18:00:22 <ryanpetrello> #startmeeting astara
18:00:23 <openstack> Meeting started Mon May 23 18:00:22 2016 UTC and is due to finish in 60 minutes.  The chair is ryanpetrello. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:24 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:00:25 <markmcclain> o/
18:00:27 <openstack> The meeting name has been set to 'astara'
18:00:32 <ryanpetrello> o/ markmcclain
18:00:38 <ryanpetrello> adam_g rods eric_lopez
18:01:03 <rods> o/
18:01:33 <adam_g> o/
18:05:55 <ryanpetrello> so what's up this week?
18:06:02 <ryanpetrello> adam_g fixed the oslo messaging thing :)
18:06:06 <ryanpetrello> thanks for digging into that
18:06:15 <eric_lopez> sorry - little late
18:06:59 <ryanpetrello> https://review.openstack.org/#/c/319443/
18:07:11 <ryanpetrello> adam_g, does this affect older versions of astara?
18:07:16 <ryanpetrello> i.e., will we need to backport anything?
18:07:22 <adam_g> ryanpetrello: it shouldn't, one sec let me check global-requirements
18:08:21 <ryanpetrello> https://review.openstack.org/#/c/317696/ got backported to kilo and liberty, which fixes an issue we saw last week
18:08:39 <markmcclain> upper constraings has 4.6.1 which should be ok
18:08:45 <adam_g> yea
18:08:51 <adam_g> so
18:08:53 <adam_g> it should be fine in the gate
18:09:14 <markmcclain> we still have this easy one hanging around
18:09:16 <markmcclain> #link : https://review.openstack.org/#/c/304927/
18:09:16 <adam_g> but the actually requirements.txt has uncapped oslo.messaging, so anyone installing from stable branches without using constraints will have issues
18:09:33 <markmcclain> yeah... wonder if we should propose cap?
18:09:36 <adam_g> we could backport the messaging fix, i dont think we'll be able to cap them
18:10:04 <markmcclain> except we can't backport because of dep on oslo.service
18:10:46 <adam_g> hmmm
18:11:05 <adam_g> markmcclain: you mean b/c we introduce a new requirement?
18:11:14 <markmcclain> adam_g: yeah
18:11:51 <adam_g> ugh
18:12:47 <adam_g> one option would be to come up with a modified backport
18:12:51 <adam_g> that doesn't use oslo.service
18:13:08 <adam_g> or, block the master patch and modify it to drop the oslo.service usage
18:13:11 <markmcclain> yeah I think that might be doable
18:13:19 <adam_g> backport that, then update master to use oslo.service (if we wanted it)
18:13:50 <markmcclain> I was also going to stand up an isolated env and make sure oslo.service is not installed by our current dependency set
18:14:26 <adam_g> one sec
18:14:30 <adam_g> i can do that right quick
18:14:32 <markmcclain> ok.. so we might be in luck
18:14:45 <markmcclain> the current stable/mitaka installs oslo.service
18:14:54 <adam_g> ryanpetrello: what do you think about putting the brakes on merging 319443 ?
18:15:08 <ryanpetrello> wfm if we want to push pause
18:15:24 <markmcclain> so we might be able to backport the change by first fixing the error in the undeclared dep in stable
18:15:30 <markmcclain> and then backporting the actual change
18:15:53 <markmcclain> that would keep the history of why we did things clear.. thtoughts?
18:15:55 <adam_g> markmcclain: what do you mean, fixing undeclared dep?
18:16:48 <markmcclain> When installing stable/mitaka in fresh env here's the current packages
18:16:51 <markmcclain> #link http://paste.openstack.org/show/498240/
18:17:16 <markmcclain> we're implicitly relying on it (guessing a drive-by dep on some other oslo lib)
18:17:34 <markmcclain> so we could explicitly declare it as a requirement for stable and then backport the fix
18:17:52 <markmcclain> it wouldn't break anyone because it's already being installed
18:18:00 <adam_g> yea
18:18:19 <adam_g> WFM
18:18:31 <adam_g> ill add +W back to the original fix then
18:18:46 * ryanpetrello thumbs up
18:19:05 <ryanpetrello> is this an issue pre-mitaka
18:19:21 <adam_g> ryanpetrello: only when using oslo.messaging >5.x
18:19:48 <ryanpetrello> k
18:19:52 <adam_g> ryanpetrello: mitaka + liberty's requirements dont declare an upper version constraint on oslo.messaging, so pip installing it currently would get you the newer 5.x oslo.msg that would break you
18:20:02 <ryanpetrello> right
18:20:23 <adam_g> BUT the upper-constraints.txt pins it to 4.x in stable/mitaka, so it gives us the illusion everything is working
18:20:29 <adam_g> ... in the gate
18:20:32 <adam_g> *sigh*
18:21:32 <adam_g> so after that messaging fix is landed, https://review.openstack.org/#/c/317744/ should fix the appliance image builds in the -src jobs
18:21:37 <adam_g> and our gate should be happy across the board
18:21:51 <adam_g> ill keep an eye and recheck that once the other merges
18:21:55 * ryanpetrello dance
18:22:02 <ryanpetrello> awesome
18:22:06 <ryanpetrello> thanks for this, adam_g :D
18:22:55 <adam_g> no prob
18:24:01 <ryanpetrello> any other updates from anybody?
18:24:55 <adam_g> not much from me, those breakages occupied most of my time last week
18:25:17 <markmcclain> just to random fixups I mentioned earlier
18:25:46 <adam_g> mestery put up https://review.openstack.org/#/c/319447/ which adds a vagrant config to spinup astara devstack. i dont use vagrant so if someone who does wants to look it over, that'd be cool
18:26:06 <markmcclain> I was working through it earlier this morning
18:27:06 <markmcclain> thinking we might want to tweak a bit of the readme since virtualbox virt-in-virt is painfully slow since it's all emulated
18:27:43 <adam_g> coool. im too married to my homegrown kvm + qcow2 snapshots + nfs mounted /opt/stack to kick tires on vagrant
18:27:54 <ryanpetrello> heh
18:28:32 <ryanpetrello> okay, sounds like we're done
18:28:37 <ryanpetrello> o/
18:28:43 <phil_h> I have been asked to present Astara to the OpenStack Ansible folks this Thursday
18:28:50 <markmcclain> phil_h: cool
18:29:10 <phil_h> They have been looking at using Octavia
18:29:12 <adam_g> phil_h: oh cool. to give an overview of it, or present ansible integration?
18:29:26 <phil_h> I have been pushing Astara to the Rackspack folks
18:29:46 <phil_h> I need to get some more info in the morning
18:30:16 <phil_h> I'll let Eric know and he can pass it on to anyone else that is interested
18:30:19 <adam_g> neat
18:30:20 <adam_g> thanks
18:30:39 <eric_lopez> phil_h: lets talk I've been playing with OSAD to figure out the architecture
18:31:19 <phil_h> If we can get them into astara using loadbalancers then we can push them all the way into astara in hte furture
18:32:59 <phil_h> Yes, lets talk later today
18:33:18 <phil_h> call me or send me a message
18:33:29 <ryanpetrello> #endmeeting