#opendaylight-meeting Meeting

Meeting started by cdub at 17:03:59 UTC (full logs).

Meeting summary

    1. Chris Price joined the meeting (chrisprice___, 17:04:15)

  2. #info in (cdub, 17:04:25)
    1. Ed Warnicke (edwarnicke___, 17:04:29)
    2. dmm (dmm, 17:04:32)
    3. Chris Wright here (cdub, 17:04:32)
    4. Chris rice (chrisprice___, 17:04:40)
    5. Colin Dixon representing IBM on behalf of Vijoy (colindixon, 17:05:34)
    6. tbachman asks about recording, dmm reminds that we are using meetbot for capturing minutes instead of call recording (cdub, 17:06:33)
    7. https://wiki.opendaylight.org/view/TSC:Main#Meeting_Agenda (dmm, 17:06:45)
    8. dmm congrats group on INTEROP Best in SDN and Best in Show (cdub, 17:07:23)

  3. agenda bashing (cdub, 17:09:42)
    1. discussing Summit date (Hyatt in Sept) and actual Helium release process (cdub, 17:10:06)
    2. this is on the agenda for today's call (cdub, 17:10:12)
    3. colindixon says we agreed to book Hyatt but didn't agree to any specific release plan (cdub, 17:11:48)
    4. edwarnicke___ agrees with colindixon's understanding (phrobb, 17:12:27)
    5. for clarity...Yes...we agreed to venue, not specific release plan (cdub, 17:12:58)

  4. board meeting (cdub, 17:13:18)
    1. board agrees to 60day delay for TSC election to strighten out issues (cdub, 17:14:39)
    2. this means that we will need to agree on what we want to take to them earlier than that to give them time to consider them (colindixon, 17:15:06)
    3. SF Hyatt booked Sep30-Oct1, hopefully w/ Helium just released, perhaps final sprint for Helium (cdub, 17:15:39)
    4. (booked for design summit) (cdub, 17:16:01)

  5. creation reviews (cdub, 17:16:05)
    1. https://wiki.opendaylight.org/view/Project_Proposals:OpenDaylight_Toolkit (Madhu, 17:16:30)
    2. https://wiki.opendaylight.org/view/Project_Proposals:OpenDaylight_Toolkit (cdub, 17:16:49)
    3. OpenDaylight Toolkit creation review (cdub, 17:17:14)
    4. actively under development, current repo in github (cdub, 17:18:11)
    5. https://github.com/opendaylight-toolkit/opendaylight-toolkit (cdub, 17:18:42)
    6. had really good demo 2wks ago (cdub, 17:19:42)
    7. general comments by TSC are this project is awesome (phrobb, 17:20:00)
    8. regXboi noted that archetypes could have long term architectural implications (i.e. this is how we do dev for internal projects) (cdub, 17:20:27)
    9. edwarnicke___ notes archetypes are awesome, but have real limitations, and unlikely to be way to build internal services because of those limitations (cdub, 17:20:56)
    10. also noted that archetypes have limits, but they are tremendously helpful for many types of apps (phrobb, 17:21:09)
    11. dmm notes these concerns are largely theoretical and benefits outweigh risks (cdub, 17:21:31)
    12. regXboi notes that he did not want to re-raise these issues and so the topic is largely tabled (colindixon, 17:21:51)
    13. colindixon and edwarnicke___ agree to share generalized blame. (edwarnicke___, 17:22:32)
    14. kentwatsen expresses concern...lots of projects already, does this make it easier to make more projects and therefore dillute current focus? (cdub, 17:23:22)
    15. clarify...this toolkit is to help newcomers to get started and build apps, not to create more ODL projects (cdub, 17:24:07)
    16. OpenDaylight Toolkit vote is unanimous (cdub, 17:25:47)
    17. AGREED: OpenDaylight Toolkit is incubated project (cdub, 17:25:59)
    18. and there is much rejoicing (cdub, 17:26:42)
    19. PCMM project review (cdub, 17:27:08)
    20. https://wiki.opendaylight.org/view/Project_Proposals:PacketCablePCMM (colindixon, 17:27:39)
    21. Thomas Kee presents on PCMM---slides will hopefully be posted later if they aren't already (colindixon, 17:31:10)
    22. https://github.com/xsited/packetcable github PCMM work underway (cdub, 17:34:05)
    23. PCMM, why OpenDaylight: SAL! (cdub, 17:35:30)
    24. https://wiki.opendaylight.org/view/Project_Proposals:PacketCablePCMM#Work_Flow_Example reviewing work flow and how ODL is updated w/ PCMM project (cdub, 17:37:31)
    25. PCMM project goals: finish sb PCMM driver, NB: CMTS provisioning, traffic profile, flow programmer de-augmentation/augmentation (cdub, 17:43:08)
    26. stay to release plan (aiming at finishing above in June) (cdub, 17:43:22)
    27. give this work to community (cdub, 17:43:32)
    28. AD-SAL vs MD-SAL? (cdub, 17:44:02)
    29. flow ids and gate ids mapping (cdub, 17:44:19)
    30. include IPv6 (cdub, 17:44:37)
    31. how to remove functionality from models rather than add (colindixon, 17:44:49)
    32. deal w/ lack of l2 (cdub, 17:45:02)
    33. open floor for questions... (cdub, 17:45:36)
    34. colindixon great showcase for SAL, and let's learn from that...great end-to-end workflow... (cdub, 17:46:25)
    35. dlenrow would like to see how we can tie policy group in here (cdub, 17:47:12)
    36. kwatsen (kwatsen, 17:48:09)
    37. LuisGomez curious if there is same QoS type enforcement capability w/out openflow (cdub, 17:48:54)
    38. some available in the past, but kind of died off, this is opportunity to revitalize (cdub, 17:49:31)
    39. edwarnicke___ asks: openflow is packet-in, cops for flow programming? (cdub, 17:50:21)
    40. yes, although some initial PoC was pure openflow (cdub, 17:50:35)
    41. netconf definition for full appliance config coming, and in short term use snmp (cdub, 17:51:16)
    42. PCMM project vote (cdub, 17:52:13)
    43. AGREED: PCMM is incubation (colindixon, 17:52:28)
    44. AGREED: PacketCablePCMM accepted as incuabtion project (cdub, 17:52:28)

  6. System Integration and Testing update (cdub, 17:53:57)
    1. sent mail w/ details for performance meetings (not perfect time, but will start w/ that) (cdub, 17:54:26)

  7. Stable release schedule/mechanics (phrobb, 17:56:00)
    1. https://wiki.opendaylight.org/view/CrossProject:Stable_Release (cdub, 17:56:19)
    2. cdub calls attention to sections 2 through 4 for TSC discussion (phrobb, 17:57:25)
    3. first, is having a branch for each project that tracks it's stable releases (colindixon, 17:58:02)
    4. cdub proposes a consistent name for Stable Branches across projects (phrobb, 17:58:44)
    5. rough consensus gained for naming convention noted on wiki (phrobb, 18:00:55)
    6. AGREED: the stable branch naming convention will stand as described in the document (namely stable/<release-name-in-lowercase>) (colindixon, 18:01:13)
    7. agreed consensus was branch name of stable/hydrogen (lower case) (edwarnicke___, 18:01:25)
    8. https://git.opendaylight.org/gerrit/#/admin/projects/controller,branches controller has correcte to stable/hydrogen from stable/Hydrogen (edwarnicke___, 18:01:47)
    9. do we simply cut what we have now? (cdub, 18:03:46)
    10. regXboi says emphatically "No!" (cdub, 18:04:01)
    11. artivact version numbers with a concrete suggestion of <major>.<minor>.<bugfix>-N where the dotted triple is the last release, i.e., hydrogen, and N represents the current stable release under that (colindixon, 18:04:03)
    12. regXboi notes that without close examination, we may add feature and/or api changes to the stable branch (phrobb, 18:05:02)
    13. colindixon asks about using a dotted triple with a dash-number on the other end (colindixon, 18:06:25)
    14. cdub says that dotted triples with a -N (where N is a number) on the end seems to actually have advantages, sorts properly, etc. (colindixon, 18:07:03)
    15. now discussing the Stable Patch Criteria section of the wiki page (phrobb, 18:08:21)
    16. we agree there's rough consensus on the version numbers barring somebody telling us it breaks maven in a way we haven'c considered (colindixon, 18:09:23)
    17. cdub wants to have good criteria and guidelines to decide what patches get moved into stable releases (colindixon, 18:10:58)
    18. edwarnicke___ is fine with that, but cautions to keep it as guidelines which we can use to shame people rather than strict laws passed and enforced by the TSC (colindixon, 18:11:31)
    19. question: are there missing criteria or any radical objections to this criteria presented? (phrobb, 18:11:47)
    20. kwatsen asks if we should change the last bullet to say you can't add new APIs or add new APIs (colindixon, 18:12:18)
    21. or change existing APIs I mean (colindixon, 18:12:36)
    22. edwarnicke___ wants to note that "induce productive discussion" would be a better way to phrase "shame people" when it comes to using these guidelines (colindixon, 18:16:05)
    23. LuisGomez and regXboi note that having a link on mechanics of cherrypicking patches would be helpful (phrobb, 18:19:56)
    24. ACTION: regXboi to document cherrypicking (phrobb, 18:20:15)
    25. , general consensus gained on stable patch criteria (phrobb, 18:20:53)
    26. now discussing Stable Release Criteria (phrobb, 18:21:10)
    27. cdub asks will we get to releasing individual projects? We are not ready to do that currently as we have things bundled (phrobb, 18:27:03)
    28. regXboi requests that all projects stay synced on the stable branch. If not, those taking the code to add to their own solutions gets very difficult (colindixon, 18:28:16)
    29. cdub coins a new term the Integrator's dillema (phrobb, 18:28:31)
    30. https://docs.google.com/spreadsheet/ccc?key=0AhywWQdJrMqedGZzREV0ZUtCSHkyOGl2a1dmWTJ4Y0E&usp=sharing#gid=0 task (cdub, 18:32:04)
    31. above link is for tracking what needs to be done (phrobb, 18:32:37)
    32. regXboi notes he will be documenting the gerrit cherrypick process (phrobb, 18:33:32)
    33. discussion ensues on how to cherrypick and document what bugs/patches have been put on stable branch (phrobb, 18:38:07)
    34. discussion about how to make developers' lives easier when a bug fix may not cherry pick as cleanly as would be liked (colindixon, 18:41:44)
    35. need to add the workflow to the wiki (phrobb, 18:42:32)
    36. ACTION: cdub to work with leena on documenting workflow on the wiki (phrobb, 18:42:58)
    37. ACTION: lr_ will update wiki w/ workflow (cdub, 18:43:08)
    38. we need a primary contact for stable releases for each project, projects please put this up on the sheet (and also the wiki?) (colindixon, 18:47:28)

  8. Helium Release Schedule (cdub, 18:47:34)
  9. Helium Release Schedule (phrobb, 18:47:35)
    1. https://wiki.opendaylight.org/view/Simultaneous_Release:Helium_Release_Plan (cdub, 18:48:09)
    2. compress above to (cdub, 18:50:49)
    3. M0 - 4/07 (cdub, 18:51:01)
    4. M1 - 5/01 (cdub, 18:51:08)
    5. rest is the same as wiki (cdub, 18:51:20)
    6. the result of this would be that projects must submit a proposal in by 4/17 in order to be able to join the Helium release (colindixon, 18:53:31)
    7. note that those changes (for M0 and M1) are now made on the wiki page (colindixon, 18:54:19)
    8. edwarnicke___ points out that there are people who are outsiders who may not be well-versed in the release process and so are going to be caught unawares by the fact that they will need to be moving quickly soon (colindixon, 18:55:58)
    9. cdub is less concerned about this (colindixon, 18:56:17)
    10. colindixon asks if this can be solve pretty easily by offering leniency for new projects that don't have people depending on them (colindixon, 19:01:00)
    11. cdub points out that this is already in the draft doc (colindixon, 19:01:12)
    12. chrisprice___ points out that we do need *a* time and that there will always be projects on the cusp of those dates, so moving things around won't likely help too much (colindixon, 19:04:40)
    13. discussion continue as how best to come to consensus on release planning and dates (phrobb, 19:07:24)
    14. kwatsen notes that this is a second release and new projects should be paying attention. Also, on hitting a Sept. date, we may need set the feature set to match the timeframe needed (phrobb, 19:09:35)
    15. Chris Wright is OK w/ plan as is (cdub, 19:13:26)
    16. colindixon is generally OK with the plan as is (noting exceptions for new projects and the fact that the major deadlines are still well in the future) (colindixon, 19:13:46)
    17. 0 (abstain), but generally link the plan currently listed on wiki (kwatsen, 19:14:08)
    18. Ok with the plan as it stands (chrisprice___, 19:14:15)
    19. Ed Warnicke is concerned that a) The current plan would have a date in the past by the time the TSC could actually vote on it. b) We actually have had no discussion of the actual content of the plan, or improvements over Hydrogen we would like to see in the plan... we seem to only be voting on end dates. c) There is insufficient space in the plan (edwarnicke___, 19:15:19)
    20. tbachman has no point (tbachman, 19:24:16)
    21. ACTION: edwarnicke___ to propose an alternate time schedule for Helium Release (phrobb, 19:26:02)
    22. colindixon proposes to set M0 date as 4/10 so that the TSC must have a plan at the end of next week's TSC meeting (phrobb, 19:29:55)
    23. AGREED: TSC will decide on the Release schedule for Helium at next week's meeting (phrobb, 19:32:12)
    24. ACTION: cdub to send an email to explain this decision to TSC and Discuss list so everyone in the community knows (phrobb, 19:32:49)


Meeting ended at 19:33:27 UTC (full logs).

Action items

  1. regXboi to document cherrypicking
  2. cdub to work with leena on documenting workflow on the wiki
  3. lr_ will update wiki w/ workflow
  4. edwarnicke___ to propose an alternate time schedule for Helium Release
  5. cdub to send an email to explain this decision to TSC and Discuss list so everyone in the community knows


Action items, by person

  1. cdub
    1. cdub to work with leena on documenting workflow on the wiki
    2. cdub to send an email to explain this decision to TSC and Discuss list so everyone in the community knows
  2. edwarnicke___
    1. edwarnicke___ to propose an alternate time schedule for Helium Release
  3. regXboi
    1. regXboi to document cherrypicking


People present (lines said)

  1. cdub (91)
  2. colindixon (53)
  3. phrobb (40)
  4. dmm (16)
  5. tbachman (10)
  6. edwarnicke___ (9)
  7. Madhu (5)
  8. abhijitkumbhare (5)
  9. odl_meetbot (4)
  10. kwatsen (4)
  11. chrisprice___ (3)
  12. regXboi (3)
  13. LuisGomez (1)


Generated by MeetBot 0.1.4.