#opendaylight-meeting: tsc
Meeting started by colindixon at 17:00:14 UTC
(full logs).
Meeting summary
- roll call and agenda bashing (colindixon, 17:00:20)
- colindixon (colindixon,
17:00:31)
- https://wiki.opendaylight.org/index.php?title=TSC:Main&oldid=48426#Agenda
the agenda in it's usual place (colindixon,
17:00:43)
- https://meetings.opendaylight.org/opendaylight-meeting/2016/tsc/opendaylight-meeting-tsc.2016-08-11-17.00.html
last week's meeting minutes (colindixon,
17:00:53)
- gkaempfer (gkaempfer,
17:01:04)
- https://wiki.opendaylight.org/view/Wiki_gardening#Search_Engine_Indexing_of_the_wiki
information on search engine indexing of the wiki (colindixon,
17:01:32)
- https://www.opendaylight.org/global-events
LinuxCon North America and Europe are in the events list (colindixon,
17:01:59)
- https://wiki.opendaylight.org/view/OpenDaylight_Toolkit:Main
https://wiki.opendaylight.org/view/Archive_Proposals/OpenDaylight_Toolkit#Archive_Vote_and_Date
colindixon added the relevant bits on the wiki to archive toolkit
(colindixon,
17:02:04)
- https://lists.opendaylight.org/pipermail/tsc/2016-August/005836.html
thread on whether we should only allow blocking bug fixes during RCs
(colindixon,
17:02:10)
- abhijitkumbhare (abhijitkumbhare,
17:02:13)
- Chris Price (ChrisPriceAB,
17:02:15)
- https://lists.opendaylight.org/pipermail/release/2016-August/007953.html
it seems like the HTTP response codes from RESTCONF is likely to be
reasonable (colindixon,
17:02:21)
- adetalhouet (adetalhouet,
17:02:25)
- mohnish anumala (mohnish__,
17:02:43)
- making sure that we can set the the nexus file
size back to 512 MB again is being tracked on integration
calls (colindixon,
17:02:43)
- snackewm (for uri) (snackewm,
17:03:16)
- https://lists.opendaylight.org/pipermail/tsc/2016-August/005842.html
phrobb's mail to projects to note that who the PTL is at the end of
boron has implications for the TSC election, so figure it out if you
have to (colindixon,
17:03:27)
- ACTION: colindixon to
e-mail out results of the APAC-time-zone poll (colindixon,
17:03:28)
- so far we only have 5 responses, so please vote
if you want to, it closes in 7 hours (colindixon,
17:03:57)
- ACTION: colindixon to
try to find somebody to help with documenting the general procedure
for the platform upgrade to Boron/Carbon? (colindixon,
17:04:03)
- ACTION: colindixon to
try to drive some/all of the ongoing TSC discussions to
completion (colindixon,
17:04:09)
- Anil Vishnoi (vishnoianil,
17:04:21)
- ACTION: anipbu to
make sure we release note the change in HTTP RESTCONF RESPONSE
codes (colindixon,
17:04:29)
- ACTION: colindixon
and edwarnicke to try to work through the corner cases and process
of the TSC elections
https://wiki.opendaylight.org/view/TSC:Election_Proposal
(colindixon,
17:05:30)
- ACTION: everyone to
review the other bits of the TSC elections, most notably the the
representing multiple groups or projects and picking represented
groups
https://wiki.opendaylight.org/view/TSC:Election_Proposal
(colindixon,
17:05:38)
- jamoluhrsen (jamoluhrsen,
17:06:47)
- TSC mailing list votes and discussions (colindixon, 17:06:56)
- https://lists.opendaylight.org/pipermail/tsc/2016-August/005809.html
Policy on who can attend DDFs (colindixon,
17:07:06)
- https://lists.opendaylight.org/pipermail/tsc/2016-August/005836.html
only allow blocking bug fixes after RCs start in Carbon (colindixon,
17:07:16)
- older ones can be found in the agenda link
above (colindixon,
17:07:28)
- events (colindixon, 17:08:13)
- https://www.opendaylight.org/global-events
(colindixon,
17:08:20)
- LinuxCon NA and OpenStack Days Easy coming up
next week (colindixon,
17:08:44)
- we've finished the CFP for OpenDaylight Forum
India, committee being formed (colindixon,
17:09:05)
- https://wiki.opendaylight.org/view/Events:Carbon_Dev_Forum
we really need topics for the Carbon DDF at the summit (colindixon,
17:09:33)
- Boron (colindixon, 17:10:34)
- We had build failures in FAAS and NETIDE but
those issues have been fixed and resolved. (anipbu,
17:10:49)
- We had a TWS to discuss the MDSAL Binding Spec
V2 (anipbu,
17:11:02)
- https://lists.opendaylight.org/pipermail/release/2016-August/007909.html
<--- Recoding for MDSAL Binding Spec V2 (anipbu,
17:11:10)
- https://lists.opendaylight.org/pipermail/release/2016-August/007941.html
<--- PowerPoint Presentation for MDSAL Binding Spec V2 (anipbu,
17:11:16)
- https://lists.opendaylight.org/pipermail/release/2016-August/007916.html
<--- Email Thread Discussion for MDSAL Binding Spec V2 (anipbu,
17:11:23)
- We continue the migration from AsciiDocs to
reStructuredText. Request folks to review relevant patches and +1.
Projects may also elect to opt out by informing the documentation
team at documentation@lists.opendaylight.org (anipbu,
17:12:01)
- https://git.opendaylight.org/gerrit/#/q/topic:adoc2rst
<--- Documentation Migration Patches (anipbu,
17:12:06)
- 6 projects affected by Restconf Response
Migration. Projects should review the impact of the changes to
Restconf Response. (anipbu,
17:12:34)
- https://docs.google.com/spreadsheets/d/1zImtd764e-hOgJAxoJKl85fxHCPu2agLfqsBtf13zQY/edit#gid=1846323581
<--- Tracking Restconf Response Impact (anipbu,
17:12:48)
- Projects please test and report blocking issues
with Boron RC0. (anipbu,
17:15:01)
- https://wiki.opendaylight.org/view/Simultaneous_Release:Boron_Release_Plan#RC0
<--- RC0 Download Information (anipbu,
17:15:07)
- https://docs.google.com/spreadsheets/d/1zImtd764e-hOgJAxoJKl85fxHCPu2agLfqsBtf13zQY/edit#gid=199721620
<--- RC0 Tracking Spreadsheet (anipbu,
17:15:11)
- The build for RC1 will start at 23:59:59 UTC on
8/18/2016 (anipbu,
17:15:22)
- Projects PTL should sign up for timeslots for
the Boron Release Review and provide links to their release notes
and release review template (anipbu,
17:15:42)
- Sign Up Sheet: https://docs.google.com/spreadsheets/d/1zImtd764e-hOgJAxoJKl85fxHCPu2agLfqsBtf13zQY/edit#gid=1708916020
(anipbu,
17:15:55)
- Release Note: https://wiki.opendaylight.org/view/Sample_Release_Notes
(anipbu,
17:16:04)
- Release Review Template: https://wiki.opendaylight.org/view/Sample_Release_Review
(anipbu,
17:16:10)
- 9 projects have made changes to their yang
files since RC0 Cutoff: groupbasedpolicy, netvirt, genius,
openflowplugin, bgpcep, yangtools, aaa, of-config, sfc (anipbu,
17:16:28)
- https://git.opendaylight.org/gerrit/#/dashboard/?title=RC+Status&Yang+File+Changes+Since+RC0=branch:stable/boron+status:merged+file:%22%255E.%252Byang%22+after:2016-08-11
(anipbu,
17:16:33)
- ChrisPriceAB says things like the HTTP response
codes should probably focus at least as much externally as
internally, in particular, most of them will see no idea of whether
this change is coming (colindixon,
17:18:35)
- ChrisPriceAB says he doesn't think we should
stop things over this, but says "we may make this change and have
the world hate us for it, but that would be fine as long as we
learn" (colindixon,
17:20:08)
- edwarnicke says that his take is that if we're
going to fix this, we shoudl fix it now, not kick the can down the
road (colindixon,
17:20:48)
- one appraoch would be to deprecate it, set off
alarms, etc. for one release (per ChrisPriceAB) (colindixon,
17:21:43)
- ACTION: colindixon to
ask how hard it woudl be to add a knob that was off-by-default that
when enabled reverted HTTP response codes to the Beryllium values
and if we can do in Boron (colindixon,
17:24:04)
- mohnish__ asks if we should do new (right)
behavior by default or old (wrong) behavior by default, the latter
would allow people to move without interruption (colindixon,
17:25:31)
- colindixon says that edwarnicke was suggesting
the former to make sure people were actively aware that things were
changing and not just replay things 6 months from now (colindixon,
17:26:21)
- vrpolak will involve changes in a YANG model
that changes RESTCONF behavior, which will involve an API change,
and to actually flip the switch, you'll need to edit an XML
file (colindixon,
17:28:06)
- system integration and test (colindixon, 17:28:19)
- all covered in infra (colindixon,
17:28:32)
- infrastructure (colindixon, 17:28:35)
- zxiiro says that there's an issue we're aware
of with rackspace public cloud where some CSIT nodes don't come
online and we can't ssh to them, seems to happen about 1 time in 6
or 7 (colindixon,
17:29:10)
- jamoluhrsen says that this will affect
distrirbution-test jobs, which is how we evaluate releases
(colindixon,
17:29:34)
- colindixon asks if CSIT will make it through
that, jamoluhrsen says that we have many jobs which bring up 4+
nodes, that menas it fails with p=0.5 or so, which makes it a huge
amount of busy (colindixon,
17:30:40)
- rackspace is working on it, we don't know of an
ETA yet (colindixon,
17:31:26)
- colindixon notes that we're still seeing some
Nexus timeouts, but much, much less (colindixon,
17:31:59)
- colindixon says he's seen 1 example and run
something like 200 verifiy jobs yesterday (colindixon,
17:32:24)
- zxiiro says he's see 3 reports so far
(colindixon,
17:32:33)
- An Ho as committer on releng/autorlease (colindixon, 17:32:50)
- https://lists.opendaylight.org/pipermail/tsc/2016-August/005833.html
(colindixon,
17:32:57)
- https://lists.opendaylight.org/pipermail/dev/2016-August/002620.html
votes (colindixon,
17:33:05)
- https://git.opendaylight.org/gerrit/#/q/project:releng/autorelease+owner:%22A+H+%253Can.ho%2540huawei.com%253E%22
(colindixon,
17:33:27)
- VOTE: Voted on "shall
the TSC approve An Ho as a committer on releng/autorelease?" Results
are, +1: 9 (colindixon,
17:34:26)
- AGREED: An Ho is a
committer on releng/autorelease (colindixon,
17:34:44)
- TSC elections (colindixon, 17:35:56)
- colindixon asks about people's readiness to
discuss it (colindixon,
17:36:16)
- consensus seems like people need to catch up
with the e-mail thread, but we could cover it for a bit (colindixon,
17:36:45)
- https://wiki.opendaylight.org/view/TSC:Election_Proposal
(colindixon,
17:36:55)
- the two major remaining topics are (1) dealing
with corner cases, (2) how we will mechanically use CIVS or
something else to run the elctions, (3) picking represented
groups (colindixon,
17:37:47)
- https://wiki.opendaylight.org/view/TSC:Election_Proposal#All_Types.2C_Mature_Lifecycle_States
this is the represented group option which seemed to get consensus,
but there others above/below it (colindixon,
17:38:07)
- abhijitkumbhare asks why we have mature seats
here if all the projects have to be mature, colindixon answers it's
just to give extra representation to mature projects and an
incentive to move to mature (colindixon,
17:41:54)
- colindixon notes that having 1/3 (5 seat) cap on mature projects,
will probalby be a problem as the number of mature projects grows
from where we are now with 8 mature projects: https://wiki.opendaylight.org/view/Category:Mature_Projects
(colindixon,
17:43:10)
- abhijitkumbhare aks if we really need caps for
project categories at all, edwarnicke says that caps are trick and
can be bad (colindixon,
17:43:45)
- to clarify... edwarnicke favors either no caps
on any RG (including company caps) or uniform caps (edwarnicke,
17:44:20)
- edwarnicke 's point was that even company caps
are *already* producing distortions where strong community members
are being bumped (edwarnicke,
17:44:57)
- ChrisPriceAB says he worries about how
complicated this is and that it might make people have a
disconnection between their vote and their results (colindixon,
17:45:22)
- in particular ChrisPriceAB notes that this
would seem to potentially punish projects for becoming core and/or
mature (colindixon,
17:46:50)
- some rehashing of the RG caps discussion
(colindixon,
17:47:02)
- colindixon notes that we're not sure how and if
CIVS actually supports scaled popularity and the like (colindixon,
17:48:41)
- also, we need to decide if somebody can be
elected to the TSC twice in two different elections? also if
somebody is the PTL of two projects in a group, do they get 2
votes? (colindixon,
17:49:30)
- edwarnicke says his take is that PTLs are
voting on behalf of the project, mohnish__ says that in that case,
the PTL might have to poll the project, colindixon notes that in
general, we let PTLs vote their conscience without having to consult
the project on every issue (colindixon,
17:52:57)
- vishnoianil says it's not clear if PTLs are
voting for themselves or voting for their project, if it's the
latter, the PTL should probably get a number votes equal to the
number of projects they're leading (colindixon,
17:53:33)
- vishnoianil asks how we solve this for
committers-at-large, colindixon says that's an election over
committers and not projects, so it's one person one vote
(colindixon,
17:55:56)
- vishnoianil asks if you could vote two
different ways with your two votes, phrobb and colindixon say that
you might really want to allow that, phrobb also says that if
somebody was PTL of projects in 2 different RGs, he would get one
vote in each, so maybe he should get 2 votes in one RG (colindixon,
17:58:40)
- another corner case would be if the votes are
open (colindixon,
18:01:01)
- cookies (colindixon, 18:01:20)
Meeting ended at 18:01:28 UTC
(full logs).
Action items
- colindixon to e-mail out results of the APAC-time-zone poll
- colindixon to try to find somebody to help with documenting the general procedure for the platform upgrade to Boron/Carbon?
- colindixon to try to drive some/all of the ongoing TSC discussions to completion
- anipbu to make sure we release note the change in HTTP RESTCONF RESPONSE codes
- colindixon and edwarnicke to try to work through the corner cases and process of the TSC elections https://wiki.opendaylight.org/view/TSC:Election_Proposal
- everyone to review the other bits of the TSC elections, most notably the the representing multiple groups or projects and picking represented groups https://wiki.opendaylight.org/view/TSC:Election_Proposal
- colindixon to ask how hard it woudl be to add a knob that was off-by-default that when enabled reverted HTTP response codes to the Beryllium values and if we can do in Boron
Action items, by person
- anipbu
- anipbu to make sure we release note the change in HTTP RESTCONF RESPONSE codes
- colindixon
- colindixon to e-mail out results of the APAC-time-zone poll
- colindixon to try to find somebody to help with documenting the general procedure for the platform upgrade to Boron/Carbon?
- colindixon to try to drive some/all of the ongoing TSC discussions to completion
- colindixon and edwarnicke to try to work through the corner cases and process of the TSC elections https://wiki.opendaylight.org/view/TSC:Election_Proposal
- colindixon to ask how hard it woudl be to add a knob that was off-by-default that when enabled reverted HTTP response codes to the Beryllium values and if we can do in Boron
- edwarnicke
- colindixon and edwarnicke to try to work through the corner cases and process of the TSC elections https://wiki.opendaylight.org/view/TSC:Election_Proposal
People present (lines said)
- colindixon (86)
- anipbu (21)
- edwarnicke (11)
- odl_meetbot (11)
- abhijitkumbhare (10)
- adetalhouet (5)
- ChrisPriceAB (5)
- jamoluhrsen (4)
- vishnoianil (3)
- mohnish__ (2)
- gkaempfer (2)
- snackewm (2)
- zxiiro (1)
- CaseyODL (1)
- phrobb (0)
Generated by MeetBot 0.1.4.