#opendaylight-meeting: tsc
Meeting started by colindixon at 17:00:14 UTC
(full logs).
Meeting summary
- agenda bashing and roll call (colindixon, 17:00:20)
- colindixon (colindixon,
17:00:24)
- https://wiki.opendaylight.org/index.php?title=TSC:Main&oldid=32699#Agenda
the snapshot of today’s agenda on the wiki (colindixon,
17:01:00)
- edwarnicke (edwarnicke,
17:01:06)
- https://meetings.opendaylight.org/opendaylight-meeting/2015/tsc/opendaylight-meeting-tsc.2015-06-18-17.00.html
last week’s meeting minutes (colindixon,
17:01:37)
- ACTION: colindixon to
try to find somebody to help with documenting the general procedure
for the platform upgrade from Helium to Lithium (for SR1)
(colindixon,
17:01:56)
- ACTION: ChrisPriceAB
to work with rovarga, jmedved, et. al. to look into leveraging OPNFV
infrastructure for performance measurements and (colindixon,
17:02:10)
- ACTION: phrobb to
investigate what it takes to get to the license for JIRA, and what
the licence would be for (colindixon,
17:02:19)
- jmedved (jmedved,
17:03:22)
- dfarrell07_pto for Red Hat, irc only atm
(dfarrell07_pto,
17:03:36)
- https://wiki.opendaylight.org/view/Events:Main#OpenDaylight_Developer_Forum_-_Santa_Clara.2C_CA_July_30th_and_31st.2C_2015
Put your design forum topics here. We will probably also create one
or more sub-pages as more topics are added (phrobb,
17:03:57)
- mohnish anumala (mohnish,
17:04:17)
- catohornet carolsanders (catohornet,
17:04:24)
- events (colindixon, 17:04:55)
- http://www.opendaylight.org/news/events/
the events list (colindixon,
17:05:12)
- major event next is the ODL summit (colindixon,
17:05:19)
- https://wiki.opendaylight.org/view/Events:Main#OpenDaylight_Developer_Forum_-_Santa_Clara.2C_CA_July_30th_and_31st.2C_2015
Put your design forum topics here. We will probably also create one
or more sub-pages as more topics are added (colindixon,
17:05:34)
- https://wiki.opendaylight.org/view/Group_Based_Policy_(GBP)/Hackathons/IETFPrague15
(alagalah,
17:05:45)
- chris price (ChrisPriceAB,
17:05:54)
- https://www.regonline.com/register/checkin.aspx?EventId=1627424&MethodId=0&EventSessionId=&startnewreg=1
please register for dev/design forum (colindixon,
17:05:57)
- https://wiki.opendaylight.org/view/Group_Based_Policy_(GBP)/Hackathons/IETFPrague15
IETF hacakathon in prague in on July 18th and 19th (GBP and SFC in
particular) (colindixon,
17:06:39)
- there will be other folks from ODL there as
well, it should be a good time (colindixon,
17:06:56)
- infrastructure (colindixon, 17:07:42)
- zxiiro says things look great at the
moment (colindixon,
17:07:50)
- Lithium Release (colindixon, 17:07:56)
- https://docs.google.com/spreadsheets/d/1KPpO9LH539Vlcoa4RvLa6PPCdLifi5JD-ihRhlybqeo/edit#gid=676729675
no active blocking bugs (colindixon,
17:08:24)
- colindixon Notest that we have no blocking
bugs.. (phrobb,
17:08:42)
- colindixon notes one issue in OpenStack and
has been moved from blocking to deferred given the fix needs to be
done in OpenStack (phrobb,
17:09:17)
- OPFLEX artifacts are not in Nexus and it's
being worked. It is not considered to be a blocker (phrobb,
17:09:42)
- documentation, and release notes are still
being worked on and will probably be done so up until our official
announcement on Monday (phrobb,
17:10:49)
- colindixon notes that making adjustments to
release notes between now Monday is fine... at some point colindixon
will cobble the wiki contents up into an asciidoc so get your last
modifications in sooner rather than later (phrobb,
17:12:49)
- https://docs.google.com/spreadsheets/d/1KPpO9LH539Vlcoa4RvLa6PPCdLifi5JD-ihRhlybqeo/edit#gid=478292312
of the 41 projects listed there, d4a is not actually in Lithium and
they know that, otherwise all projects have voted go (colindixon,
17:14:09)
- defense4all has stepped out of the release due
to their own development issues. As a free-standing app, there is
no other impact on the Li release (phrobb,
17:14:39)
- Youcef Laribi (Youcef,
17:14:45)
- All other projects have voted "go" so we are
ready to go re Li (phrobb,
17:15:09)
- https://nexus.opendaylight.org/content/repositories/automatedweeklyreleases-1074/org/opendaylight/integration/distribution-karaf/0.3.0-Lithium/
this is the lithium release distribution which everyone has tested
(colindixon,
17:15:27)
- core tutorials voted n/a because there’s only
one small piece included which is only used by integration
(colindixon,
17:15:47)
- VOTE: Voted on "Shall
the TSC approve the Release of Lithium?" Results are, +1: 7
(phrobb,
17:16:46)
- note that Uri and Youcef vote +1 as well
(colindixon,
17:17:13)
- AGREED: Liithum is
released re the artifacts located here:
https://nexus.opendaylight.org/content/repositories/automatedweeklyreleases-1074/org/opendaylight/integration/distribution-karaf/0.3.0-Lithium/
(phrobb,
17:17:27)
- colindixon congratules for first on time
Opendaylight major release with 40+ projects (gzhao,
17:17:56)
- colindixon says everyone should go and
celebrate!!!! (gzhao,
17:18:21)
- zxiiro notes that the release engineering team
will be tagging and bumping versions as part of the release
(colindixon,
17:19:11)
- zxiiro asks project don't merge patches on
stable/lithium (gzhao,
17:19:22)
- https://lists.opendaylight.org/pipermail/release/2015-June/002990.html
the e-mail explaining the details (colindixon,
17:20:54)
- edwarnicke asks how long version bumping and
patches will take, Thanh says today, colin says it will take at
least 8 hours, colindixon says it’s going to take maybe closer to a
few days (colindixon,
17:22:42)
- colindixon says early next week is a resonable
estimate for when it could be done (colindixon,
17:22:58)
- Martin Sunal as a comitter on SXP (colindixon, 17:24:05)
- TSC tabling SXP committer discussion until SXP
members present (phrobb,
17:24:32)
- creation review fro NetIDE (colindixon, 17:24:40)
- https://wiki.opendaylight.org/view/Project_Proposals:NetIDE
the project proposal (colindixon,
17:24:53)
- https://wiki.opendaylight.org/images/9/9c/NetIDE_ODL_proposal.pdf
slides at the bottom of the proposal (colindixon,
17:25:10)
- https://lists.opendaylight.org/pipermail/project-proposals/2015-June/000318.html
proposed on May 19th, so much more than two weeks ago (colindixon,
17:25:40)
- NetIDE consortium is broader than just Intel,
many other academics and companies (colindixon,
17:26:10)
- looking to create an IDE for SDN
applications (colindixon,
17:26:38)
- https://github.com/fp7-netide
it’s already an open source project on github (colindixon,
17:27:22)
- it’s already under the EPL (colindixon,
17:27:35)
- colindixon notes that they have 4 repos on
github, but would likely only get one repo in ODL, will that be an
issue? (colindixon,
17:28:50)
- alec says he doesn’t think so, because not all
of the stuff on github will be migrated to ODL (colindixon,
17:29:17)
- jmedved asks if this is targeting apps that use
the REST API or the ones that are plugings, Alec says ideally
both (colindixon,
17:29:58)
- two main parts: network engine and IDE
(colindixon,
17:30:23)
- NetIDE network engine (colindixon, 17:30:30)
- starts on slide 7 (colindixon,
17:30:38)
- the architecture has a notion of a client SDN
controller and server SDN controller using backends and shims for
them to be layered (colindixon,
17:31:12)
- Alec says he can run a demo of this on mininet
using ODL as the server SDN controller with Floodlight, Pyretic, and
Rya as client SDN controllers (colindixon,
17:32:46)
- demo app is branch office connectivity
(colindixon,
17:32:53)
- the language being used between the shims and
the backends (glue between client and server SDN controllers) is
currently the pyretic language (colindixon,
17:33:19)
- the pyretic language for now is OpenFlow 1.0
only, which has limitations (colindixon,
17:33:36)
- jmedved asks if this is OpenFlow specific, Alec
says yes for now (colindixon,
17:33:55)
- jmedved asks what they mean by supporting
NETCONF, Alec says that they would think about forking Pyretic to
add in 1.3 and some new features, e.g., NETCONF messages
(colindixon,
17:34:42)
- jmedved says that things aren’t very useful
with OpenFlow alone, Alec says maybe, but there are lots of OpenFlow
controllers out there and this enables those to run on topo of
ODL (colindixon,
17:35:38)
- edwarnicke asks about the value of using
pyretic messages vs. translating directly to models for ODL
(colindixon,
17:36:37)
- Alec answers because there is another shim for
Ryu and pox (it’s not just ODL) (colindixon,
17:36:54)
- edwarnicke notes that that he believes that
using pyretic as the carrier message will result in pain since it is
“semantically poor” when compared to OpenDaylight’s models
(colindixon,
17:39:52)
- colindixon makes sure to note that this
technical discussion is good and important, but the technical
discussion will not much (if any) bearing on whether the TSC
approves the project and should be taken as engineer-to-engineer
feedback (colindixon,
17:40:50)
- there is conversation about openflow-only vs.
more general (colindixon,
17:41:13)
- to be clear, Alec says that their plans with
the pyretic language would be to try to incorporate things back to
the main project (colindixon,
17:44:04)
- edwarnicke asks how the pyretic guys are
feeling about these modifications and extensions, Alec says
generally thats’s a good relationship (colindixon,
17:44:46)
- zxiiro and dfarrell07_pto brought up the
question about forking pyretic, not me just to give credit where
credit is due (colindixon,
17:45:27)
- jmedved points out that different controllers
have different clustering/HA models, Alec says they haven’t looked
at that yet (colindixon,
17:45:52)
- the NetIDE IDE (colindixon, 17:46:35)
- starts on slide 10 (colindixon,
17:47:00)
- this is a suite of plugins to Eclipse (colindixon, 17:47:18)
- shows a screenshot of an interface GUI with
switches, hosts, etc. (colindixon,
17:47:41)
- this then auto-generates the relevant code
using Django to generate python for Ryu in the screentshot
(colindixon,
17:48:11)
- this could be done with YANG and Java in a
similar way for OpenDaylight (colindixon,
17:48:22)
- another tool they aren’t showing, allows you to
take an ODL topology and then deploy it in mininet to check your
real topology in simulation (colindixon,
17:49:13)
- Just to make sure it gets in the notes
(regarding Pyretic fork), dfarrell07_pto notes that long-lived forks
are typically harder to integrate than working upstream initially,
and suggests that Pyretic committer rights seem appropriate for
people doing such substantial refactoring (dfarrell07_pto,
17:50:18)
- the scope seems to be NetIDE consortium’s shim
and backend for allowing the client/server SDN controller
architecture and Eclipse-based IDE tools to help developers
(colindixon,
17:52:29)
- jmedved says we could expand things if we could
consider topology stacks beyond the typical L2/L3, openflow
scope (colindixon,
17:53:40)
- Alec says he’s open to that (colindixon,
17:53:45)
- VOTE: Voted on "Shall
the TSC approve NETIDE?" Results are, +1: 8 (dfarrell07_pto,
17:55:36)
- note that Uri_ also voted +1 (colindixon,
17:55:45)
- AGREED: The TSC vote
above was to move the NetIDE project to incubation (dfarrell07_pto,
17:57:04)
- Lith cookies! (dfarrell07_pto, 17:58:55)
- ACTION: me to send a
mail (ChrisPriceAB,
17:59:20)
Meeting ended at 17:59:52 UTC
(full logs).
Action items
- colindixon to try to find somebody to help with documenting the general procedure for the platform upgrade from Helium to Lithium (for SR1)
- ChrisPriceAB to work with rovarga, jmedved, et. al. to look into leveraging OPNFV infrastructure for performance measurements and
- phrobb to investigate what it takes to get to the license for JIRA, and what the licence would be for
- me to send a mail
Action items, by person
- ChrisPriceAB
- ChrisPriceAB to work with rovarga, jmedved, et. al. to look into leveraging OPNFV infrastructure for performance measurements and
- colindixon
- colindixon to try to find somebody to help with documenting the general procedure for the platform upgrade from Helium to Lithium (for SR1)
- jmedved
- ChrisPriceAB to work with rovarga, jmedved, et. al. to look into leveraging OPNFV infrastructure for performance measurements and
- phrobb
- phrobb to investigate what it takes to get to the license for JIRA, and what the licence would be for
- rovarga
- ChrisPriceAB to work with rovarga, jmedved, et. al. to look into leveraging OPNFV infrastructure for performance measurements and
People present (lines said)
- colindixon (91)
- dfarrell07_pto (25)
- alagalah (22)
- odl_meetbot (17)
- ChrisPriceAB (16)
- phrobb (13)
- tykeal (10)
- gzhao (8)
- ebrjohn (7)
- edwarnicke (7)
- zxiiro (5)
- catohornet (4)
- mohnish (4)
- jmedved (3)
- Youcef (3)
- abhijitkumbhare (3)
- drizzt__ (3)
- rovarga (1)
- odl-casey (1)
- uri_ (1)
Generated by MeetBot 0.1.4.