13:00:31 <fuqiao> #startmeeting High Availability Project Meeting 13:00:31 <collabot`> Meeting started Wed Mar 21 13:00:31 2018 UTC. The chair is fuqiao. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:31 <collabot`> Useful Commands: #action #agreed #help #info #idea #link #topic. 13:00:31 <collabot`> The meeting name has been set to 'high_availability_project_meeting' 13:00:39 <fuqiao> #chair Fuqiao 13:00:39 <collabot`> Current chairs: Fuqiao fuqiao 13:01:26 <georgk> #info Georg Kunz 13:01:34 <georgk> hello, hello 13:01:35 <Kanglin> #info Kanglin 13:01:37 <mardim> #info Dimitrios Markou 13:01:44 <fuqiao> Hi, all. 13:01:58 <fuqiao> #topic HA test case update 13:02:14 <mardim> I can start if you want guys 13:02:20 <mardim> I will be quick 13:02:44 <mardim> #info ODL noHA testcase is merged 13:02:54 <fuqiao> #info we would like to go through all the test cases we have so far. See which cases are ready for merge. and ask Ross to merge certain project before next week 13:03:10 <fuqiao> Mardim, thanks. 13:03:14 <mardim> #info ODL HA testcase I have a draft 13:03:26 <mardim> #link https://gerrit.opnfv.org/gerrit/#/c/53995/ 13:03:46 <fuqiao> #info TC087 ODL noHA test case is merged 13:03:59 <mardim> #info TO merge this in we should fix apex integration with yardstick 13:04:17 <mardim> #info we need apex pod.yaml for that 13:04:36 <fuqiao> is it possible to finish for F release? otherwise we may submit a release exception 13:04:37 <mardim> #action mardim will fix the Apex integration with yardstick 13:04:37 <Kanglin> so apex supports odl ha scenario? 13:04:57 <mardim> it will take time because I need to fix to many things 13:05:09 <mardim> not sure if it will be ready for F 13:05:14 <mardim> probably not\ 13:05:17 <mardim> not* 13:05:33 <fuqiao> #info TC085 & TC086 can't merge for F, because of lack of installer supporting VM HA and Hypervisor HA. 13:05:42 <fuqiao> Mardim: got that. 13:05:57 <mardim> fuqiao, that's all 13:06:09 <fuqiao> Mardim: thanks. 13:06:11 <mardim> thanks guya 13:06:15 <mardim> guys* 13:06:26 <Kanglin> I think tc090 and tc091 is ready to be merged 13:07:19 <fuqiao> #info VM HA feature comes out with Queen release. and we are currently working with compass and XCI to make it integrated for G release. therefor TC085 & TC086 probabally should targeting for G 13:07:37 <fuqiao> #info TC090 & TC091 ready for merge 13:07:55 <fuqiao> Kanglin: how about TC088 & TC089? 13:08:03 <fuqiao> I think you also submit these two? 13:08:51 <Kanglin> yes, but it seems that there is no ha mechanism for recovery of nova-scheduler and nova-conductor 13:09:22 <fuqiao> OK. same problem with TC085 & TC086 13:09:39 <Kanglin> these two cases can be merged but i'm afraid ci will fail if we add these cases to test suites 13:10:00 <fuqiao> I actually think this makes no sense, we can only test something that can certainly pass the test. Then why should we test that... 13:10:04 <mardim> yardstick has no mechanish to disable testcases 13:10:37 <mardim> so it they are failing they will try to fix them and not disable them 13:10:42 <mardim> if* 13:10:44 <georgk> mardim: doesn´t Yardstick h ave a config file for this? 13:10:52 <mardim> georgk, no 13:11:18 <mardim> I think they are not working the same way as functest 13:11:22 <fuqiao> Kanglin: could you help go to Ross and ask for merging TC090 & TC091 before next week? 13:11:44 <Kanglin> fuqiao: sure 13:11:55 <fuqiao> So for Fraser, we will only have TC087 TC090 TC091 13:12:19 <fuqiao> #action Kanglin will reach out to Ross to have TC090 & TC091 merge before next week. 13:12:31 <Kanglin> fuqiao: is there any other patches ready to be merged? I can make a list to Ross 13:12:49 <mardim> it's a no from my side 13:12:55 <fuqiao> Kanglin, I think no 13:12:59 <Kanglin> such as the description file of SDN no-ha patch 13:13:00 <georgk> I need to get the test spec merged 13:13:06 <georgk> Kanglin: yes 13:13:32 <fuqiao> So we can get the test spec merged first, but with no test case? 13:13:56 <mardim> fuqiao, That looked pretty weird in my eyes 13:14:00 <mardim> :) 13:14:07 <Kanglin> fuqiao: test case for SDN no-ha is already merged 13:14:22 <fuqiao> Oh yes. got that. I misunderstood 13:14:42 <fuqiao> Georg, could please put a link here for Kanglin? 13:15:12 <georgk> #link https://gerrit.opnfv.org/gerrit/#/c/49245/ 13:15:37 <fuqiao> #info for Fraser release, TC085 TC090 TC091 are ready to merge. other test cases have issue with passing CI 13:15:46 <Kanglin> BTW, is there any requirements for yardstick when you implementing HA test cases? yardstick is planning for new features of g release now 13:16:11 <fuqiao> #action fuqiao will submit release exception for the other test cases 13:16:36 <fuqiao> Kanglin: I will ask Wangmeng. They suggest some features for HA test cases 13:16:45 <mardim> Kanglin, Form my side there are missing some pod.yaml 13:16:50 <mardim> from* 13:16:52 <fuqiao> He is not at the meeting today 13:19:27 <georgk> fuqiao: regarding the release exception, you are referring to David´s email 13:19:33 <georgk> ? 13:20:01 <fuqiao> Yes. He ask me to submit a release exception. I have never done this before, not quite sure what that is? 13:21:24 <georgk> I think, he is looking at this page: https://wiki.opnfv.org/display/SWREL/Documentation+Compliance+for+Fraser 13:21:59 <georgk> do we point to the Yardstick documentation in this case? 13:22:59 <fuqiao> Georg: I add the HA doc git repo. You mean I should also add yardstick doc repo? 13:24:15 <fuqiao> Wangmeng: Kanglin asks if we have any suggestiong for new feature in yardstick. Yardstick begin discussion on features for G release 13:27:15 <wangmeng> feature? 13:28:15 <fuqiao> Yes. I think we discuss this before. e.g. the yardstick framework support sending centain packets 13:29:17 <Kanglin> fuqiao: Another thing I want to konw. What's the status about the HA Analysis document I submitted? Can it be merged? 13:29:45 <fuqiao> Kanglin: yes, I think it is ready for merge. I will work on it this week. 13:29:54 <Kanglin> fuqiao: thanks 13:30:03 <fuqiao> I may modify it a little and reupload to the docs repo 13:32:21 <georgk> I do have one more topic I´d like to bring up (just to mention it already) 13:33:34 <fuqiao> #action Fuqiao will look into the HA analysis doc and merge it before this week 13:34:14 <wangmeng> Oh, I see. I think yardstick framework have to support keypairs provided by owner while creating VM in the G release. 13:35:03 <Kanglin> wangmeng: you can conclude all the issues and mail them to me later 13:35:13 <wangmeng> OK 13:36:56 <fuqiao> Georg: do you have another topic? 13:37:32 <georgk> yes, 13:38:11 <georgk> I have been in discussions with folks at OPenStack 13:38:22 <georgk> at the last PTG and at the last summit 13:38:42 <georgk> there are efforts to create a new framework for ¨extreme testing¨ as they call it 13:38:57 <georgk> it also covers resilience tests 13:39:16 <georgk> I advertised Yardstick quite a bit to avoid that they start from scratch 13:39:35 <georgk> so, at the end of the day, there are cross-community discussion planned 13:39:54 <georgk> and it would be good if more representtation from OPNFV would be there 13:40:00 <fuqiao> OK 13:40:16 <fuqiao> when will the discussion happen? 13:40:17 <georgk> since you guys and girls have build the HA part of Yardstick, it might be interesting for you 13:40:26 <georgk> #link https://etherpad.openstack.org/p/extreme-testing-contacts 13:40:28 <Kanglin> georgk: Is there any document I can refer to this topic? my current academic research is also about cloud reslience 13:41:04 <georgk> Kanglin: not sure, mostly prototyping work and email discussions 13:42:01 <georgk> This is driven from Adam Spiers, who also runs the self-healing SIG 13:42:23 <Kanglin> georgk: ok, I'll read the etherpad first. I'm interested about it 13:42:29 <georgk> the first step would be a join call to talk about requirements, goals and tools 13:42:38 <georgk> joint 13:43:04 <georgk> if you have more questions, just reach out to me or to those guys 13:43:40 <fuqiao> OK. I will first sign up in the wikipage 13:43:52 <Kanglin> georgk: got it, thanks 13:45:35 <fuqiao> got that. 13:45:42 <fuqiao> Any other topics? 13:46:09 <fuqiao> If not, I will close the meeting. 13:46:18 <fuqiao> Thank you all! 13:46:51 <georgk> thank you! Bye 13:46:56 <mardim> Bye 13:46:57 <fuqiao> #endmeeting