13:00:49 <fuqiao> #startmeeting OPNFV high availability project meeting
13:00:49 <collabot`> Meeting started Wed Mar  7 13:00:49 2018 UTC.  The chair is fuqiao. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:00:49 <collabot`> Useful Commands: #action #agreed #help #info #idea #link #topic.
13:00:49 <collabot`> The meeting name has been set to 'opnfv_high_availability_project_meeting'
13:01:00 <fuqiao> #chair fuqiao
13:01:00 <collabot`> Current chairs: fuqiao
13:01:22 <fuqiao> hi, all. please info yourself
13:01:25 <fuqiao> #info fuqiao
13:01:44 <georgk> #info Georg Kunz
13:01:44 <GregWaines> #info GregWaines
13:01:46 <Stefan___> Hi, this is Stefan
13:01:48 <mardim> #info Dimitrios Markou
13:01:55 <georgk> hello, hello!
13:02:00 <mardim> hello !
13:02:26 <fuqiao> hi, all. I would like to use this call to syncrinize the current progress.
13:02:44 <fuqiao> Mardim: could you please help give an update for the SDN HA test cases?
13:03:00 <mardim> Yes sure
13:03:02 <fuqiao> are the bugs figured out last week solved?
13:03:09 <mardim> So the testcase is ready
13:03:19 <mardim> let me find the link
13:03:27 <mardim> #link https://gerrit.opnfv.org/gerrit/#/c/52159/
13:03:34 <mardim> I marked it as -1
13:03:43 <mardim> because it has too many dependent patches
13:03:52 <mardim> so I should wait until those patcjes get merged
13:03:59 <mardim> Regarding the bugs
13:04:13 <mardim> the aforementioned patches solve some of them
13:04:19 <mardim> let me paste the links
13:04:22 <fuqiao> #topic HA testcase update
13:04:34 <mardim> #link https://gerrit.opnfv.org/gerrit/#/c/52933/
13:04:44 <mardim> #link https://gerrit.opnfv.org/gerrit/#/c/53027/
13:04:52 <mardim> #link https://gerrit.opnfv.org/gerrit/#/c/53029/
13:04:55 <fuqiao> #info SDN HA testcase is ready
13:05:03 <mardim> #link https://gerrit.opnfv.org/gerrit/#/c/53031/
13:05:12 <mardim> #link https://gerrit.opnfv.org/gerrit/#/c/53099/
13:05:18 <fuqiao> #info waiting for the bug fix patches merged
13:05:20 <mardim> nad one link which is merged
13:05:32 <mardim> #link https://gerrit.opnfv.org/gerrit/#/c/52931/
13:05:40 <mardim> that;s all for now
13:05:48 <mardim> georgk, did you tested the testcase
13:05:49 <mardim> ?
13:05:51 <mardim> BTW ^
13:06:04 <fuqiao> so there are 6 bugs related to the current test case?
13:06:18 <mardim> these are not only bugs
13:06:26 <mardim> these are missing features too
13:06:45 <fuqiao> OK.
13:07:00 <georgk> not yet, sorry
13:07:07 <fuqiao> another question is , I think we have 2 test cases, but I see you give only one link for the test case
13:07:09 <mardim> georgk, no problem just asking :)
13:07:27 <mardim> Yes didn't started the other one yes
13:07:29 <mardim> yet*
13:07:39 <mardim> because I need a High availability installer
13:07:46 <mardim> and right now I have only one server
13:07:54 <mardim> with a noHA installer deployed
13:08:01 <mardim> also Fuel doesn't support
13:08:06 <mardim> ODL clustering yet
13:08:26 <mardim> which means that I should find an installer with ODL clustering enabled
13:08:32 <mardim> in a HA deployment
13:08:42 <georgk> mardim: I think Apex does
13:09:04 <mardim> Yes but right now I have only one server
13:09:07 <georgk> mardim: at least Peri told me this - however, with the intended changes of Apex, I am not sure how stable that installer will be
13:09:15 <fuqiao> OK. seems that it is quite challege for us to catch the MS6 deadline
13:09:23 <mardim> and I want first to merge the noha testcase before undeploy fuel
13:09:48 <mardim> georgk, from internal feedback I know that s not so stable
13:10:07 <georgk> fuqiao: yes, the test cases was quite complex to implement
13:10:08 <mardim> But I prefer to merge the noha first before undploy
13:10:16 <mardim> undeploy*
13:10:27 <fuqiao> Georgk: OK. got that
13:10:49 <georgk> I think the non-HA test case created a good basis for the other one
13:10:50 <fuqiao> #info challenging to meet MS6 for SDN HA test cases
13:11:05 <mardim> georgk, Yes I think the same
13:11:15 <mardim> because all the required fetaures are the same
13:11:21 <mardim> so the only think we will need to do
13:11:32 <mardim> is just to write the testcase yaml I think ideally
13:12:09 <georgk> ok, cool
13:12:13 <mardim> :)
13:12:13 <wangmeng> #info Wangmeng
13:12:22 <georgk> let´s try to get the non-HA test in before the deadline
13:12:38 <georgk> and then, there are still 3 weeks after that, I think
13:12:43 <georgk> if the HA one is simple, fine
13:12:47 <mardim> I aggree
13:12:50 <mardim> georgk, BTW
13:12:52 <fuqiao> for the second usecase, from my understanding, the biggist difficulty is you can't find a sutible testing environment
13:13:02 <georgk> otherwise, we at least have a good basis for the next release
13:13:20 <mardim> fuqiao, Yes exactly
13:13:30 <fuqiao> #info plan to get the 1st SDN HA testcase before deadline
13:13:36 <georgk> but do you need a baremetal deployment?
13:13:37 <mardim> georgk, If the testcase find problems in Jenkins
13:13:47 <mardim> we can merge Bug fixes after the merge right ?
13:14:19 <mardim> georgk, You do not need a baremetal
13:14:29 <mardim> But you need an installer
13:14:38 <mardim> with a POD that yardstick supports
13:14:46 <mardim> such as Fuel or compass
13:14:48 <fuqiao> Mardim: can the pharos lab help with such issue? I guess we can find some Pharos environment support ODL cluster?
13:15:12 <mardim> fuqiao, Maybe compass support clustering
13:15:25 <mardim> and Yardstick has a POD suport for compass
13:15:30 <mardim> let me paste the link
13:15:43 <fuqiao> OK. do you need me to reach out to the compass team and ask for help?
13:15:52 <mardim> #link https://github.com/opnfv/yardstick/tree/master/etc/yardstick/nodes
13:16:02 <mardim> pod support for the installers ^
13:16:51 <georgk> mardim: interesting, I thought Yardstick was a little bit more flexible
13:17:12 <mardim> georgk, I can not find a pod support for Apex
13:17:15 <mardim> so that means
13:17:25 <mardim> that if we need the HA ODL testcases
13:17:37 <mardim> to run on Apex we should create the POD yaml too
13:17:47 <mardim> this is my understanding
13:18:06 <fuqiao> #info the main difficulty for the 2nd test case of SDN HA is lack of testing environment. will have to figure out which installer support ODL cluster
13:18:58 <georgk> mhh, ok
13:19:12 <georgk> the pod.yaml is needed to find the nodes, right?
13:19:15 <fuqiao> Wangmeng: could you please also give an update for the VM HA and hypervisor HA testcase?
13:19:22 <mardim> georgk, exaclty
13:19:31 <mardim> the attack script for instance
13:19:33 <georgk> ok, folowing this up with you after the meeting
13:19:47 <mardim> needs to know in which node the ODL runs to ssh and kill the process
13:19:55 <mardim> georgk, sure !!
13:21:18 <fuqiao> Wangmeng? are you still here?
13:22:00 <wangmeng> I'm working on calculating the vm reset time, the others have been done.
13:23:04 <fuqiao> OK
13:23:07 <Kanglin> #info Kanglin
13:23:38 <fuqiao> do you think its possible to finish before March 16?
13:24:33 <wangmeng> yes,I think I can finish the two tcs before March 16.
13:24:51 <fuqiao> we may have to leave some time for the community to review the test case. so better the testcase be uploaded before mid next week
13:25:26 <wangmeng> I will finish it as soon as possible for the community to review.
13:25:38 <fuqiao> OK. Great
13:26:14 <fuqiao> # VM HA and Hypervisor HA test case can possibly meet the deadline
13:27:03 <fuqiao> Greg: I recall we had discussion with the masakari project for ansible integration. do you have any updates on that?
13:31:01 <fuqiao> #info VM HA and hypervisor HA test case can possibly meet the deadline
13:31:31 <fuqiao> Greg: are you still on line?
13:32:50 <fuqiao> OK never mind. I will ping Greg afterwards. If there are no other topics, I will close the meeting.
13:32:57 <fuqiao> Thank you for everyone!
13:33:10 <fuqiao> #endmeeting