18:00:25 <tenobreg> #startmeeting Sahara
18:00:26 <openstack> Meeting started Thu Jun  8 18:00:25 2017 UTC and is due to finish in 60 minutes.  The chair is tenobreg. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:27 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:00:30 <openstack> The meeting name has been set to 'sahara'
18:00:54 <vgridnev> o/
18:02:33 <tenobreg> lets wait a couple more minutes to see if anyone else shows up, otherwise start just the three of us
18:03:16 <jeremyfreudberg> tenobreg, three of us = me, you, __?
18:03:31 <tenobreg> and vgridnev
18:03:39 <jeremyfreudberg> right
18:04:25 <jeremyfreudberg> should we ping our outreachy intern?
18:05:08 <tenobreg> jeremyfreudberg, just did
18:05:15 <jeremyfreudberg> yup
18:05:16 <tenobreg> ok, lets start
18:05:24 <tenobreg> #topic News/Updates
18:06:01 <tenobreg> I'm still working on image creation and I have a question about that to ask vgridnev later. Also, I will cut Pike-2 today, since we got gate fixed
18:06:09 <vgridnev> so I've been working to fix several jobs: spark dib, added new scenario job with py3 (experimental) , and grenade, but it will not pass because heat can't delete floating ip ...
18:06:27 <jeremyfreudberg> vgridnev, where's the log for that?
18:06:40 <vgridnev> https://review.openstack.org/#/c/471371/
18:07:17 <tenobreg> gate-grenade-dsvm-sahara-ubuntu-xenial-nv right?
18:07:31 <vgridnev> yep
18:07:45 <jeremyfreudberg> i'll take a look
18:07:48 <tenobreg> I will take a look on the logs as well
18:08:12 <vgridnev> http://logs.openstack.org/71/471371/2/check/gate-grenade-dsvm-sahara-ubuntu-xenial-nv/0587927/logs/new/screen-h-eng.txt.gz
18:09:05 <tenobreg> thanks vgridnev
18:09:11 <jeremyfreudberg> rebase the changes maybe, I'm seeing a nova floating ip association there
18:09:50 <vgridnev> that is heat logs
18:10:30 <jeremyfreudberg> 2017-06-08 10:35:35.899 16016 INFO heat.engine.resource [req-d7f19a00-521b-4c83-a259-7e98efca16d2 - sahara_grenade - - -] DELETE: NovaFloatingIpAssociation "floating_ip_assoc"
18:10:47 <vgridnev> jeremyfreudberg, these logs are from heat, not from sahara
18:10:59 <tenobreg> looks to me like a client version problem, it tries to call self.client().floatingips but that doesn't exists anymore
18:11:03 <jeremyfreudberg> vgridnev, i understand, that kind of heat resource is broken
18:11:52 <jeremyfreudberg> vgridnev, i think a rebase would fix it, your parent commit is a month old
18:12:29 <vgridnev> jeremyfreudberg, nope
18:13:07 <tenobreg> does it make sense to be nova client version issue?
18:13:29 <jeremyfreudberg> yes, tenobreg, heat had the same problem with the new novaclient is us
18:13:41 <jeremyfreudberg> so they deprecated Nova floating IP associations in favor of Neutron ones
18:13:53 <jeremyfreudberg> anyway, vgridnev has just rebased the patch so we will see
18:14:08 <tenobreg> jeremyfreudberg, so you are saying that this might be old heat version we are using
18:14:58 <tenobreg> ?
18:15:16 <jeremyfreudberg> tenobreg, i'm not sure actually, because even though grenade failed with that error, our other jobs passed
18:15:34 <tenobreg> I see
18:15:36 <tenobreg> lets wait and see
18:16:34 <tenobreg> #topic Pike-2
18:16:55 <tenobreg> just want to make sure that we are ready to cut Pike-2
18:17:00 <vgridnev> jeremyfreudberg, job use the following scenario : we are creating cluster using Sahara from ocata + Heat from ocata. After that we update Sahara and heat to master and then we delete cluster
18:17:16 <vgridnev> tenobreg, ok
18:18:04 <jeremyfreudberg> ah, vgridnev, I understand now... An Ocata cluster would create a Nova floating IP resource, but a master-branch Cluster can't delete it, since a master-branch Heat pulls down the newest novaclient
18:18:09 <jeremyfreudberg> so the job will never pass
18:18:12 <jeremyfreudberg> for ocata->pike
18:18:17 <jeremyfreudberg> it will pass pike->queens
18:18:22 <jeremyfreudberg> (when that happens)
18:18:34 <vgridnev> Something like this, yep
18:18:35 <tenobreg> jeremyfreudberg, makes perfect sense
18:18:39 <tenobreg> do we have any thing else to do to cut Pike-2? From what I see we can cut it, gate is fixed and we have it stable with a reno about the breakage of the ironic
18:18:48 <vgridnev> I think no
18:18:48 <tenobreg> do we all agree?
18:19:08 <vgridnev> Breakage of ironic?
18:19:10 <vgridnev> What?
18:19:39 <tenobreg> vgridnev, this is somewhat of a doubt we have... in order to fix gate jeremyfreudberg had to use pre created port on the Heat template
18:19:46 <jeremyfreudberg> vgridnev, this is about using pre-created Neutron ports in our heat templates, we had to do that to get a stable gate
18:19:48 <tenobreg> but that used to be an issue with ironic
18:19:55 <vgridnev> https://bugs.launchpad.net/heat/+bug/1544195
18:19:56 <openstack> Launchpad bug 1544195 in heat "User can not provision ironic node via nova when providing pre-created port" [Undecided,Confirmed]
18:20:19 <vgridnev> Nova guys state that it might be fixed already
18:20:25 <jeremyfreudberg> exactly, it "might" be
18:20:33 <jeremyfreudberg> and we haven't tested it, so we are just putting a warning in release note
18:20:35 <tenobreg> we are not sure that if it still the case
18:20:41 <tenobreg> but hopefully it won't be
18:21:00 <tenobreg> we set a release note stating that and we will make sure it works by P-3
18:21:18 <vgridnev> tenobreg, you don't need such release note
18:21:39 <vgridnev> Just fix it if it's broken
18:21:57 <tenobreg> even if we are releasing a version with it broken?
18:22:02 <tenobreg> and we don't have time to fix?
18:22:20 <vgridnev> yes
18:22:24 <tenobreg> P-2 deadline is today
18:22:30 <jeremyfreudberg> vgridnev, i'm not sure, reno has a special section to mention these kinds of things, I think it's a nice thing to do
18:22:47 <jeremyfreudberg> and we don't call it broken, we just warn that it is untested
18:22:48 <vgridnev> Just file the bug
18:23:04 <tenobreg> I agree with jeremyfreudberg, we may not need it, but it is good to have that written
18:23:57 <tenobreg> jeremyfreudberg, can you file a bug?
18:25:12 <tenobreg> so, we all agree we can cut Pike-2?
18:25:28 <vgridnev> yes
18:25:30 <jeremyfreudberg> yes, i think pike-2 is fine, unless you want to me to remove release note
18:25:40 <tenobreg> jeremyfreudberg, lets leave it there
18:25:42 <tenobreg> it won't hurt
18:25:49 <tenobreg> once we fix it we can remove it
18:26:13 <tenobreg> #topic Open Discussion
18:26:38 <tenobreg> vgridnev, I have a question for you and maybe jeremyfreudberg as well... I'm working on image validation for the new image creation process
18:26:57 <tenobreg> the validation process gets some scripts and runs on the instances
18:27:33 <tenobreg> the problem I'm facing is the following: I export some variables distro and test_only(former reconcile0
18:28:21 <tenobreg> but the variables aren't set during script execution
18:29:05 <tenobreg> does paramiko has issues with export during its execution? is there a better way to do this?
18:29:52 <vgridnev> I'm not an expert in the new mechanism of image generation, but it seems that there is some issue in our image generation code, doesn't it?
18:30:14 <vgridnev> Variables on the node?
18:30:28 <vgridnev> Node created with new image, I mean
18:30:46 <tenobreg> what do you mean? The image creation is working properly
18:31:09 <tenobreg> at least I tried with CDH here and the image created was working
18:31:36 <vgridnev> So, I mean, your created the instance with new image. And then you test that instance with our scripts right?
18:31:38 <jeremyfreudberg> the scripts generally work as far as i know,  there's the experimental gate which seems to be doing something
18:32:14 <tenobreg> the validation process uses the same scripts that we use for image creation
18:32:28 <tenobreg> it just makes sure that everything that should be there is there already
18:32:48 <vgridnev> no idea then
18:32:52 <jeremyfreudberg> tenobreg, is this code that you have issues with somewhere
18:33:40 <tenobreg> the problem here is basically with running a script remotely
18:34:25 <jeremyfreudberg> tenobreg, nothing to do with this bug, right? https://review.openstack.org/#/c/446017/
18:34:40 <jeremyfreudberg> it sounds similar but i think yours is a different context
18:34:57 <tenobreg> jeremyfreudberg, not really
18:35:42 <tenobreg> it is similar, yes, we had the same issue during image creation because the "bash <<_SIV_" creates a new bash
18:35:43 <vgridnev> Check which shell you script is using
18:36:10 <vgridnev> It might just sh where you vars are not set
18:36:29 <tenobreg> it uses /bin/bash
18:37:03 <jeremyfreudberg> where are you "exporting" these variables from?
18:37:14 <tenobreg> let me paste here
18:37:30 <vgridnev> You also might be need to put these vars into some script in /etc/profile.d/ (how we do that with java)
18:37:46 <vgridnev> https://github.com/openstack/sahara/blob/8b4af0daf3688dc79c8b15818dd817b01f5cc823/sahara/swift/swift_helper.py#L91
18:38:06 <tenobreg> https://github.com/openstack/sahara/blob/master/sahara/plugins/images.py#L674-L681
18:38:56 <tenobreg> vgridnev, I see, I will take a look and try to figure out why this is happening
18:39:04 <vgridnev> ookay
18:39:50 <tenobreg> so jeremyfreudberg, on that part of the code, we concatenate the script with the export of the variables
18:40:27 <tenobreg> whole cmd is executed remotely
18:42:17 <jeremyfreudberg> not sure really, it seems like it's fine
18:42:33 <tenobreg> it does, but it is driving me crazy
18:42:34 <tenobreg> lol
18:43:12 <jeremyfreudberg> i assume you've thrown in some debug print statement already that prints script variable right before execution?
18:43:26 <tenobreg> yes, as blank as it can be
18:44:08 <vgridnev> So, I've got some sad news: next Tuesday is my last working day with Mirantis, btw I will try to stay around Sahara project, so you can always ping me by mail ( gridnevvvit at gmail )
18:44:35 <jeremyfreudberg> vgridnev :(
18:45:04 <tenobreg> vgridnev, sad to see you go, hopefully we will not lose you completely
18:45:36 <tenobreg> good luck on new endeavors
18:45:53 <jeremyfreudberg> best of luck
18:45:54 <vgridnev> But as I said, I will continue attending meetings sometimes if it would be possible. And will do code reviews!
18:46:03 <vgridnev> tenobreg, jeremyfreudberg thanks!
18:46:06 <tenobreg> vgridnev, we very much appreciate it
18:47:08 <tenobreg> do we have any other topics to discuss today?
18:47:15 <jeremyfreudberg> well, ptg
18:47:31 <jeremyfreudberg> i assume we'll just have the one day (and not two or three)
18:47:34 <jeremyfreudberg> if anything
18:48:03 <tenobreg> true
18:48:21 <tenobreg> We could get it all on one day for sure
18:48:29 <jeremyfreudberg> i guess when enough people are back, we will find out if enough saharans will come
18:48:38 <jeremyfreudberg> if it's just 2 people then not much point
18:50:12 <tellesnobrega> sorry, internet went down
18:50:19 <jeremyfreudberg> yeah, you've changed your name
18:50:32 <tellesnobrega> jeremyfreudberg, I don't think skipping is the right thing to do
18:50:33 <jeremyfreudberg> make sure the meeting ends properly, i've seen issue there before
18:51:10 <vgridnev> I will clarify possibility of my attendance later. BTW, there are 3 open questions about that: I will need new visa, Travel support program, and new company - not sure they will allow me to attend.
18:51:45 <tellesnobrega> vgridnev, completely understand, hopefully we will see you there
18:52:03 <jeremyfreudberg> vgridnev, hope to see you there too.
18:52:17 <jeremyfreudberg> hopefully we start having ptg in an "easier" country too
18:52:24 <tellesnobrega> I think we should have, even if it is only two or three or four people, it is nice to get some face to face time and we could organize some coding, bug fix day, bug triage and so on
18:52:47 <tellesnobrega> jeremyfreudberg, I think next year it will be in europe, lets see
18:53:05 <tenobreg> changed to make sure I can end the meeting
18:53:32 <tenobreg> jeremyfreudberg, how about we think on two days and organize some event for us, bug, documentation, or something like that
18:53:59 <tenobreg> documentation seems like an interesting thing and easily fixable without lots of resources
18:54:17 <jeremyfreudberg> i agree, docs would be good, and if we are at ptg we could talk to docs people
18:54:33 <tenobreg> true
18:55:47 <jeremyfreudberg> got to go
18:55:52 <tenobreg> thanks guys
18:55:58 <tenobreg> see you all next week
18:56:02 <tenobreg> best of luck vgridnev
18:56:17 <vgridnev> thanks
18:56:20 <tenobreg> #endmeeting