[01:11] <blackboxsw> powersj: done last pass on comments completed w/ https://code.launchpad.net/~powersj/cloud-init/+git/cloud-init/+merge/324136
[01:12] <blackboxsw> yeah you really did a heavy lift here thanks
[01:19] <powersj> Thanks!
[17:52] <powersj> https://code.launchpad.net/~powersj/cloud-init/+git/cloud-init/+merge/324201
[17:52] <powersj> smoser: ^
[18:37] <smoser> powersj, responded there..
[18:38] <smoser> http://paste.ubuntu.com/24594336/
[18:38] <powersj> smoser: Will do, thx
[18:48] <smoser> larsks, is akaris avaailable on irc ?
[18:52] <larsks> smoser: i can try to find out.  Give me a second...
[18:54] <akaris> smoser hey
[18:54] <akaris> how are you?
[19:29] <rharper> powersj: had some discussion today around the daily repo for cloud-init rpms ... is there work to do there? discussions ?
[19:30] <smoser> akaris, hey.
[19:30] <akaris> smoser so I just pushed a verification / valueerror if the protocol is not in ipv4 or ipv6
[19:30] <akaris> smoser I'll have to test centos/rhel 5/6/7 again
[19:31] <smoser> akaris, awesome. thank you. that really does look good
[19:31] <smoser> i was looking at the mask2cidr stuff
[19:31] <smoser> that is a mess
[19:31] <smoser> :)
[19:31] <larsks> akaris: thanks for taking on that patch.  That whole file is a bit hairy!
[19:31] <smoser> trying to make it into some sense.
[19:32] <smoser> yeah. the tools/net-convert.py is quite helpful
[19:32] <powersj> rharper: ah yes there is work. We need to figure out how to automate the creation daily. I still am in the investigation phase of how to make the automation happen.
[19:32] <powersj> rharper: trying to see what APIs are available or how to get those builds to happen in an automated fashion using COPR
[19:33] <powersj> more than willing to add you to the group on COPR and let you play with it as well
[19:33] <powersj> https://copr.fedorainfracloud.org/coprs/g/cloud-init/cloud-init/
[19:33] <rharper> cool
[19:33] <akaris> smoser that mask2cidr stuff ... I think I triggered it because I deleted the wrong data from cloud-init's cache
[19:33] <akaris> I was just figuring out via trial/error what i needed to delete to trigger a rerun
[19:34] <akaris> and when I finally got there, it turns out that I somehow still must have had some data (because the mask was now converted to an integer)
[19:34] <akaris> and that obviously doesn't fit into that method
[19:35] <smoser> but i dont know how you got the wrong stuff.
[19:35] <akaris> smoser I think I added that in the bug report, sec
[19:36] <akaris> https://bugs.launchpad.net/cloud-init/+bug/1684349       rm -Rf /var/lib/cloud/data/* ; cloud-init --force init
[19:36] <akaris> this was on RHEL with cloud-init 0.7.9
[19:36] <akaris> and an OSP setup with config drive
[19:37] <akaris> I'm likely just hitting some corner case because I don't know how to delete that data correctly ;-)
[19:37] <akaris> as a matter of fact, if I delete /var/lib/cloud/* then with the config drive setup, I cannot rerun cloud-init
[19:38] <akaris> is there some other way to make it run just as if this was right after the very first boot?
[19:38] <larsks> akaris: that's odd; I'm pretty sure that should work...what happens when you try?
[19:38] <akaris> lasks I think nothing happened. I'll have to bring my env back up and I can reconfirm
[19:39] <akaris> these instances cannot reach the metadata agent
[19:39] <akaris> the customer has a scenario where there is no L2 adjacency between the instance and OSP
[19:39] <larsks> ...which should be okay, if you're using config drive.
[19:39] <smoser> akaris, well yeah, that should have worked.
[19:39] <smoser> i dont know what state it would have found.
[19:40] <akaris> may I need to delete that data and then reboot the instance ? because I did  rm -Rf /var/lib/cloud/*   ;  cloud-init --force init     and that didn't work ... but that's weeks ago btw
[19:40] <larsks> akaris: for what it's worth, I rm /var/lib/cloud/* and re-run cloud-init all the time, and I haven't run into a particular problem.
[19:40] <akaris> let me recreate that test environment and test again. that will be better than speculating
[19:41] <smoser> yea, that is what is wierd. that really should work. if you could get the network_data.json that'd be helpful
[19:41] <akaris> larsks ok cool, I'll retry that
[19:41] <akaris> smoser ack
[19:41] <akaris> gimme a few hours to get that back up running