. 16:58:47 #info colindixon has some ideas around focus areas, we can keep them 16:59:19 #info odl-casey says we really need to focus on making sure we can get source for third-party packages 16:59:34 #info casey cain says working with 3rd party packages need to be built with src jar files. 17:00:57 #inifo casey casin If the 3rd party package doesn't have the source we can't use because we can't redistribute the code. We can't release without the source for 3rd party packages. 17:01:04 https://docs.google.com/spreadsheets/d/1q_vxUjo4gIxpyAGq43Yiu77o1GHdLaSlxVMysd6G5bQ/edit?usp=sharing 17:06:03 #topic talking about 3rd party libarires and source 17:06:29 #info everyone can leave at this point since it’s not an MD-SAL topic 17:07:27 #action colindixon to make sure that the no 3rd-party librarires for Beryllium have the source or else projects will be dropped 17:08:00 #info ttkacik1 says the trick is to do use the CLM process to at least start this 17:09:40 #info colindixon Better tooling would help this effort. 17:15:02 #action colindixon to e-mail odl-casey, ttkacik1, zxiiro, and phrobb about 3rd-party source, the last thing to do is to open blocker bugs against projects that use them 17:16:12 #link https://docs.google.com/spreadsheets/d/1q_vxUjo4gIxpyAGq43Yiu77o1GHdLaSlxVMysd6G5bQ/edit?usp=sharing this is a list fo what is missing in both our local repo and in maven central 17:16:38 #info this list came from zxiiro and we’re going to produce a more accurate list tomorrow 17:16:53 Sorry, I'm trying to get to the airport. My only question today is what's the status of openflowplugin branch cutting 17:17:10 zxiiro: good question, abhijitkumbhare? 17:18:35 #info zxiiro is working the orbit and and eclipse jars 17:19:09 #info odl-casey says the two bigs ones he can’t find are: com.sun:tools:jar:sources:1.5.0:system and xom:xom:jar:sources:1.0:test 17:20:05 https://drive.google.com/a/linuxfoundation.org/file/d/0B1vn4mVLKKmfaU9Zdl9ZdlVyTjA/view?usp=sharing 17:20:35 I'm pretty sure com.sun.tools comes from inside the JDK. It's the "tools.jar" file I believe in a packaged JDK 17:22:14 I'm not sure how to get the sources from that though. I guess if you compile your own JDK maybe 17:23:03 I looked though the package and didn't see it. But I was using Mac's crappy search instead of cli find. 17:23:50 https://wiki.opendaylight.org/view/Simultaneous_Release:Lithium_Release_Plan#Lessons_from_Hydrogen.2C_Helium_and_Lithium_for_future_releases 17:24:01 odl-casey: as far as I understand every JDK provider has their own too. but we use OpenJDK on the infrastructure servers 17:24:25 https://wiki.opendaylight.org/view/Simultaneous_Release:Lithium_Release_Plan#Project_Status 17:30:47 #link https://wiki.opendaylight.org/view/Simultaneous_Release:Lithium_Release_Plan#Lessons_from_Hydrogen.2C_Helium_and_Lithium_for_future_releases this is where we can start to catalog what needs to change strategically 17:31:08 #link https://docs.google.com/spreadsheets/d/1KPpO9LH539Vlcoa4RvLa6PPCdLifi5JD-ihRhlybqeo/edit#gid=676729675 this is where to track tactical issues 17:31:46 #chair ttkacik1 17:31:46 Current chairs: catohornet colindixon ttkacik1 17:32:01 ttkacik1: that way you can do the #endmeeing 17:32:18 #endmeeting