14:08:00 #startmeeting OFplugin Daily Scrum 14:08:00 Meeting started Wed Aug 6 14:08:00 2014 UTC. The chair is edwarnicke. Information about MeetBot at http://ci.openstack.org/meetbot.html. 14:08:00 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:08:00 The meeting name has been set to 'ofplugin_daily_scrum' 14:08:11 Could everyone please #info in for roll call 14:08:15 #info edwarnicke 14:08:20 #info abhijitkumbhare 14:08:28 #info jmedved 14:08:53 #info prasanna 14:08:56 colindixon: oflibMichal prasanna avishnoi are you guys here? 14:08:57 #info avishnoi 14:09:15 yup 14:09:23 no 14:09:25 :p 14:09:35 #info oflibMichal for openflowjava 14:09:38 We were just discussing edwarnicke - - can you please go over the problems with the current extensibility framework ? 14:09:55 abhijitkumbhare: OK... so here's where we are 14:09:57 (in reality I’m just opening my laptop between things in doctor’s offices, I may vanish at any point) 14:09:58 ok - lets finish the info 14:10:21 #link https://git.opendaylight.org/gerrit/#/c/9393/ - is a patch which *should* give us extensibility (including examples) 14:10:29 Kamal__: can you # info in? 14:10:35 The extension mechanism for actions is pretty clean 14:10:46 The extension mechanism for matches, because of the API difficutiles, is deeply ugly 14:11:01 But I don't think we can change the API for matches for Helium at this point 14:11:24 Kamal__: Did you #info in? 14:11:34 before we start discussing extensibility (which will probably take the rest of the call), can we go over the action item from the call yesterday? 14:11:37 so what do we do to get the OVSDB unstuck? 14:11:39 We do have consumers who desperately need extensions for Helium though 14:11:40 is kamal on the chat? 14:12:08 hi , yes i am on chat 14:12:09 I saw him come in 14:12:14 So my recommendation would be that we run with 9393 and just acknowledge we will have to do something much smarter for Lithium 14:12:24 Kamal__: Could you #info in for the attendance record? 14:12:34 i think i see him in the attendee list, but not #info himself 14:12:37 Sure - that’s good enough 14:12:40 shague: Are you here? 14:13:22 However - is it good enough for whatever needed by OVSDB? shague may have an idea? 14:13:35 kamal: did you have a chance to do anything about the FRM, as discussed yesterday? 14:14:15 well... i cant stay long today. Just want to give an update: Bug 1477 has been taken by Vaclav as there is a gerrit already being done by him. 14:14:16 Guys do we want to start up a tracking spreadsheet like we did for Hydrogen, or do we want to do something else (Trello?) 14:14:28 Kamal__: Could you #link the gerrit in? 14:15:02 Either is fine 14:15:05 Trello 14:15:13 prasanna: Do we have a Trello? 14:15:17 Yes 14:15:21 yes 14:15:21 https://git.opendaylight.org/gerrit/#/c/8174 14:15:25 I'm totally down to use Trello... but need a hand getting going 14:15:30 xls sheet is also fine 14:15:32 Give me a moment 14:15:36 #link https://git.opendaylight.org/gerrit/#/c/8174 - Bug 1477 fix 14:15:42 because lot of people are not there on trello 14:15:43 Vaclav has updated the defect as well 14:15:51 prasanna: I'm fine either way... but if Trello is better I'm down for better :) 14:15:56 https://trello.com/b/TtnyEJZr/helium-release 14:16:05 jmedved: Did you commit your scripts? 14:16:08 anything edwarnicke :) 14:16:46 ewarnicke: no, not yet. ran out of time yesterday. will try today 14:16:47 prasanna: So it looks like we have a good Trello... does everyone who is doing work have an account? 14:16:58 jmedved: Could you at least email them out now? 14:17:07 I have you on the Trello edwarnicke as a member - you probably need to accept 14:17:15 edwarnicke yes, hope so 14:17:29 we will need vaclav on the chat to give update 14:17:33 #link https://git.opendaylight.org/gerrit/#/c/9723/ 14:17:41 let me see if i can get reach him 14:18:04 any idea of 9723?...Rob are you there? 14:18:38 prasanna: Lets get the framework in first, and then we can look at Robs patch :) 14:19:00 Guys... I would suggest the following agenda: 14:19:16 1) Decide if we are using Trello (it looks like we already decided to, which is awsome) 14:19:30 2) Check to see if all the folks doing work have Trello accounts 14:19:52 3) Check in on Kamal__ and 1477 (it looks like we have a patch... which is awesome) 14:20:00 4) Decide what we are doing about extensions 14:20:13 i am im 14:20:13 5) Figure out other action items and get them assigned 14:20:18 Does that sound good to folks? 14:20:29 i am im’ ing vaclav - looks like he’s online 14:20:30 Yes 14:20:32 #chair prasanna abhijitkumbhare jmedved Kamal__ 14:20:32 Current chairs: Kamal__ abhijitkumbhare edwarnicke jmedved prasanna 14:20:33 he’s coming 14:20:45 #topic Using Trello 14:20:54 Did we decide previously to use Trello? 14:21:07 yes 14:21:08 Yes 14:21:11 Perfect 14:21:21 #agreed We previously decided to use Trello for tracking 14:21:23 We have the tasks on Trello currently 14:21:34 #info We have the tasks on Trello currently 14:21:37 But we will now need more granular tasks 14:21:47 Anything else for that topic, or can we move to #2? 14:21:51 abhijitkumbhare: Totally :) 14:22:37 ? 14:22:45 I think the following folks do not have Trello accts - edwarnicke, jmedved , oflibMichal 14:22:49 Going once... going twice... 14:23:06 #topic Getting everyone a Trello account 14:23:12 also debolina & muthu 14:23:49 OK... I *think* I've subscribed now on trello... abhijitkumbhare prasanna can you confirm? 14:24:04 Not as per that sheet 14:24:12 Yes I have account 14:24:28 prasanna is on https://trello.com/b/TtnyEJZr/helium-release 14:24:28 abhijitkumbhare: I'm hagbard 14:24:35 OK 14:24:44 Then let me re-include you 14:25:14 oflibMichal: jmedved Do you have Trello accounts? 14:25:24 creating one right now 14:25:37 oflibMichal: Many thanks :) 14:25:57 I added edwarnicke as hagbard 14:26:05 abhijitkumbhare: Many thanks :) 14:26:16 jmedved: Can you create a Trello account? 14:26:40 avishnoi: are you in on the Trello? 14:26:49 yup 14:26:54 OK 14:26:54 not sure. maybe 14:27:04 Add Rob also he is pushing lot of code into Nicira extension 14:27:16 prasanna: Good point 14:27:21 But I don't think we have Rob here right now 14:27:29 OK 14:27:57 add martin sunal too - he did some (quite abit) of the work rob is pushing 14:28:58 #topic Bug 1477 - FRM/Statistics Manager: figure out flow indexing - https://bugs.opendaylight.org/show_bug.cgi?id=1477 14:29:10 #link https://bugs.opendaylight.org/show_bug.cgi?id=1477 - FRM/Statistics Manager: figure out flow indexing 14:29:29 Kamal__: You said that https://git.opendaylight.org/gerrit/#/c/8174 was for 1477 14:29:35 But it claims to be for 537 14:29:48 jmedved: added martin sunal 14:30:46 Kamal__: You stil there? 14:30:50 edwarnicke: you can consider 1477 duplicate of 537 14:30:59 The FRM changes mentioned in the 1477 bug desc are part of 537 14:31:03 avishnoi: Could you mark it as such in Bugzilla? 14:31:20 okay 14:31:38 #action avishnoi - mark 1477 as a dup of 537 in bugzilla 14:31:54 OK.. someone needs to review and try out https://git.opendaylight.org/gerrit/#/c/8174 14:31:59 Who can do that for tomorrow? 14:32:15 i can try 14:32:21 and run my scripts against it 14:32:47 avishnoi review https://git.opendaylight.org/gerrit/#/c/8174 ? :-) 14:32:59 ouc :-) 14:33:06 i am looking at it 14:33:09 ok 14:33:19 v_demcak: is it ready to be tested? 14:33:19 #action jmedved To run his scripts against https://git.opendaylight.org/gerrit/#/c/8174 for tomorrow 14:33:37 8174, i mean 14:33:37 avishnoi: did you agree to review and try out https://git.opendaylight.org/gerrit/#/c/8174 for tomorrow? 14:33:53 edwarnicke: 1477 tagged duplicate of 537 14:33:54 yes it is ... i've retested and make some last fix ... so I hope it is prepare for merge 14:34:05 abhijitkumbhare: Would you take on converting the #actions from the meeting minutes into Trello cards? 14:34:14 sure 14:34:33 ok, let me build and apply the gerrit 14:34:37 #action abhijitkumbhare - Convert #action from meeting minutes to Trello cards 14:34:39 i am going to give it a workout 14:34:47 jmedved: Many thanks :) 14:35:00 avishnoi: Can you try out the patch for tomorrow? 14:35:20 yes 14:35:56 #action avishnoi To try out and review https://git.opendaylight.org/gerrit/#/c/8174 for tomorrow 14:36:23 Shall we move onto extensions? 14:36:28 yes 14:36:49 #topic Extensions 14:37:26 #link https://git.opendaylight.org/gerrit/#/c/9393/ - should give us extensions... pretty for actions, less pretty for matches 14:37:39 #info our users desperately need extensions 14:37:51 So guys... what are your thoughts on getting this in? 14:38:37 abhijitkumbhare: avishnoi prasanna jmedved v_demcak opinions? 14:39:03 i think we should ask this to madhu, how much code are they writing on top of match extension 14:39:32 if that's something which they can re-write once we re-work match 14:39:32 Yes avishnoi - sounds right. 14:39:44 avishnoi: I asked networkstatic last night 14:39:51 avishnoi: He said some 14:39:52 i believe whatever plan ed proposed sounds good to me 14:40:00 avishnoi: But his general opinion was something is better than nothing 14:40:07 I also know SFC needs match extensions 14:40:14 edwarnicke: i agree 14:40:17 OK... so can we do this 14:40:19 What are requirements from ovsdb team, can they do with actions for now, until we look into matches 14:40:34 Can we get some folks to signup to try the patch out for tomorrow, and if it checks out we'll merge it? 14:40:51 prasanna: Note... matches are reported to *work*... they are just ugly as hell 14:41:01 oh ok 14:41:58 So... who an signup to try this out for tomorrow, and what do you need to be able to successfully do so? 14:42:04 btw. why is the Rob 14:42:18 v_demcak: I don't understand your question 14:42:41 sorry why i the Rob's nicira support abandoned ? 14:42:53 v_demcak: I don't know 14:43:10 prasanna: Could you try out 9393 for tomorrow? 14:43:12 edwarnicke , abhijitkumbhare what are the issues we are seeing with matches 14:43:18 jmedved: Could you run your scripts against it? 14:43:51 against what? 14:43:59 I can review https://git.opendaylight.org/gerrit/#/c/9393/ - but I think prasanna & probably networkstatic /madhu should also review it 14:44:04 https://git.opendaylight.org/gerrit/#/c/9393/ 14:44:10 abhijitkumbhare: Can you build it and try it out? 14:44:16 will try 14:44:29 OK - later in the day though 14:44:37 #action jmedved - try out https://git.opendaylight.org/gerrit/#/c/9393/ 14:44:43 #action abhijitkumbhare - try out https://git.opendaylight.org/gerrit/#/c/9393/ 14:44:51 prasanna: Are you still there? 14:44:52 will build and try it .. 14:45:06 #action prasanna - try out https://git.opendaylight.org/gerrit/#/c/9393/ 14:45:47 since prasanna is the last one on the chain - probably # action for prasanna to merge it as well? 14:46:17 will do it ... but can we ask ... ovsdB team member to also try it 14:46:20 abhijitkumbhare: Do we want to decide that tomorrow morning together? 14:46:24 since they are immediate users 14:46:27 OK 14:46:38 abhijitkumbhare: Can you take an action to get the OVSDB guys to try it out? 14:46:46 yes 14:46:59 Guys... I have some helpful instructions for trying it otu 14:47:07 Where should I put them in the wiki? 14:47:07 #action abhijitkumbhare to reach out to OVSDB folks to try out https://git.opendaylight.org/gerrit/#/c/9393/ 14:47:25 that would help ....edwarnicke 14:47:30 Michal has the wiki 14:47:37 will just find it’ 14:47:45 Cool... I can update it 14:48:13 https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Backlog:Extensibility 14:48:49 should be moved out of backlog - but can be moved later 14:48:58 i meant the wiki 14:49:46 #action edwarnicke to update instructions to try out https://git.opendaylight.org/gerrit/#/c/9393/ at : https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Backlog:Extensibility 14:51:13 abhijitkumbhare: Done: https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Backlog:Extensibility#Helpful_hints_for_reviewing_https:.2F.2Fgit.opendaylight.org.2Fgerrit.2F.23.2Fc.2F9393.2F 14:51:29 Great - thank edwarnicke ! 14:52:07 Guys... please look quickly at that and let me know if you are left with questions 14:53:00 Anything else on this topic? 14:53:38 No ... will go through the wiki and ping you offline 14:53:43 Do we want to talk briefly about other things we need to work on for the Release? 14:53:50 Try and build out some forward plans 14:53:51 ? 14:53:56 The critical thing is to check if this will satisfy OVSDB 14:54:22 abhijitkumbhare: Yes... but honestly... I think that (modulo bugs) this is the best we can do without changing the match API 14:54:33 OK 14:54:42 #topic Other items 14:54:55 So... what other things do we need to do 14:55:00 We need to get the VTN folks unstuck 14:55:09 abhijitkumbhare: In what way are they stuck? 14:55:26 They seem to be having a problem in adding flow to 1.3 switches 14:55:36 I am aware of one... we need to enforce action order based on the order field... 14:55:49 avishnoi had given them hints - not sure if they helped? 14:56:12 abhijitkumbhare: I have one patch for the order problem here: https://git.opendaylight.org/gerrit/#/c/9214/ 14:56:16 But I didn't get a chance to test it 14:56:21 OK 14:56:30 prasanna: Would you be willing to pick it up? 14:57:04 I think avishnoi had some ideas about this - can you summarize avishnoi ? 14:57:12 looking into the gerrit 14:57:25 prasanna: this patch may even work out of the box 14:57:33 the net net is we should respect the order being set 14:57:58 abhijitkumbhare: are you point to the PUSH_VLAN action related issue 14:58:17 yes - the issue that hideyuki mentioned 14:58:19 abhijitkumbhare: Can you take an action to get the vtn folks on the meeting tomorrow so we can crisply sort out their issues? 14:58:24 edwarnicke: remember the issue which we discussed over mailing list regarding how to handle PUSH_VLAN 14:58:32 Oh.. yes 14:58:36 The semantic issue 14:58:42 Did we reach a conclusion? 14:58:43 yes 14:58:45 Yes agree Ed 14:59:05 avishnoi: prasanna What are you agreeing with me about ? ;) 14:59:23 edwarnicke: about the issue ) push_vlan 14:59:23 sure - will ask them (and actually the OVSDB folks for tomorrow - OVSDB more urgent) 14:59:30 abhijitkumbhare: Also good 14:59:46 Did we reach a conclusion about PushVlan though? 15:00:32 avishnoi's "yes" was for conclusion I suppose 15:00:36 we have a solution, but the issue i see is , for one config flow, we are installing two flows, how we will augment statistics in that case 15:00:43 What was the conclusion? 15:00:56 avishnoi: Any smart ideas there? 15:01:10 conclusion was to install two flows 15:01:19 I need to think about it 15:01:25 OK... do we have something in the API to express the semantics they are looking for? 15:01:38 because if we augment statistics with the same id, they will stomp each other 15:02:12 avishnoi: More than that... there should be a correspondence between the configured flow and the operational flow 15:02:27 no, i think this is something which need to be handled when we convert md-sal flow to of-flow 15:03:09 so how about OVSDB folks we start of the discussion tomorrow (about extensibility - first half hour) 15:03:15 avishnoi: And back again? ofFlow to mdSal flow? 15:03:20 VTN for 10-15 min after that 15:03:27 abhijitkumbhare: Why not interleve them 15:03:32 No need to block off different times ;) 15:03:35 sure 15:03:37 ok 15:03:49 Guys.. do we have the API semantics for the fix for the VTN guys? 15:04:00 edwarnicke: we can associate these flows using the flow-id 15:04:33 and ideally we should augment statistics of both the flows to same config flow 15:04:39 Cool... so the critical question is... do we have the semantics in the API for this? 15:04:42 and we can't do it with same flow-id 15:04:43 If so, what are they? 15:05:07 #info Item - PushVlan semantics for VTN guys 15:05:51 Guys? 15:06:12 I think the critical path to figuring out how to do the PushVlan thing for VTN is to figure out if we have the API semantics 15:06:14 edwarnicke: can you explain your question a bit 15:06:16 If we do, we can probably do it 15:06:19 If not we cant 15:06:20 OK 15:06:35 avishnoi: What action would we translate to two flows? 15:06:56 SET_VLAN_ID 15:07:08 do we have SET_VLAN_ID currently in our API? 15:07:38 yes, i think so, and frm converts it to appropriate openflow flow 15:07:44 Will run off to write to Madhu_offline & networkstatic about the code review, match/actions, etc. 15:07:46 Cool 15:07:47 prasanna: please correct me if i am wrong 15:07:55 So in principle we can do this 15:08:06 before the 8:30 am Pacific release meeting 15:08:15 yes 15:08:15 avishnoi your right we have it 15:08:28 Do we have a bug for this yet? 15:10:10 Yes, i think so, need to extract is from mail 15:11:02 Who can take an action to find the bug from the mail, put a pointer in the bug to the email thread, and create a Trello card pointing to the bug? 15:11:34 edwarnicke: i can do that 15:11:40 For tomorrow? 15:12:03 yes, hopefully creating trello card is not very complex :) 15:12:08 #action avishnoi find the bug from the mail for VTN PushVlan issue, put a pointer in the bug to the email thread, and create a Trello card pointing to the bug 15:12:29 abhijitkumbhare: Can you prepare a link for us for bugs we need to scrub for tomorrow? 15:12:46 OK 15:13:21 abhijitkumbhare: Could you queue up the actions for today for us to review at the top of tommorrows meeting? 15:13:28 here is the bug link : https://bugs.opendaylight.org/show_bug.cgi?id=1421 15:13:36 OK 15:13:48 #action abhijitkumbhare to prepare a link for us to use for bug scrub tomorrow 15:14:00 #action abhijitkumbhare to queue up actions from todays meeting for us to go over tomorrow 15:15:08 OK guys... anything else before we wrap for the day? 15:15:48 I am good 15:15:54 prasanna: You good ? 15:15:56 nothing from my side 15:16:11 #endmeeting