/srv/irclogs.ubuntu.com/2016/04/12/#cloud-init.txt

=== rangerpbzzzz is now known as rangerpb
dmsimardHi ! I was looking to get an OpenStack environment going with CentOS guests with config-drive provided network configuration. It looks like I would need the RHEL version of http://bazaar.launchpad.net/~cloud-init-dev/cloud-init/trunk/revision/1189?start_revid=118913:39
dmsimardharlowja: ^ I was told you might be able to help ?13:40
smoserharlowja, ^14:56
harlowjasup16:32
smoserharlowja, dmsimard wants you to fix centos for network config16:56
smoserand config drive :)16:56
harlowja:-P16:56
harlowjaya, damn16:56
harlowjalol16:56
harlowjai knew it, ha16:56
dmsimardoh hai16:56
dmsimardwell, I'm willing to help and all16:56
dmsimardfriend here mgagne pointed me to an (older?) iteration of that implementation I guess, https://github.com/mgagne/cloud-init-fedora-pkg/blob/epel7/cloud-init-0.7.5-network-info-support.patch16:57
dmsimardBut the way it was just merged (three weeks ago, talk about timing) looks similar yet different16:58
mgagneharlowja: original version here: https://github.com/racker/cloud-init-fedora-pkg16:58
dmsimardcloud-init will be in core OS packages in RHEL 7.3 and CentOS 7.3 this year so it'd be great to have a version cut with that support baked in16:59
harlowjamgagne dmsimard ya, i'd like to have that in also17:01
harlowjado u guys want to work on that, or should i?17:01
harlowjai moved companies recently, so i might have to do some of the legal CCLA crap again smoser17:01
harlowjai gotta aks on that17:01
harlowja*ask17:01
larsksharlowja: which goes back to my question yesterday asking about a 0.7.7 release... :)17:02
dmsimardI can help testing but I'm not knowledgeable enough about those bits to do it myself17:02
harlowjasmoser is mr.release ;)17:02
harlowjahe's gonna do git soon also17:02
harlowjalol17:02
harlowjai just know it17:02
harlowjai believe17:02
harlowjai'll send smoser a cookie if he does git17:02
harlowjalol17:02
larsksI would contribute to that cookie fund.17:03
harlowja:)17:04
smoserharlowja, git will come17:07
smosergit willi come17:07
harlowjasmoser tommorow?17:07
harlowjalol17:07
harlowjatoday?17:07
smoserafter 16.04 releases (next thursday)... not before.17:07
harlowja16.0417:07
harlowjadamn it17:07
harlowjaanother LTS17:08
harlowjai just reinstalled 14.0417:08
harlowjalol17:08
smoserif spandhe joins maybe we can convince her to do it.17:08
smosershe's still yahoo, right ?17:08
harlowjaya17:08
harlowjadmsimard mgagne if u guys want to get a patch working17:09
harlowja https://github.com/mgagne/cloud-init-fedora-pkg/blob/epel7/cloud-init-0.7.5-network-info-support.patch i guess ?17:09
harlowjajust maybe a few more tests ?17:09
harlowjacause that patch doesn't seem to have to many17:09
harlowjathen maybe by the time those tests happen, smoser will have git17:10
harlowjaand we can all have cookies17:10
harlowjai believe!17:11
harlowjalol17:11
mgagneharlowja: let me check if have a more up2date version17:13
harlowjamgagne  np17:13
mgagnebecause we are reading network_info.json now, not vendor_data.json17:13
harlowjaright17:13
harlowjathat's fine, some network_info.json test cases, that give a sample network_info.json in and check the output would be cool17:14
harlowja*and/or multiple sample inputs ---> check the outputs17:14
mgagnethat's scary, I think I fixed it for debian but not rhel17:15
mgagneharlowja: for trusty, for what it's worth: https://gist.github.com/mgagne/46748012efa1ff3389b380a25bedb14d17:17
harlowjacool, ya, tests for these would be nice ;)17:17
harlowjalike a bunch of input data files and expected output files17:17
harlowjafor debian and ubuntu and redhat and ...17:18
harlowjamaybe to far-fetched, butttt17:18
mgagneI might not be the right person to ask to invest time on it atm ^^'17:18
harlowjakk17:18
harlowjadmsimard ?17:18
mgagnedon't know, doesn't work with us anymore =)17:19
harlowjadelegation not working17:19
mgagneshould patch be targeted for v2 or 0.7.x is fine?17:19
harlowjamust delegate harder, lol17:19
mgagnestill waiting for my clone to arrive from mail17:19
harlowjaUPS lose it?17:20
harlowjalol17:20
mgagneunfortunate yes =( they also say we can't ship bio stuff :P17:20
harlowjaah17:20
harlowjagood point17:20
klindgrenI hear fedex or DHL is looser on the restrictions17:23
harlowjanice nice17:23
harlowjadmsimard let me know if u have time for those kinds of tests17:24
harlowjaif not, i can try, but no guarantees on time for that :-P17:24
harlowjaor smoser can get git and then we can accept patch, and i can start adding tests17:24
harlowjaand cookies17:24
larsksOut of curiousity, what is the effort involved in moving to git?  Is it more than just import-from-bzr-export-to-github?17:31
harlowjalarsks we already have that17:31
harlowjabut its not in sync with bzr17:31
harlowjalarsks https://github.com/openstack/cloud-init is that (export from bzr, import to github)17:31
larsksRight, I have noted that :).  Is the goal to maintain the bzr repo and keep it in sync automatically?17:31
harlowjabut bzr has a native mode17:31
larsksOr why not just perform the manual sync once, then kill the bzr repo and go on our merry way?17:32
harlowjai delegate question to smoser17:32
harlowja:)17:32
larsksI am using git locally (via git-remote-bzr), which works great for me but confused the heck out of dmsimard yesterday when I forgot and gave him a git commit id from my local repo...17:33
harlowjai think the idea was to keep 0.7.x on bzr (and bzr-git) and anything new using that git above17:34
larsksThat seems clunky, because people are going to continue to make patches against the 0.7.x code for packages, and it would be nice to have it in the same repo as newer stuff for ease for cherry-picking, etc.17:35
harlowjalarsks  i know, its either that or we talk about killing bzr again :-P17:36
smoserlarsks, it would be git on launchpad.17:37
larskssmoser: yeah, that's fine.17:38
harlowjabut only for 0.7.x17:38
harlowjaand then i think i'd ask the openstack-infra guys to kill the 0.7.x branch on  https://github.com/openstack/cloud-init version17:38
smoserright.17:38
harlowjawhich hopefully they would do17:38
smoserthe other thing that is needed is the revision number thing...17:38
larsksWhich revision number thing?17:39
dmsimardSorry was afk getting food, reading backlog17:39
smoserharlowja, had something there. but basically i quite often make use of the bzr revision number for things (snapshots) and want some equivalent17:39
smoseri forget what it is... theres a apython module that gives snapsthos numbers.17:39
harlowjathere's a few17:39
harlowjapbr17:39
harlowjaa few others17:39
harlowjahttps://github.com/habnabit/vcversioner (another one)17:40
harlowjaand i'm sure a few more17:40
smoserthat wasnt it i dont htink17:41
larsksOr just use short git commit ids (which are nice because they can be used directly in git commands, etc)17:41
smoseryeah, but they dont increase.17:41
dmsimardok, I read all that17:41
larsksTrue.17:41
dmsimardI definitely can't write python tests to save my life, sorry :p17:41
larsksThat would be a great movie.  "Do you expect me to write a unit test? No, Mr. Bond; I expect you to die."17:42
harlowjalol17:43
* smoser writes down not to trusty larsks's movie recommendations17:49
smosers/trusty/trust/17:49
smoser(fingers type a y if i type trust)17:49
larsksYeah, Trusty is old news.17:49
dmsimardharlowja: so was that patch a good fit for what landed in trunk, then ?18:10
dmsimardminus the debian bits18:10
* harlowja hasn't looked over to much yet18:10
harlowjalol18:10
harlowjadmsimard i'll get into looking at it more18:11
dmsimardno emergencies or anything, I resorted to using dhcp for the time being :(18:12
harlowjak18:12
dmsimardDo let me know if you'd like to test anything though, this is something I can help with.18:13
harlowjak18:13
harlowjathx18:13
=== rangerpb is now known as rangerpbzzzz

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!