16:00:45 #startmeeting ironic_neutron 16:00:46 Meeting started Mon Aug 17 16:00:45 2015 UTC and is due to finish in 60 minutes. The chair is Sukhdev. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:48 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:50 The meeting name has been set to 'ironic_neutron' 16:01:12 #topic: Agenda 16:01:16 #link: https://wiki.openstack.org/wiki/Meetings/Ironic-neutron 16:01:27 Welcome everybody 16:01:34 hi 16:01:46 o/ 16:01:51 Anybody like to add anything to agenda? 16:02:18 #topic: Announcements 16:02:29 I got no announcement this week - 16:02:37 Does anybody have any? 16:03:26 other than that some of us were in mid-cycle sprint - which we will cover in next topic 16:03:45 * jroll is here now 16:03:54 #topic: Ironic mid-cycle Sprint update 16:04:27 bunch of us participated in the mid-cycle sprint 16:04:36 please see the etherpad for details - https://etherpad.openstack.org/p/ironic-neutron-mid-cycle 16:05:01 I saw updates to our patchset, was that a result of the sprint? 16:05:13 we have been reviewing them 16:05:20 We made one decision - not to use bind_requested flag 16:05:35 yhvh: yes, we had a chance to test your patches 16:06:08 thanks to all involved 16:06:10 Sukhdev: did you end up getting things working? :) 16:06:30 jroll: we came very close :-) 16:06:47 what kind of problems did you encounter? 16:07:02 yhvh: One issue we noticed with your patch was - 16:07:25 When we create the port groups, we did not see link-information going into the DB 16:07:46 We had to manually fill in the DB to make forward progress in testing 16:07:57 this is something you may want to look into 16:08:02 will do 16:08:35 interesting 16:08:57 yhvh: We were also confused about the exact syntax as to how to specify the link-information - so, we guessed few things - 16:09:09 yhvh: it may the side effect of that as well :-) 16:09:46 noted 16:09:50 so, another feedback for you is - for the help string for these APIs, you may want to specify the format of the input requred 16:10:45 For those who were not there, just to give you a quick run-down 16:10:47 this is all good, any more? 16:11:45 we took all the relevant patches listed on the etherpad and put them all together in one work space with the intent 16:11:53 to test end-to-end flows 16:12:55 After we filled in the link-connection information manually in the DB, we were able to see that network flip logic driver was able to fetch this information 16:13:08 and package it correctly to ship to neutron 16:13:44 In the process we realized that we needed a patch in Nova to define the attributes for the vnic_type 16:14:02 So, we pushed a patch for that - that made neutron/nova handshake happy 16:14:36 We ran into some entry point issues for the driver, which consumed a lot of time 16:14:59 eventually, we ended up running out of time and could not get the end-to-end flow to work 16:15:46 The good news is that I have saved the entire setup and will try to pick it up from where we left off - and will see how far do I go with this 16:16:09 will need updated patches from lazy_prince 16:16:25 lazy_price was here - looks like he disappeared 16:17:07 I was going to ask lazy_prince to add to it - but, looks like he dropped off 16:17:50 yhvh: so, if you could look into the issue I mentioned and push an updated patch, I will try to test it later 16:18:04 I'm on it 16:18:18 I'll be fixing up the nova patch this week 16:19:00 jroll: can you see if you can have my one line nova patch approved by nova team? 16:19:50 jroll: do you want to add any update from the sprint? 16:20:06 Sukhdev: I'll put it on a special list 16:20:12 nope, I think you covered everything :) 16:20:26 Ah one more thing - 16:21:19 had a discussion with devananda jroll and others that we will push for merging all of neutron and ironic code into liberty 16:21:38 even if we are not able to get everything merged in nova 16:22:08 this makes it easier for the customers/vendors to test this solution with liberty (along with one or two patches for nova) 16:22:14 seems like a good way to keep pressure up 16:22:37 So, we should really push to make into liberty 16:22:41 yhvh: yup 16:23:23 Anybody wants to add anything? 16:23:56 * Sukhdev waiting 16:24:19 nope, just thanks for all your efforts! 16:24:20 Any questions/concerns 16:24:57 moving right along, then - 16:25:04 #topic: Tesing 16:25:18 We also discussed about testing during the sprint 16:25:38 mitchjameson captured the testing part on the etherpad as well - 16:25:59 please have a look and update, if needed 16:26:57 mitchjameson jroll : do you guys want to elaborate anything about the testing discussion? 16:28:12 please do look at the etherpad for details 16:28:14 I think the etherpad says it all 16:28:40 jroll: yup 16:28:48 yeah.. etherpad has all.. 16:29:07 lazy_prince: welcome back - was looking for you earlier 16:29:26 Sukhdev: system rebooted.. so lost connection 16:29:29 lazy_prince: want to add anything about the sprint testing - other than what I have already said 16:29:32 now its all good.. 16:29:41 nope.. 16:29:43 There are some open questions about testing on the etherpad, go ahead and comment if you have thoughts there 16:30:08 mitchjameson: good point - 16:30:10 sure.. will take a look.. still recovering from travel.. 16:30:40 everybody - please have a look into the testing section, and please try to share your wisdom in answering those questions 16:31:19 They are mostly a question of how rigorous we want to be 16:32:23 jroll: Under the TODO list there is one action from nova - set the vnic_type on the initial port-create reqest 16:32:36 umm.. i would want to make it work first with simple test cases and then worry about making them rigorous.. 16:32:54 lazy_prince: ok 16:32:57 err 16:32:58 Sukhdev: ok 16:34:49 jroll: so, shall we have a separate patch for this or you want to incorporate this fix into your nova patch - that you will be updating? 16:36:13 Sukhdev: I'll have to see how much code it is, maybe a separate patch maybe not 16:36:27 jroll: ok 16:36:41 moving right along - 16:36:50 #topic Documentation 16:37:36 I noticed amotoki is not here - he had signed up for the neutron side documentation 16:38:03 I will follow up with him offline and see if he has made any progress on this 16:38:13 other than that I do not have any update on this 16:38:39 anybody has anything on the documentation? 16:38:42 * Sukhdev waiting 16:39:02 #topic Open Discussion 16:39:21 OK - we are at the end of agenda 16:39:40 Anybody wants to discuss anything or bring up anything? 16:40:00 lazy_prince: I will test a bit more from where we left off last week 16:40:26 lazy_prince: Will see how far can I proceed 16:40:37 yeah.. Sukhdev: can you drop me the diff for the code we had..? that would make my life a bit easy.. 16:40:59 lazy_prince: In the mean time, if you can work on updating the patches 16:41:15 Sukhdev: sure.. 16:41:18 lazy_prince: yes, right - I will shoot you the diff of everything that I have 16:41:44 Sukhdev: thanks.. 16:41:49 lazy_prince: I do not know if I have your email - send me a hello email at sukhdev@arista.com 16:42:07 lazy_prince: I will reply back with the diff 16:42:40 Anything else? 16:42:53 Anybody wants to cover anything? 16:42:57 yhvh: for the doc part, i need some insight on how to populate data for the portgroup and we can use the ironicclient.. 16:43:30 yhvh: this way you get active feedback from all of us.. 16:43:50 np, updating the help strings should be a good way to communicate this? 16:44:04 yup.. that should do.. 16:44:29 if you need us to pick up some patches, do let us know.. 16:45:14 ok thanks! 16:46:08 Anything else? 16:46:16 we have 14 min 16:46:20 Sukhdev: hello mail sent. 16:46:21 * Sukhdev sukhdev 16:46:29 lazy_prince: cool - thanks 16:46:36 * Sukhdev waiting 16:48:14 OK - I guess, we are done 16:48:21 Thanks folks.. 16:48:22 bye 16:48:30 #endmeeting