04:00:46 <krtaylor> #startmeeting third-party
04:00:47 <openstack> Meeting started Wed Feb 11 04:00:46 2015 UTC and is due to finish in 60 minutes.  The chair is krtaylor. Information about MeetBot at http://wiki.debian.org/MeetBot.
04:00:48 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
04:00:50 <openstack> The meeting name has been set to 'third_party'
04:01:03 <krtaylor> who is here for the Third Party CI Working Group meeting?
04:02:09 * krtaylor thinks this may be a quick meeting
04:09:31 <patrickeast> sry i’m late, is the third party meeting still going on?
04:09:44 * patrickeast goes to check eavesdrop
04:09:54 <krtaylor> hi patrickeast
04:10:00 <patrickeast> krtaylor: hey
04:10:15 <krtaylor> yeah, I guess, unusually light attendance tonight
04:10:40 <patrickeast> gotcha
04:10:44 <krtaylor> so I may defer the agenda if it is just us here
04:11:27 <patrickeast> yea, can’t get through much without more folks around
04:11:34 <krtaylor> patrickeast, thanks for showing up, did you have anything to talk about?
04:11:59 <patrickeast> not much, i was hoping maybe someone would have advice wrt to this change set https://review.openstack.org/#/c/153086/
04:12:08 * krtaylor looks
04:12:50 <patrickeast> basically if your nodepool template instance is on a 10.0.0.x address image caching is broken
04:13:40 <patrickeast> and i can’t figure out a good way to get environment variables in there since the list image script is called by python with a subproccess call that doesn’t do shell=True
04:13:58 <patrickeast> not sure if i should try and fix nodepool scripts so i can hack around it or push harder to get something in devstack
04:14:19 <patrickeast> or give up and download images every time :D
04:14:45 <krtaylor> patrickeast, so you were wondering about sdague 's comment
04:14:53 <patrickeast> yea
04:15:12 <patrickeast> i don’t think we can actually set that easily without modifying the infra scripts
04:15:56 <krtaylor> hm, I guess I need to go look at the bug
04:16:46 <krtaylor> so, why doesnt it get HOST_IP?
04:18:00 <patrickeast> it does, but its address is something like 10.0.0.82, which then gets filtered out in the get_default_host_ip method because it falls into the default devstack FIXED_RANGE
04:18:57 <patrickeast> https://github.com/openstack-dev/devstack/blob/master/functions-common#L765
04:19:05 <krtaylor> ah, hm, well actually, I kinda like this fix  :)
04:19:42 <krtaylor> right, now I understand
04:20:33 <krtaylor> ok, so I also understand the comment about changing the fixed range, is that not an option?
04:21:17 <krtaylor> or...
04:21:29 <patrickeast> the problem is that there isn’t a way to inject environment variables in, short of modifying the list image script
04:21:38 <patrickeast> since it is called by cache_devstack.py
04:21:57 <patrickeast> so i could try and tackle it by modifying the infra nodepool scripts
04:22:14 <krtaylor> it does seem hacky that the filter is fixed, maybe that range should be passed instead of assumed
04:23:22 <krtaylor> yeah, in this case, I'd find sdague and discuss the worth statement, he may have something else in mond
04:23:29 <krtaylor> mind
04:23:30 <patrickeast> sounds good
04:23:44 <patrickeast> i’ll try and catch him tomorrow morning
04:25:29 <krtaylor> patrickeast, did you see last week's discussion highlighting a third party CI service?
04:25:42 <patrickeast> oo no, i missed that
04:26:15 <krtaylor> rfolco from my team presented on our environment, how we did things and fix problems for our environment
04:27:19 <krtaylor> I'd like to continue that, in Paris the response was very positive for getting everyone to share their details and scripts that made their life easier, etc
04:27:44 <krtaylor> patrickeast, so you can prob see where I am headed :)
04:27:48 <patrickeast> yea thats really cool, it definitely is beneficial to share that kind of stuff
04:27:58 <patrickeast> i’m happy to share what we’ve got
04:28:00 <krtaylor> would you be willing to share your environment,
04:28:06 <krtaylor> cool :)
04:28:10 <krtaylor> you beat me to it
04:28:14 <patrickeast> haha
04:28:22 <krtaylor> is this slot easier for you?
04:28:26 <patrickeast> i’ll have to double check that there isn’t anything super secret
04:28:31 <patrickeast> uhh some days
04:28:43 <patrickeast> the earlier one works if i remember to get up in time
04:28:53 <patrickeast> since its right before the cinder meeting i also attend
04:29:48 <krtaylor> yeah, that was another scheduling problem, I was surprised that this time won the vote honestly
04:30:38 <krtaylor> patrickeast, I'll ping you later and figure out a time/week that would be good for you
04:30:49 <patrickeast> sounds good
04:30:53 <krtaylor> I was going to set up a template for stuff to talk about
04:31:02 <krtaylor> make it easier for folks to share
04:31:07 <patrickeast> thats a good idea
04:31:49 <krtaylor> and, on the agenda for tonight, I'd like to get a repo for teams to share their scripts that help them run tests, check logs, status, whatever
04:32:22 <krtaylor> it doesnt make sense for each of us to re-invent the wheel every time we roll out another CI
04:32:52 <krtaylor> asselin 's in-tree proposal will help that a lot also
04:32:55 <patrickeast> yea definitely, we probably all have made tools to monitor them
04:33:33 <krtaylor> check out -> https://review.openstack.org/#/c/139745/
04:33:45 <krtaylor> oh, hehheh
04:33:48 <krtaylor> #link https://review.openstack.org/#/c/139745/
04:34:03 <krtaylor> thats the in-tree spec, its a good read
04:35:17 <patrickeast> ok cool, i’ll make sure to check it out
04:35:26 <patrickeast> that will be great to have
04:35:41 <krtaylor> agreed
04:37:50 <patrickeast> i’ve got to head out
04:38:05 <krtaylor> yeah, time to wrap this up I guess
04:38:16 <krtaylor> patrickeast, thanks for showing up!
04:38:23 <patrickeast> thanks for the help on my review, and i’ll be in touch about presenting our ci system
04:38:30 <krtaylor> perfect
04:39:11 <patrickeast> krtaylor: cya
04:41:58 <krtaylor> anyone else?
04:43:30 <krtaylor> alright then, that's a wrap
04:43:55 <krtaylor> we'll defer the agenda till next Wed
04:44:05 <krtaylor> #endmeeting