16:01:11 #startmeeting tsc 16:01:11 Meeting started Thu Aug 23 16:01:11 2018 UTC. The chair is skitt. Information about MeetBot at http://ci.openstack.org/meetbot.html. 16:01:11 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:01:11 The meeting name has been set to 'tsc' 16:01:20 #topic Roll-call and agenda-bashing 16:01:28 TSC *members* and proxies, please #info in 16:01:35 Project representatives, please #project in 16:01:40 #link Agenda: https://wiki.opendaylight.org/view/TSC:Meeting#Agenda 16:01:43 #info skitt 16:01:45 #info anipbu 16:01:57 #project odlparent controller netvirt ovsdb 16:02:02 #project usc 16:02:29 #info gzhao 16:02:42 #info Thanh Ha 16:02:47 #info rovarga 16:03:15 #project odlparent yangtools mdsal controller infrautils coretutorials honeycomb/vbd bgpcep 16:03:22 #info faseelak proxy for Abhijit 16:03:30 #info jamoluhrsen 16:03:32 #topic Impromptu infrastructure update 16:03:41 #info we need to disable all our CLM jobs 16:03:42 #project serviceutils infrautils coe genius 16:04:01 #info we found out that the way we use the jobs doesn't respect Sonatype's ToS 16:04:25 #info LuisGomez 16:04:41 #info vina_ermagan 16:04:48 #project lispflowmapping 16:04:56 #info Anil Vishnoi 16:04:56 #project distribution 16:05:17 #info vina_ermagan proxy for lori 16:05:22 #project openflowplugin 16:06:16 #info gzhao 16:06:34 #topic General topics 16:07:50 #undo 16:07:50 Removing item from minutes: 16:07:53 #topic Action items 16:08:20 #info LuisGomez provided a patch for the managed project documentation to document the steps involved in a release 16:08:24 #link https://git.opendaylight.org/gerrit/#/c/75312/ 16:09:23 #info CaseyLF followed up with Lisa re marketing concerns, Lisa will join the call in a few minutes 16:09:57 #info OPNFLWPLUG-1022 has a fix which is merged 16:10:21 #topic ODL Parent updates 16:10:36 #info ODL Parent 4 is nearly ready, waiting for a Karaf release and a couple of bug fixes 16:10:48 #info YANG Tools is ready for Neon, MD-SAL is 50% there 16:11:06 #info the integration window patches for kernel projects are in progress 16:11:39 #topic ONAP CVE reports 16:12:11 #info apparently some of our third-party tooling has open CVEs 16:12:56 #info but more importantly, CaseyLF would like us to formalise security relationships with our sister communities 16:15:09 #link our vulnerability management process: https://wiki.opendaylight.org/view/Security:Vulnerability_Management 16:17:37 #info the conversation started because ONAP is shooting for silver CII and needs security issues reported in CLM to be resolved in ODL 16:18:29 #info shague 16:18:34 #info realistically, ONAP can get silver CII only if ODL gets it too, which means investment on our end, hopefully with help from ONAP 16:20:31 #info CaseyLF thinks our CLM reports might be different, which would explain why ONAP sees more issues than we do 16:21:30 #info other LFN projects share a CLM instance, we have our own 16:22:12 #action zxiiro to open a bug to migrate our instance to the shared infra 16:22:15 #undo 16:22:15 Removing item from minutes: 16:22:25 #action zxiiro to open a bug to migrate our CLM instance to the shared infra 16:26:33 #info so far the security team aren't aware of this discussion with ONAP 16:26:49 #topic Update on SSH/VM access 16:27:16 #info there's a workaround to get SSH access to VMs 16:27:39 #info apparently the LF hasn't offered SSH access in the past because of an OpenStack bug 16:28:36 #info once that's dealt with we'll be able to have SSH access 16:28:48 #topic Release dashboard 16:29:10 #info today is the Fluorine final checkpoint 16:30:17 #action aadam to find a way to identify the releases affected by MRI blocker bugs 16:30:53 #info Fluorine autorelease is looking good 16:31:52 skitt: the builds are "similar" not "the same" :) 16:32:05 rovarga, true, until we have reproducible builds ;-) 16:32:09 or perhaps equal, but definitely not the same :) 16:32:14 rovarga, the code that went into them should be the same 16:32:21 skitt: which we realistically won't have... ever :) 16:32:56 rovarga, I know a few people who are betting otherwise 16:33:22 rovarga, or rather, that it will become possible to have reproducible builds 16:33:26 skitt: well, until I see some serious effort being put into it ... :) 16:33:53 rovarga, https://reproducible-builds.org/ (not specific to ODL of course) 16:34:13 #info the release integration window starts on September 7 16:34:43 #info for Neon, the affected projects are ODL Parent, YANG Tools and MD-SAL 16:35:00 #info they should all be ready by September 7 and we’ll have a fair idea of the amount of work involved to integrate 16:36:44 #topic Clustering 16:36:51 #info CONTROLLER-1742 might be fixed 16:37:31 #info CONTROLLER-1858 needs more debugging info, which will take some time 16:37:50 #info CONTROLLER-1849 involves using Artery in a three-node cluster, and jamoluhrsen might be the first person on earth to try this 16:39:07 #info jamoluhrsen needs an Artery and Docker and Linux shared memory expert to help him out with this issue 16:39:59 #info OPNFLWPLUG-1022 should be fixed, but there's a CSIT issue to fix too 16:40:29 CaseyLF: If we have time at the end of the meeting I'd like to circle back on the CLM migration request. 16:40:36 #action vishnoianil_ to update OPNFLWPLUG-1022 16:40:58 Ok. We need to give Lisa a few minutes too. 16:43:24 #info NETVIRT-1015 might correspond to expected messages since the test involves killing two nodes 16:44:07 #info the key question is how to monitor a cluster 16:46:29 #action shague to downgrade NETVIRT-1015 16:47:36 #topic Addressing CSIT at the start of a release 16:48:07 #info aadam thinks that now that we've stabilised autorelease, we should address issues in our CSIT 16:49:21 #info rovarga fears removing tests; if we end up disabling tests, we need some workflow to ensure they're re-enabled 16:49:37 #info any lost test coverage will come back as bug reports later 16:50:24 #info shague thinks we need to identify those tests which never pass and are always signed off on 16:50:33 #info there are tests which haven't passed in two years 16:50:46 #info those probably don't have much value at all 16:51:22 #info if we end up removing all the tests for a project, that will have to be addressed by a system test waiver 16:52:13 #info vishnoianil_ suggests opening a JIRA every time a test is removed or disabled 16:53:10 #info LuisGomez suggests running two test jobs, one with only known-good tests to spot regressions, the other with all tests 16:55:38 #topic Marketing 16:56:16 zxiiro: for the change to use a ahitelist for the tests ro run, that would dovetail into this conversation 16:56:16 skitt: aadam: autorelease build for us to use is #189 (https://jenkins.opendaylight.org/releng/view/autorelease/job/autorelease-release-fluorine/189/) 16:56:47 jamoluhrsen, owe yeah :-) 16:56:52 #topic Marketing 16:56:53 #link ODL 2019 planning notes for marketing: https://docs.google.com/document/d/10JQ2CkUX11Mr-64g0ugu6J18uR24LALYpmSzdRpzlf8/edit?usp=drive_web 16:57:07 ... I'm not chaired. :) 16:57:19 But you go it, skitt. Thanks. 16:57:26 #info the marketing teams wants to understand what the project goals are for 2019, where the project is expected to focus 16:57:37 shague: yep 16:57:38 #info both in technical terms and marketing terms (new verticals etc.) 16:57:59 #info without input from the TSC, marketing will use its best guesses 16:58:13 #info the document needs to be completed by the end of August 16:58:47 #action CaseyLF to send an email to the TSC list reminding everyone that the planning document needs to be completed 16:59:30 #info the important aspect is the goals, tactics can be addressed later 16:59:39 #info next Monday’s TWS will be dedicated to this 16:59:54 #topic CLM 17:00:12 #info there is no migration path if we migrate to the shared LFN system, so if we migrate we'll lose our history 17:00:19 #info zxiiro sent an email to the mailing list 17:00:33 #topic Cookies 17:00:36 #endmeeting