19:00:01 <NobodyCam> #startmeeting Ironic
19:00:02 <NobodyCam> #chair devananda
19:00:02 <openstack> Meeting started Mon Jan 13 19:00:01 2014 UTC and is due to finish in 60 minutes.  The chair is NobodyCam. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:00:02 <NobodyCam> Welcome everyone to the Ironic meeting.
19:00:03 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:00:05 <openstack> The meeting name has been set to 'ironic'
19:00:06 <openstack> Current chairs: NobodyCam devananda
19:00:24 <NobodyCam> Welcome back everyone
19:00:36 <NobodyCam> who's here for the ironic meeting
19:00:46 <lucasagomes> o/
19:00:48 <GheRivero> o/
19:00:51 <agordeev1> o/
19:00:52 * devananda waves from Australia
19:00:52 <ifarkas> o/
19:00:52 <matty_dubs> \o
19:01:04 <rloo> o/
19:01:17 <NobodyCam> welcome everyone
19:01:30 <k4n0> o/
19:01:34 <max_lobur> o/
19:01:44 <vaibhav> o/
19:01:53 <NobodyCam> #topic announcements
19:02:06 <NobodyCam> Icehouse release progress // graduation?
19:02:17 <NobodyCam> devananda: is that you?
19:02:36 <devananda> well. it's all of us :)
19:02:54 <NobodyCam> :-p
19:03:03 <devananda> tl;dr - the TC is going to be looking at all the incubated projects this week
19:03:10 <devananda> to guage progress towards graduation
19:03:21 <haomeng2> me:)
19:03:25 <devananda> particularly to compare everything to the newly-approved guidelines
19:03:41 <devananda> which, in case folks haven't seen them, are here
19:03:44 <devananda> #link http://git.openstack.org/cgit/openstack/governance/tree/reference/incubation-integration-requirements
19:03:52 <dkehn> devananda: do you hae a link to the new guidelines?
19:04:00 <dkehn> devananda: nevermind
19:04:02 <devananda> :)
19:04:39 <devananda> i also want to remind / inform folks taht we, as a project, need to start being better about how we communicate our plans
19:04:54 <devananda> which is done by targeting thigns (bugs, BPs, etc) towards releases and milestones
19:05:00 <devananda> you can see what's targeted here:
19:05:02 <devananda> #link https://launchpad.net/ironic/icehouse
19:05:03 <devananda> and
19:05:13 <devananda> #link https://launchpad.net/ironic/+milestone/icehouse-2
19:05:26 <devananda> anyone should be able to set the "target" for bugs
19:06:05 <devananda> with I2 coming up, I'd like to take a few minutes to review the targeted things, particularly BPs
19:06:10 <devananda> and see if we need to retarget anything
19:07:03 <devananda> #topic Blueprints for I2 milestone
19:07:16 <devananda> let's jsut go down the list
19:07:25 <devananda> dkehn: hi!
19:07:26 <NobodyCam> devananda: add-neutron-support is it really essential?
19:07:37 <dkehn> devananda: hi
19:07:49 <devananda> dkehn: you've got a patch up for neutron. is there much more to do?
19:07:56 <devananda> dkehn: or does it just need testing?
19:08:13 <dkehn> It just needs testing, as you can see
19:08:22 <devananda> NobodyCam: without neutron support, tripleo won't be able to use ironic. so, IMO, yes.
19:08:34 <devananda> dkehn: great. want to update the BP status?
19:08:35 <NobodyCam> devananda: ack
19:08:36 <dkehn> there is more tessting code that actual code, and I ran out of time to complete it all
19:09:02 <dkehn> I will be working on it but this week is the neutron sprint in Montreal
19:09:21 <dkehn> so there will be limited time to complete until I get back
19:09:27 <dkehn> assuming I get back
19:09:34 <dkehn> thats it
19:09:42 <devananda> dkehn: taht's fine. I'll set it to "needs testing" :)
19:09:49 <devananda> err, "needs code review", i mean
19:09:59 <dkehn> k
19:10:14 <devananda> PXE driver BP...
19:10:20 <devananda> has no one assigned.
19:10:30 <devananda> that one landed long ago, but we're all jsut now testing it
19:10:37 <dkehn> devananda: I'll get it out of WIP then
19:10:53 <devananda> anyone want to take point on that BP? GheRivero ? NobodyCam ?
19:11:05 <GheRivero> I can take it
19:11:16 <devananda> GheRivero: thanks
19:11:17 <NobodyCam> :0 faster then I
19:12:21 <devananda> romcheg: around?
19:12:30 <romcheg> devananda: yup
19:12:33 <NobodyCam> devananda: we'll need to get the "get service url" working for that
19:12:37 <devananda> sunjing: around?
19:13:00 <devananda> romcheg: I think the "intelligent routing for RPC" BP is superceded by my consistent hash BP. what do you think?
19:13:17 <haomeng2> jing is not around i think
19:13:22 <romcheg> I think we discussed that already :)
19:13:27 <romcheg> I agree with you
19:13:30 <devananda> romcheg: k :) I'll close that BP
19:13:35 <GheRivero> I will have something about "get service url" soon
19:13:46 <NobodyCam> GheRivero: W00t
19:13:47 <romcheg> devananda: thanks
19:13:49 <lucasagomes> :)
19:13:57 <lucasagomes> I'm also fixing the ironic-rootwrap
19:14:00 <lucasagomes> couple of problems with it
19:14:06 <lucasagomes> https://bugs.launchpad.net/ironic/+bug/1268696
19:14:06 <devananda> GheRivero: great! I saw some discussion in channel about it
19:14:09 <uvirtbot> Launchpad bug 1268696 in ironic "On Ubuntu: /usr/local/bin/ironic-rootwrap: Executable not found: /sbin/iscsiadm" [Undecided,In progress]
19:14:10 <lucasagomes> https://bugs.launchpad.net/ironic/+bug/1268674
19:14:13 <uvirtbot> Launchpad bug 1268674 in ironic "ironic-rootwrap: No section: 'Filters'" [Undecided,In progress]
19:14:26 <lucasagomes> that's needed for pxe
19:14:46 <NobodyCam> lucasagomes: was that the reviews you just put up?
19:14:54 <lucasagomes> NobodyCam, yes
19:15:04 <NobodyCam> :) will look at after meeting
19:15:05 <lucasagomes> there's one more needed but on devstack
19:15:06 <devananda> lucasagomes: ahh. IIRC, that has a different path
19:15:19 <rloo> lucasagomes: don't forget to specify the milestone on those bugs :-)
19:15:29 <lucasagomes> rloo, +1
19:15:41 <haomeng2> :)
19:16:21 <linggao> devananda, I can try to answer your questions for sun jing for relate them to her.
19:16:28 <lucasagomes> devananda, yup it does... but as nova also does not use absolute paths on the filters
19:16:37 <devananda> linggao: hi!
19:16:37 <lucasagomes> I think we might want to go with the same approach
19:16:45 <linggao> hi
19:17:04 <devananda> linggao: how's progress on the "use ipmitool for serial console" work? I've seen a patch up but haven't had time to test it yet
19:17:14 <devananda> linggao: is that patch a complete implementation, or is there mroe to do?
19:18:13 <devananda> lucasagomes: could you also set importance for your bugs?
19:18:18 <linggao> devananda, from the conversation I had with her last wek, she has put the implementation for serial console using ipmitool in the patch.
19:18:30 <lucasagomes> devananda, +1 yes
19:19:16 <devananda> lucasagomes: i trust that you can guage the importance of a bug -- and really, most of us should be able to triage bugs now :)
19:19:26 <devananda> linggao: great!
19:19:31 <devananda> linggao: i'll update teh BP
19:19:32 <linggao> devanandan, she was going to change it to use the Console interface to start or stop the condole instead of puting them in the power on/off funcitions like nova did
19:19:44 <linggao> but I am not sure they check that code in or not.
19:19:50 <devananda> linggao: never mind -- nothing to update. it already has the right status :)
19:20:08 <devananda> linggao: ah. it should indeed use the console interface now
19:20:17 <devananda> sounds like it may not be complete by I2 then
19:20:22 <lucasagomes> devananda, right :) yea I need to take a better look at our bug list. I fixed the ones assigned to me awhile ago.
19:20:46 <linggao> the to-do for that BP is to implement the console interface usng the ipminative, that will be done after icehouse
19:21:15 <devananda> linggao: that's fine
19:21:32 <devananda> ok, i'd like to move on, so we dont spend the whole meeting on BP's
19:21:37 <devananda> thanks everyone :)
19:21:40 <NobodyCam> devananda: do the seamicro bp need to me tagged as i2
19:21:45 <NobodyCam> ok
19:22:06 <devananda> NobodyCam: let's talk about the seamicro BPs in a bit
19:22:11 <NobodyCam> ack
19:22:12 <devananda> they were jsut proposed a few days ago
19:22:26 <devananda> #topic Tempest API tests
19:22:31 <haomeng2> devananda, i will try to triage some bugs
19:22:42 <devananda> I want to thank romcheg for all the work he did on this
19:22:45 <romcheg> I have exciting news about the test
19:23:04 <romcheg> But everyone already knows them :)
19:23:13 <max_lobur> :)
19:23:17 <NobodyCam> :)
19:23:21 <devananda> romcheg: hehe, but please, say it anyway :)
19:23:26 <romcheg> ok
19:23:30 <haomeng2> great ;)
19:23:38 <romcheg> So everything got merged
19:24:00 <NobodyCam> awesome romcheg
19:24:06 <lucasagomes> good stuff :)
19:24:09 <romcheg> Tempest tests can be manually triggered for every ironic patch by leaving "check experimental" on review
19:24:09 <max_lobur> congrats!
19:24:20 <devananda> also, we will soon have tempest API checks in our gate, once this lands
19:24:23 <devananda> #link https://review.openstack.org/#/c/65845/
19:24:40 <devananda> so every patch to Ironic will spin up devstack. expect jenkins to become slower ...
19:24:52 <NobodyCam> ieek
19:24:59 <devananda> the same tests will also run for every tempest and devstack patch -- but in non-voting mode
19:25:34 <haomeng2> inclouding pxe_ssh driver?
19:25:49 <devananda> haomeng2: right now, this is only for exercising our API
19:26:02 <haomeng2> ok got
19:26:05 <devananda> it's an integration test -- make sure all the components of Ironic integrate with each other
19:26:06 <romcheg> haomeng2: agordeev1 is working on that
19:26:17 <haomeng2> ok
19:26:24 <devananda> functional test (can it actually DO a deployment) is being worked on still
19:26:31 <romcheg> I'm trying to help them since I got more familiar with infra
19:26:36 <devananda> i'm hoping we haev that by I3 milestone
19:27:05 <devananda> #topic Nova driver
19:27:25 <devananda> NobodyCam: that's you - any announcements? besides how awesome it is? :)
19:27:33 <NobodyCam> :) I need to actually start writing tests
19:27:39 <NobodyCam> lol
19:27:42 <devananda> heh
19:27:44 <NobodyCam> its getting there
19:28:10 <devananda> i had a brief chat with a few nova folks at LCA -- as I thought, we'll need to split the patch up for it to land
19:28:19 <NobodyCam> still 100% but I think we worked out a way to map the drivers requirments so nova will know what each driver needs
19:28:29 <NobodyCam> oh ya
19:28:32 <devananda> nice
19:28:41 <lucasagomes> nice
19:28:51 <NobodyCam> the full review is up at
19:28:57 <lucasagomes> would be that something that we might want our drivers to say as well?
19:28:58 <NobodyCam> looks for link
19:29:15 <NobodyCam> #link https://review.openstack.org/#/c/51328/
19:29:18 <devananda> lucasagomes: you mean in our /drivers/ api?
19:29:26 <lucasagomes> for example, by GET /drivers, would also return a list of the mandatory parameters
19:29:32 <lucasagomes> with an description etc
19:29:34 <lucasagomes> devananda, yes
19:29:35 <devananda> lucasagomes: i think we talked about that early on. Yea, eventually I think they should
19:30:04 <NobodyCam> lucasagomes: yes I would like to make that dynamic
19:30:07 <lucasagomes> ah right, yea that would be very useful
19:30:29 <devananda> lucasagomes: so that other clients can retrieve said list, then populate driver_info, then call validate to see if they got everything
19:30:37 <NobodyCam> but the good news is we are acryally calling deploy from the driver
19:30:55 <devananda> lucasagomes: but nova will still need some static mapping there, I expect
19:31:41 <NobodyCam> I am would like feedback on the mapping and anyother bit of the driver
19:32:01 <devananda> NobodyCam: one thing on building that review into a dib image that I learned last week
19:32:21 <devananda> NobodyCam: you can tell dib to pull in a patch/rev via env vars now
19:32:23 <devananda> just set this
19:32:30 <devananda> export DIB_REPOLOCATION_nova=https://review.openstack.org/openstack/nova
19:32:33 <devananda> export DIB_REPOREF_nova=refs/changes/28/51328/8
19:32:36 <lucasagomes> right, maybe having a bug ticket to implement that ( wishlist ) would be great, I can file that later
19:32:40 <NobodyCam> awesome
19:32:43 <devananda> instead of hard-coding it in your element install script
19:32:51 <devananda> lucasagomes: ++
19:33:54 <NobodyCam> devananda: I can look at refactoring the element i have up in my repo
19:34:19 <NobodyCam> or I saw you may have started a new one
19:35:32 <NobodyCam> devananda: just dropped
19:36:01 <NobodyCam> are there anyquestions on the nova driver
19:36:11 * devananda switches to cellular
19:36:45 <devananda> #topic functional testing
19:36:58 <devananda> just a quick note here - i'll be talking with agordeev1 later today about this
19:37:06 <devananda> and romcheg hopefully, too
19:37:18 <romcheg> +1
19:37:39 <devananda> functional testing is a requirement for graduation, and we should have it in our gate ASAP. certainly no later than I3
19:37:55 <devananda> #topic tripleo integration
19:38:03 <devananda> again, just a quick note -
19:38:24 <devananda> i think we can simplify the current walkthrough in our wiki
19:38:47 <devananda> and merge most of that up into tripleo at this point
19:39:03 <devananda> in fact, we will need to at some point have tripleo support both baremetal and ironic in parallel
19:39:08 <devananda> with an ENV var to toggle it
19:39:14 <devananda> so that folks using tripleo can migrate
19:39:43 <devananda> NobodyCam: you've been maintaining that mostly, so once i'm back in the states, let's hash those changes out
19:39:55 <NobodyCam> devananda: I have had a hard time in the past getting env var thru the oOo revirew processs
19:40:07 <NobodyCam> devananda: ack
19:40:16 <devananda> with 20 minutes left...
19:40:24 <devananda> #topic SeaMicro proposals
19:40:42 <k4n0> Hi folks
19:40:47 <k4n0> Rohan here
19:40:49 <devananda> hi!
19:40:54 <romcheg> Hi!
19:40:55 <NobodyCam> hi rohan
19:41:19 <k4n0> Seamicro is proposing ironic drivers for its products
19:41:21 <NobodyCam> k4n0: do you think you can have all that by i2
19:41:30 <k4n0> yes
19:41:44 <k4n0> working very hard on that
19:42:32 <devananda> k4n0: i just saw your reply on the BP regarding the seamicro client lib
19:42:56 <k4n0> devananda: yes, does that sound ok?
19:43:00 <devananda> k4n0: good stuff! I'm thrilled to see support for seamicro getting added
19:43:07 <devananda> k4n0: yep, sounds fine
19:43:13 <lucasagomes> # link https://blueprints.launchpad.net/ironic/+spec/seamicro-power-driver
19:43:35 <k4n0> we will need talk about functional testing of drivers though
19:43:43 <devananda> right
19:43:52 <NobodyCam> also:
19:43:56 <NobodyCam> #link https://blueprints.launchpad.net/ironic/+spec/seamicro-vendor-passthru-interface-implementation
19:44:18 <devananda> k4n0: we (will) have the same requirement that Nova does -- all third-party drivers need functional testing on real hardware
19:44:26 <romcheg> Regarding to testing: I think we will have to chose the neutron's ways
19:44:35 <devananda> k4n0: tripleo has been working on some specifications for how folks contribute it
19:44:46 * devananda looks for the link
19:44:52 <devananda> romcheg: what do you mean exactly?
19:45:00 <romcheg> When vendors set up their own environment which is open for OpenStack CI
19:45:09 <devananda> neutron has had a lot of difficulties in the past, but working on improving it...
19:45:28 <romcheg> Then we'll be able to run TripleO's periodic tasks againts that hardware
19:46:04 <NobodyCam> devananda: are the requirments for the hardware venders need to put up?
19:46:16 <k4n0> that would be helpful
19:47:46 <devananda> NobodyCam: i haven't written them up specifically yet (on my list...)
19:48:06 <devananda> but lifeless has written it up for tripleo, and I'd like to point fokls to that for the time being
19:48:17 <k4n0> will do
19:48:50 <devananda> k4n0: broadly speaking, there are two kidns of testing which jenkins can support
19:48:53 <devananda> voting and non-voting
19:49:14 <devananda> infra has very strict requirements on things which can vote in the gate, since that will affect ALL projects
19:49:48 <k4n0> right, so where do the third party functional tests go?
19:49:49 <devananda> we're still discussing exactly where in the matrix of possible tests vendor drivers need to fit for ironic
19:49:53 <devananda> right :)
19:49:57 <k4n0> :)
19:49:58 <k4n0> ok
19:50:05 <devananda> my current view is that it's the intersection of
19:50:10 <devananda> *somewhere near the
19:50:13 <k4n0> something like turbo-hipster?
19:50:30 <devananda> check + gate for all patches in Ironic
19:51:06 <romcheg> devananda: wouldn't that be too expensive for vendors?
19:51:31 <devananda> romcheg: how else can we know taht some patch doesn't break support for that vendor?
19:51:54 <devananda> romcheg: how else do we prevent regressions and ensure support for the things which are landed in trunk?
19:52:17 <romcheg> devananda: there's no other way I think :(
19:52:30 <devananda> k4n0: what do you think?
19:52:47 <devananda> k4n0: will you guys be able to test, at a minimum, all patches to ironic (both check and gate pipelines)
19:53:08 <k4n0> I think so, will need to check on this
19:53:16 <devananda> ignoring for the moment that we're still working out bugs in our deploy process, and dont have functional tests written into devstack :)
19:53:32 <k4n0> Does neutron do the same?
19:53:41 <devananda> k4n0: neutron is moving towards this model
19:53:54 <devananda> k4n0: and nova is kicking out all the drivers which aren't doing this
19:54:02 <k4n0> :) ok, i can talk to my folks, get back to you on this
19:54:19 <devananda> thanks :)
19:54:31 <devananda> k4n0: any other questions for us?
19:55:14 <k4n0> we can get the blueprints in after i2 right?
19:55:29 <k4n0> i mean, do we have to wait till i3?
19:55:37 <NobodyCam> i was thinking maybe we should tag as i3 for now
19:56:12 <devananda> k4n0: targeting a BP is just a way of communicating expectations
19:56:14 <NobodyCam> they can ofc land sooner
19:56:26 <k4n0> ok, we will try to finish sooner
19:56:37 <k4n0> i2 is good for now
19:57:01 <devananda> k4n0: i'm fine with it targetted to i3
19:57:19 <k4n0> ok
19:57:40 <devananda> k4n0: oh! also, here's the link for tripleo's requirements for additional regions -- not goign to be exactly the same as ours, but may be worth reading, particularly if you gys are also looking at tripleo support
19:57:44 <devananda> #link https://wiki.openstack.org/wiki/TripleO/TripleOCloud/Regions
19:58:00 <k4n0> devananda: thanks !
19:58:11 <k4n0> I am done with my questions
19:58:24 <NobodyCam> and we have two minutes
19:58:31 <devananda> :)
19:58:34 <k4n0> thank you everyone, I will try to hang around every meeting :)
19:58:35 <devananda> #topic open discussion
19:58:52 <NobodyCam> k4n0: great also hang out in channel too
19:58:55 <devananda> k4n0: please join us in #openstack-ironic too
19:59:03 <max_lobur> as for race condition problem
19:59:05 <k4n0> ohh cool, didnt know that channel
19:59:09 <max_lobur> there are 3 links in agenda
19:59:13 <devananda> k4n0: and thanks again -- I'm really thrilled to see SeaMicro contributing :)
19:59:14 <lucasagomes> #link https://bugs.launchpad.net/ironic/+bug/1259910
19:59:15 <uvirtbot> Launchpad bug 1259910 in ironic "race condition when changing node states" [Medium,In progress]
19:59:22 <NobodyCam> devananda: +1
19:59:33 <lucasagomes> I don't think we have time to talk about the race condition (1min left)
19:59:34 <NobodyCam> max_lobur: you put up a
19:59:37 <NobodyCam> poc today?
19:59:39 <NobodyCam> on that
19:59:50 <lucasagomes> devananda, could you take a look at that bug ticket when u get some free time?
19:59:55 <devananda> lucasagomes: will do
20:00:02 <lucasagomes> NobodyCam, he did already
20:00:11 <lucasagomes> there's one patch fixing it using the thread approach
20:00:13 <NobodyCam> hae the link for the log
20:00:27 <max_lobur> :)
20:00:27 <devananda> lucasagomes: let's discuss the bug back in channel though
20:00:34 <lucasagomes> #link https://review.openstack.org/#/c/66368
20:00:35 <max_lobur> +
20:00:38 <NobodyCam> great meeting every one
20:00:41 <devananda> thanks everyone!
20:00:44 <NobodyCam> ty lucasagomes
20:00:45 <k4n0> thanks
20:00:47 <haomeng2> :)
20:00:47 <max_lobur> thanks!
20:00:50 <lucasagomes> thanks
20:00:52 <haomeng2> tks
20:00:55 <romcheg> thanks
20:01:00 <devananda> #endmeeting