16:01:12 <Sukhdev> #startmeeting ironic_neutron
16:01:13 <openstack> Meeting started Mon Aug 24 16:01:12 2015 UTC and is due to finish in 60 minutes.  The chair is Sukhdev. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:15 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:01:17 <openstack> The meeting name has been set to 'ironic_neutron'
16:01:21 <cragusa> hi Sukhdev, this is carmelo
16:01:30 <Sukhdev> cragusa hello
16:01:32 <lazy_prince> o/
16:01:54 <Sukhdev> #topic: Agenda
16:02:01 <Sukhdev> #link: https://wiki.openstack.org/wiki/Meetings/Ironic-neutron
16:02:24 <Sukhdev> Anybody like to add anything to the agenda?
16:02:45 <Sukhdev> #topic: Announcements
16:03:22 <Sukhdev> Neutron FF is next week - which means basically we have this week to push anything into neutron tree for Liberty release
16:03:44 <Sukhdev> I have pushed what ever is needed for this integration based upon our discussions in mid-cycle
16:04:03 <Sukhdev> Can anybody think of anything that needs to be addressed in this regard?
16:04:14 <lazy_prince> k.. cool.. but we need those documented somewhere..
16:04:31 <Sukhdev> lazy_prince: correct -
16:04:42 <lazy_prince> say what to do when writing a mech driver for baremetal..
16:05:20 <Sukhdev> lazy_prince: I am looking for amotoki - he is not here
16:05:47 <lazy_prince> can that be done before FF gets in effect..? also we had to update the specs, can we get that done too..?
16:06:03 <Sukhdev> I chatted with amotoki off-line last week about the documentation - he mentioned he had some family emergency - hence, delays
16:06:17 <Sukhdev> he mentioned he will push the documentation patch this week for review
16:06:19 <lazy_prince> ohh.. no worries then..
16:06:52 <Sukhdev> I will follow up with him - and, if he is not able to foliow through with this, I will try to push it myself
16:06:59 <lazy_prince> but at least, spec update should be there..
16:07:20 <Sukhdev> I will also check the deadline for documentation
16:07:43 <lazy_prince> fine with me..
16:08:00 <Sukhdev> BTW, for ML2 drivers for vendors, this deadline is not applicable - we can push them later
16:08:21 <Sukhdev> Those drivers go into vendor repos
16:09:09 <Sukhdev> Anybody has any announcement?
16:09:15 <Sukhdev> jroll: ?
16:09:51 <Sukhdev> #topic: Integration Effort
16:10:09 <Sukhdev> This is the new topic I added for this week
16:10:28 <Sukhdev> I have been testing the patches - and have hit the blockers
16:10:42 <Sukhdev> I sent an email last week to select people about it
16:11:09 <lazy_prince> yeah.. so I have been working on them this week... and I will be pushing a patch in few days..
16:11:31 <Sukhdev> lazy_prince: did you understand the issue? was I clear?
16:11:50 <lazy_prince> yes.. you were.. for the pep8 issues, that is coming from the one of the patches posted by Laura..
16:12:17 <lazy_prince> if everyone is okay with it, i can push a patch to fix that..
16:12:20 <yhvh_> we have not made any progress on the db bug, I've written a few extra unit tests to try reveal the bug but no luck so far
16:13:06 <Sukhdev> yhvh_: it is very easy to reproduce - I can help you in reproducing the bug -
16:13:12 <lazy_prince> yhvh_: is this about the local_link_connection using ironic-client..?
16:13:27 <Sukhdev> lazy_prince: yup
16:13:43 <yhvh_> yes and also 'ironic portgroup-list' fails for me with the newest posted patches
16:13:56 <lazy_prince> i saw some indentation issues with one of the patches.. could that be the real cause..?
16:14:21 <yhvh_> lazy_prince: could be since this worked before the sprint afaik
16:14:26 <yhvh_> I will try bisect
16:14:36 <Sukhdev> lazy_prince: ah ha.. those can really throw you off - wonders of python :-)
16:14:46 <lazy_prince> i have put it as review comment...
16:15:16 <lazy_prince> https://review.openstack.org/#/c/206243/14/ironic/conductor/manager.py
16:15:24 <lazy_prince> L1624
16:15:50 <Sukhdev> yhvh_: In terms of reproducing the bug, simply execute the CLI to create a port-group and then go to mysql to list the contents of DB table, and you will see those columns are empty
16:15:57 <Sukhdev> yhvh_: very easy to reproduce
16:17:15 <lazy_prince> anyways.. yhvh_ will take a look at it and see if it helps..
16:17:19 <yhvh_> ok thanks
16:17:21 <Sukhdev> lazy_prince: that indentation can be an issue - but, this is for update - we saw the create/add not writing data into DB
16:18:20 <lazy_prince> I have couple of stuffs to work on.. check for binding extensions enabled in neutron and the admin context for port create...
16:18:37 <Sukhdev> I have been able to work around this bug, by manually entering data into DB - but, lazy_prince the create_port() for neutron is real blocker for me now
16:18:47 <lazy_prince> apart from that, fixing test cases to ensure py27 passes..
16:19:08 <Sukhdev> lazy_prince: I can not make any forward progress until we resolve the authentication issue
16:19:17 <lazy_prince> yeah.. i am working on priority to get you unblocked..
16:19:46 <Sukhdev> lazy_prince: perfect - this will help me figure out if we still need anything additional on the neutron side
16:20:10 <Sukhdev> lazy_prince: I am not anticipating anything, but, you never know - until I test it for sure
16:20:14 <lazy_prince> will drop an email once i push the fix..
16:20:22 <lazy_prince> yeah.. true..
16:20:41 <Sukhdev> lazy_prince: there is Silicon Valley OpenStack summit this week here in the valley
16:20:54 <Sukhdev> lazy_prince: I may have to go there for a day or so -
16:21:04 <Sukhdev> it is from Wed thru Friday
16:21:12 <lazy_prince> ohh...k.. good..
16:21:16 <Sukhdev> So, the sooner I get it, the better
16:21:29 <Sukhdev> lazy_prince: pressure pressure :-):-):-)
16:21:48 <Sukhdev> lazy_prince: I am just kidding - take your time - lets do this right
16:21:53 <lazy_prince> so Sukhdev, what is the right time to find you on irc..?
16:22:13 <Sukhdev> lazy_prince: most of the time during day
16:22:28 <lazy_prince> you mean your day.. ?? :)
16:22:41 <Sukhdev> yup
16:22:56 <Sukhdev> lazy_prince: I just created a private channel for you - you should have seen my hello there
16:23:12 <lazy_prince> aha.. and i keep looking for you in my day time..
16:23:19 <lazy_prince> yes.. i saw..
16:23:22 <Sukhdev> lazy_prince: even if I am not on-line, if you drop a message there, I will get it and will respond as soon as I can
16:23:37 <lazy_prince> okay.. will do so...
16:23:41 <Sukhdev> Hope that will help
16:24:05 <Sukhdev> cragusa, yhvh_ : Are you guys blocked on anything?
16:24:39 <Sukhdev> please do not be shy, if you need any help in producing any bug, please let us know - we will jump in to help out
16:25:06 <yhvh_> cragusa has been on vacation and is just back today, I've been busy with other work concerns, hopefully I get to give this more time this week
16:25:35 <Sukhdev> yhvh_: ah I see - no worries
16:25:57 <Sukhdev> yhvh_: did you get the pointer about reproducing the DB issue?
16:26:09 <yhvh_> Sukhdev: yes thanks, looks clear
16:26:21 <Sukhdev> cool
16:26:55 <Sukhdev> Ok - moving on...
16:27:17 <Sukhdev> #topic: Discussion on plan for the Liberty
16:27:31 <Sukhdev> I added this topic on the agenda as well
16:27:49 <Sukhdev> looks like jroll disappeared today :-):-)
16:28:04 <lazy_prince> I think he is on vaccation this week..
16:28:28 <Sukhdev> oh - I did not know that - that make sense -
16:28:57 <Sukhdev> Since we are down to the final weeks of release cycle, I thought it will be good if we discussed and planned
16:29:40 <Sukhdev> Since we are working on various parts of the system - it will be better if we coordinated our efforts so that the we can be effective in testing all the parts
16:30:06 <Sukhdev> Basically, I have been doing the integration testing - i.e. taking all the patches and try to test end-to-end
16:30:20 <Sukhdev> no success so far because of the issues covered already
16:30:57 <Sukhdev> based upon the discussion here, this is how I see - please provide your wisdom as well
16:31:36 <Sukhdev> 1) yhvh_  to go get the DB issue worked out and push a new patch - please send email so that we can jump on it right away
16:31:58 <yhvh_> ok np
16:32:07 <Sukhdev> 2) lazy_prince to push the updated patch to deal with authentication issue - as per plan, please send email so that we can jump on it
16:32:26 <lazy_prince> gotcha
16:32:57 <Sukhdev> 3) Everybody, please look at the etherpad and make sure we have all the right patches listed - if anything does not belong there, please remote it. If something is missing, please add
16:33:15 <Sukhdev> I am going by only the etherpad - lets make sure we are all on the same page
16:34:01 <lazy_prince> so there needs to be a TODO section on the etherpad.. to track pending stuff..
16:34:03 <Sukhdev> 4) When both of those updated patches are available Sukhdev to run the integration to see where the next issue is
16:34:14 <lazy_prince> :)
16:34:27 <Sukhdev> lazy_prince: there is TODO section, please feel free to add to it -
16:34:48 <lazy_prince> aha.. cool..
16:35:12 <lazy_prince> so i will start adding to the TODO ..
16:35:13 <Sukhdev> #link: https://etherpad.openstack.org/p/ironic-neutron-mid-cycle
16:35:43 <Sukhdev> lazy_prince: you had mentioned that you might be able to get somebody from your team to look into the testing
16:35:52 <Sukhdev> lazy_prince: any update on that?
16:36:47 <lazy_prince> yeah.. i have someone working on it.. however, i need the docs for mech driver stuff so that we can start updating ovs mech driver for testing..
16:37:44 <Sukhdev> lazy_prince: I can help…even if the documentation is not ready yet, I can provide needed information
16:37:59 <lazy_prince> or at least a mech driver dummy code to do end to end test simulations...
16:38:14 <Sukhdev> lazy_prince: feel free to kick off an email with him on the email, and I will be happy to provide details via email - so that he can make progress
16:38:43 <Sukhdev> lazy_prince: I will copy amotoki on the email as well - this way we will make sure everybody is on the same page
16:39:07 <lazy_prince> sure.. will check with him.. for now, the guy is going through the specs and etherpad..
16:39:40 <Sukhdev> lazy_prince: cool - just let me know - whenever is the right time - we'll jump on it
16:39:52 <lazy_prince> very well..
16:40:15 <Sukhdev> #topic: Patches under review
16:40:51 <Sukhdev> As discussed earlier - please make sure that etherpad has the correct information - please correct it, if it is not
16:41:09 <Sukhdev> yhvh_ cragusa: I see this - https://review.openstack.org/#/c/206163/
16:41:53 <Sukhdev> I think this is obsolete - can you check, if it is can you abandon it? if not, this is ignored
16:42:16 <yhvh_> ok cragusa or I will check
16:42:34 <Sukhdev> I am worried there may be more cases like this - hence, requesting that please go through the ehterpad and make sure this reflects the correct information
16:43:11 <Sukhdev> any other thoughts?
16:43:47 <Sukhdev> #topic: Testing
16:44:02 <Sukhdev> We have discussed about the testing part already -
16:45:10 <Sukhdev> mitchjameson updated the etherpad with testing strategy - lazy_prince please have the person who is looking at testing to review this section and provide feedback
16:45:24 <Sukhdev> Anything else on the testing?
16:45:33 <lazy_prince> k..
16:46:01 <Sukhdev> We have covered documentation, so I will skip that over and go to last topic on the agenda
16:46:10 <Sukhdev> #topic: Open Discussion
16:46:32 <Sukhdev> We have 15 min. left - we can use it to cover any Q/A
16:47:19 <lazy_prince> yeah.. so the patches under review will not merge unless the tescases pass.. and they will not pass unless we configure them to use none provider..
16:47:33 <lazy_prince> or we tweak the test cases..
16:47:37 <Sukhdev> lazy_prince devananda: do you know what is the absolute last day for code submission to be included in Ironic tree for Libery?
16:48:01 <devananda> Sukhdev: nope. we actually haven't done a good job with that this cycle --
16:48:18 <devananda> in large part because we inteded to move to an independent release model, but haven't actually done a release yet
16:48:33 <devananda> jroll and i are working on that, and meeting with dhellmann later today to see if we can push a 4.0.0 release out this week
16:48:40 <Sukhdev> devananda: we are trying to figure out the plan for remaining weeks of the release cycle - it will help if we know the deadline :-)
16:49:34 <devananda> Sukhdev: the best I can do right now is point you to a) the openstack general timeline, which we're not exactly following, and b) our aspirational description of the process we intend to follow but haven't gotten working yet
16:49:49 <devananda> https://wiki.openstack.org/wiki/Liberty_Release_Schedule
16:50:06 <Sukhdev> devananda: when you work out the details, please let us know so that we are aware of how many days do we have to get patches merged/approved
16:50:18 <devananda> http://specs.openstack.org/openstack/ironic-specs/specs/liberty-implemented/feature-based-releases.html
16:51:06 <Sukhdev> devananda: it seems we can push all the way to the end of Sept - if there are no FF guidelines for Ironic, right?
16:51:14 <devananda> potentially yes
16:51:49 <devananda> Sukhdev: also note that I will be offline from roughly Aug 27 - Sept 8
16:51:53 <Sukhdev> devananda: cool - this helps - we have neutron deadline next week - so, trying to work out the plan so that we don't miss the critical deadlines
16:52:19 <devananda> so my hope is we can do a 4.0.0 release in the next two days, and then either 4.1.0 or 5.0.0 at the end of Liberty
16:52:22 <Sukhdev> devananda: so noted
16:52:59 <Sukhdev> devananda: thanks for clarifying
16:53:45 <Sukhdev> lazy_prince: going back to your point - can you please put this (in BOLD) on the etherpad that we much configure the provide network as none to make the tests pass
16:54:07 <Sukhdev> lazy_prince: or any other hook that is critical
16:54:27 <lazy_prince> point noted...
16:55:01 <Sukhdev> lazy_prince: one more question - how will we make jenkins happy if this needed?
16:55:25 <Sukhdev> lazy_prince: won't jenkins keep voting -1 on all patches and prevent merge?
16:56:28 <lazy_prince> true.. ci job will need to execute this with none provider or else we will have to push ovs baremetal driver to get the network flipping work..
16:56:32 * Sukhdev time check 4 min. left
16:57:24 <Sukhdev> lazy_prince: so, this is what you were mentioning the person from your team is looking at, right?
16:57:41 <lazy_prince> thats the reason, i put it forward.. by default, ironic patches will configure ironic to use neutron network provider which will always upset the ci...
16:58:35 <lazy_prince> kind of.. and to do that, i either need the specs updated and merged or the doc to be ready on how to write a bm mech driver..
16:59:09 <lazy_prince> i hope everyone understand the issue i am talking about..
16:59:25 <Sukhdev> lazy_prince: there is no difference in bm driver or any other existing driver - except for one thing
16:59:59 <Sukhdev> lazy_prince: that is bm driver will look at the binding profile, while the others do not look/act on it
16:59:59 <lazy_prince> yeah.. i just need it to be captured.. so that i avoid unnecessary rework..
17:00:27 <Sukhdev> lazy_prince: agree - let me ping amotoki - later today
17:00:34 <Sukhdev> We are out of time -
17:00:43 <Sukhdev> Thanks for joining today
17:00:46 <cragusa> thanks
17:00:48 <lazy_prince> we already found ir-regularities in the spec..
17:00:56 <yushiro> bye
17:00:59 <Sukhdev> please do send emails for urgent issues, or ping on IRC
17:01:01 <Sukhdev> thanks
17:01:01 <lazy_prince> so wanted to avoid that going forward..
17:01:02 <Sukhdev> by
17:01:09 <Sukhdev> #endmeeting