08:31:22 <kubi001> #startmeeting Yardstick Work Meeting
08:31:22 <collabot`> Meeting started Mon Nov 14 08:31:22 2016 UTC.  The chair is kubi001. Information about MeetBot at http://wiki.debian.org/MeetBot.
08:31:22 <collabot`> Useful Commands: #action #agreed #help #info #idea #link #topic.
08:31:22 <collabot`> The meeting name has been set to 'yardstick_work_meeting'
08:31:37 <kubi001> #topic roll call
08:31:54 <kubi001> #info kubi
08:32:02 <JingLu> #info Jing
08:32:16 <JackChan> #info Jack
08:32:18 <zhihui_> #info zhihui
08:32:21 <rbbratta> #info Ross
08:32:43 <Kanglin> #info Kanglin
08:33:18 <kubi001> rbbratta: welcome to the meeting
08:33:33 <rbbratta> hi
08:33:54 <JingLu> welcome!
08:34:28 <JackChan> rbbratta:welcome~
08:34:49 <zhihui_> welcome:)
08:35:07 <Kanglin> welcome : )
08:35:08 <kubi001> rbbratta: could you give a briefly self-introduction to the team :)
08:35:29 <rbbratta> is gotomeeting started?
08:35:40 <kubi001> today is IRC only
08:35:53 <kubi001> rbbratta: today's meeting is IRC only
08:36:08 <rbbratta> Hi, I'm Ross from Intel, starting working on yardstick for some addition VNF testing we are working on.
08:36:10 <kubi001> rbbratta: we could start the GotoMeeting if you need
08:36:20 <rbbratta> no, IRC is okay
08:36:47 <kubi001> rbbratta: that's great
08:36:59 <rbbratta> I think someone did a demo last week
08:37:56 <rbbratta> I'm also signed up to go to the plugfest at UNH in December
08:38:05 <kubi001> rbbratta: great
08:38:19 <kubi001> rbbratta: Mingjiang also will be there
08:38:42 <Mingjiang6> i'll  be there
08:39:14 <kubi001> Mingjiang6:  rbbratta you could have a F2F meeting during the plugfest
08:39:34 <kubi001> zhihui_: would you go to the plugfest?
08:40:06 <kubi001> Due to some reason, I won't attend the plugfest
08:40:06 <rbbratta> sure, sounds good
08:40:39 <Mingjiang6> yes thats great
08:41:18 <rbbratta> I've just started with the code, trying to get it working in our lab.
08:41:32 <rbbratta> I've been having some ssh issues
08:41:57 <JingLu> can you share more details?
08:41:59 <rbbratta> but if you have agenda, I can wait
08:42:08 <zhihui_> kubi001, I will not attend plugfest.
08:42:21 <kubi001> rbbratta: IRC channel will be a good place to discuss the issue
08:42:28 <zhihui_> our company has not decided to attend plugfest.
08:42:54 <kubi001> zhihui_: got it, thanks
08:44:09 <rbbratta> I was trying to push a patch, but git-review broke on me.
08:44:26 <Mingjiang> what's you problem?
08:44:31 <rbbratta> We have a VNF that we send commands to periodically, so we need to keep stdin open
08:44:39 <kubi001> #info rbbratta and Mingjiang will attend the plugfest
08:44:50 <rbbratta> but ssh.py closes stdin if is .read() returns empty string
08:45:26 <rbbratta> https://gerrit.opnfv.org/gerrit/24311
08:46:53 <rbbratta> need to add unittest for that
08:46:55 <Mingjiang> rbbratta: i can't access to the link
08:46:58 <kubi001> rbbratta: It seems that I can't open it
08:47:20 <JingLu> me either
08:47:32 <rbbratta> https://gerrit.opnfv.org/gerrit/#/c/24311/
08:48:04 <JingLu> still cannot access
08:48:20 <kubi001> zhihui_: Kanglin could you access the link?
08:48:23 <rbbratta> hmm, I pushed it as a draft
08:48:57 <kubi001> It works
08:48:58 <rbbratta> should just be change 24311
08:49:02 <Kanglin> yes i could
08:49:08 <JingLu> the link works now
08:49:17 <zhihui_> kubi001, yes.
08:49:29 <rbbratta> maybe permissions for gerrit drafts are limited
08:49:53 <Mingjiang> it works now
08:50:52 <kubi001> rbbratta: so, you have fix the issue in that patch?
08:51:37 <OPNFV-Gerrit-Bot> Ross Brattain proposed yardstick: ssh.py: add flag to keep stdin open  https://gerrit.opnfv.org/gerrit/24311
08:52:40 <rbbratta> yes, with the patch it seems to keep stdin open.  I haven't done a full duration test, so stdin might close for some other reason, we still might need keepalives.
08:53:13 <rbbratta> I was trying to figure out if there was better way.
08:53:19 <kubi001> #action all, please help to review the patch https://gerrit.opnfv.org/gerrit/#/c/24311/
08:53:34 <kubi001> rbbratta: sounds good
08:54:14 <kubi001> #topic Yardstick support Newton
08:54:58 <kubi001> #info Yuenan Li send a email about Functest and Yardstick can't run on the Newton Openstack
08:55:41 <kubi001> #info due to some API version confliction. such as Keystone
08:55:59 <Mingjiang> I've noticed that Lu Jing have created a jira about this
08:56:36 <Mingjiang> do we have a deadline to fix this problem?
08:56:58 <JingLu> Yes The JIRA ticket is YARDSTICK-410
08:57:28 <JingLu> #link:https://jira.opnfv.org/browse/YARDSTICK-410
08:57:55 <kubi001> It would be great if we could make it work before the MS3
08:58:54 <kubi001> MS3 is end at 11/30
08:59:52 <kubi001> we may need to consider how to dynamically select version
09:00:05 <kubi001> zhihui_: any ideas?
09:00:57 <JackChan> Maybe we need to discuss with functest and other projects to come to a common conclusion
09:01:13 <kubi001> JackChan: that's a good idea
09:01:17 <Mingjiang> agree
09:01:21 <JingLu> agree
09:01:43 <kubi001> do we have some draft idea about this issue?
09:02:29 <zhihui_> I don't think we need to support all the openstack verison.
09:03:03 <JackChan> But maybe at least v2 and V3?
09:03:14 <zhihui_> Different branch already did it.
09:04:40 <zhihui_> we just need support the latest openstack cli in branch master
09:04:48 <Mingjiang> i think we should keep up with the newest openstack, since opnfv would switch to newer version each release.
09:06:11 <Mingjiang> for the previous version, we'd support it in a openstack way, which means we should use openstack api to be compatible, not by our own ways.
09:07:15 <JingLu> I think compatiability is import here
09:07:47 <JackChan> Mingjiang: Yes, making it compatible is necessary
09:07:55 <JingLu> import->important
09:10:23 <kubi001> zhihui_: do you mean we only need to support the latest Openstack CLi in lastest Yardstick Release?
09:10:50 <zhihui_> yes , If openstck verison is Mitaka, you can use yardstick with branch c
09:11:15 <JackChan> But I think since yardstick is a tools for testing, we can not choose the environment. My opinion.
09:11:16 <zhihui_> If openstack verison is Newton, you can use yardstick with branch master.
09:12:28 <kubi001> zhihui_: that would be much easier for us,  but that would be a problem If some user want to use new feature of yardstick at old version cloud
09:13:01 <zhihui_> yes, this is a problem.
09:13:30 <JackChan> zhihui: so the new yardstick version will not support M release?
09:14:45 <zhihui_> JackChan, it support M.
09:15:05 <zhihui_> I don't find the cli problem with fuel.
09:15:35 <kubi001> zhihui_: do you know if the fuel have update to Newton?
09:15:53 <zhihui_> I am not sure, I need to check.
09:15:59 <kubi001> zhihui_: It looks like it would be happened with Newton
09:16:15 <JackChan> zhihui: Sorry, I mesunderstand what you mean.
09:17:47 <zhihui_> compatibility is a long term work.
09:17:48 <JackChan> zhihui; But how to achieve?
09:18:19 <Mingjiang> i think we should try our best to be compatible since yardstick is a tool, not specific to a specific env. I suggest yardstick team to support Newton at a high priority and then try to be compatible in the long term.
09:18:26 <zhihui_> I am not against compatibility.
09:19:12 <zhihui_> Link: https://build.opnfv.org/ci/job/yardstick-fuel-zte-pod1-daily-master/138/consoleFull
09:19:56 <zhihui_> I remember the cli warning has existed a long time, seems from mitaka
09:21:09 <zhihui_> Maybe it is a block issue in Newton.
09:22:44 <zhihui_> If it is, so let's update openstack cli.
09:23:56 <JingLu> yes, the keynote cli will be deprecated in Newton
09:24:39 <JackChan> Yes, the openstack community info that they willl not support the old version in N release.
09:25:02 <kubi001> OK, It seems that we need to update
09:25:18 <JingLu> I think there is a milestone for achieving support for Newton.
09:25:32 <kubi001> JingLu: Yes, MS3
09:27:51 <kubi001> let's have more discussion with the email
09:27:52 <JackChan> So if you run under N release, it will failed due to keystone client version. Also maybe there some problem in heat template(not sure).
09:28:17 <kubi001> JackChan:thanks for pointing out this
09:28:55 <kubi001> #action we will have more discussion with email
09:29:04 <kubi001> #topic others
09:29:11 <kubi001> any others topic?
09:30:26 <kubi001> thanks all for today
09:30:39 <kubi001> #endmeeting