Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2013-11-29

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:19 rmoe joined #puppet-openstack
00:20 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::mon create or destroy monitors  https://review.openstack.org/56841
00:52 * dachary hit by https://bugs.launchpad.net/ubuntu/+source/virtualbox/+bug/1220724
00:52 dachary & upgraded...
00:56 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::mon create or destroy monitors  https://review.openstack.org/56841
01:59 rongze joined #puppet-openstack
02:04 openstackgerrit A change was merged to stackforge/puppet-cinder: Improve Cinder Backup support  https://review.openstack.org/58792
02:20 rongze joined #puppet-openstack
03:06 rongze joined #puppet-openstack
04:08 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-cinder: Add Ceph as backup backend support  https://review.openstack.org/59006
04:12 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-neutron: Ensure ml2 is set in core_plugin  https://review.openstack.org/58881
04:17 rongze joined #puppet-openstack
05:08 rongze joined #puppet-openstack
05:15 sputnik13 joined #puppet-openstack
05:35 michchap joined #puppet-openstack
06:06 Jean-Roger joined #puppet-openstack
07:06 rongze joined #puppet-openstack
07:28 rongze joined #puppet-openstack
07:50 rongze_ joined #puppet-openstack
08:11 rongze joined #puppet-openstack
08:11 bauzas joined #puppet-openstack
08:21 bauzas joined #puppet-openstack
08:42 bauzas joined #puppet-openstack
08:45 rongze joined #puppet-openstack
09:06 bauzas joined #puppet-openstack
09:19 derekh joined #puppet-openstack
09:22 larsks_ joined #puppet-openstack
09:22 francois1 joined #puppet-openstack
09:25 cherriges joined #puppet-openstack
09:27 odyi joined #puppet-openstack
09:27 Jean-Roger joined #puppet-openstack
09:30 bauzas joined #puppet-openstack
09:34 pabelanger joined #puppet-openstack
09:48 bauzas joined #puppet-openstack
10:38 michchap_ joined #puppet-openstack
12:24 xingchao joined #puppet-openstack
12:26 LAnthony joined #puppet-openstack
12:39 xingchao joined #puppet-openstack
13:11 rongze joined #puppet-openstack
13:13 LAnthony left #puppet-openstack
13:16 LAnthony joined #puppet-openstack
13:16 LAnthony left #puppet-openstack
14:05 xingchao joined #puppet-openstack
14:23 mjblack joined #puppet-openstack
14:24 mjblack joined #puppet-openstack
14:27 xingchao joined #puppet-openstack
14:43 xingchao joined #puppet-openstack
14:52 rmoe joined #puppet-openstack
15:10 xingchao joined #puppet-openstack
15:29 rongze joined #puppet-openstack
16:14 tnoor1 joined #puppet-openstack
16:20 LAnthony joined #puppet-openstack
16:20 LAnthony left #puppet-openstack
16:27 mgagne joined #puppet-openstack
16:27 mgagne joined #puppet-openstack
17:42 bauzas joined #puppet-openstack
18:06 dmsimard joined #puppet-openstack
18:19 dachary dmsimard: hi. I fixed the cephx of ceph::mon + added integration tests.
18:21 dachary michchap_: did you catch the answer from Dan ? I roughly summarized it at https://wiki.openstack.org/wiki/Puppet-openstack/ceph-blueprint#Support_scenario_based_deployment
18:33 dmsimard dachary: oi
18:33 dmsimard dachary: Will check when I have time
18:33 dachary thanks !
18:38 rongze joined #puppet-openstack
19:04 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: ceph::mon create or destroy monitors  https://review.openstack.org/56841
19:08 dmsimard Is there a way to do a side-by-side diff of two different patchsets from the gerrit interface ?
19:08 dmsimard Gets confusing to see what changed from the previous patchset(s) when there are so many :)
19:17 dachary yes, gerrit allows you to do that dmsimard
19:18 dachary dalgaaf_: https://review.openstack.org/#/c/56841/ is ready for your review
19:18 dmsimard dachary: How ? I can't see
19:18 mgagne dmsimard: Old Version History: select previous patch set
19:19 mgagne dmsimard: won't be useful if commit got rebased on master though.
19:20 dachary mgagne: hi :-) I realize https://review.openstack.org/#/c/56841/ is difficult to review without intimate knowledge of Ceph. I would very much appreciate a style / conventions review if you have time though.
19:22 mgagne dachary: I sure have one comment and it's about the commit message itself: it's too huge. If there is any useful info for the end user, it should be in the repo itself (maybe in docs/), not the commit message. as for coding convention described in commit message, commit messages aren't tutorials. if the info is worth spreading and sharing, it should be put in the wiki or somewhere else.
19:23 mgagne dachary: imagine the guy reading "ceph::mon create or destroy monitors" only to find out it's a tutorial on resource ordering
19:23 dachary mgagne: ok. The patch is way too big.
19:24 mgagne dachary: I don't know how people want to use launchpad (or not) but I would suggest splitting the wiki page in multiple blueprints so people can get a grasp of what's in progress and what has been done.
19:25 mgagne dachary: and reference said blueprint in commit message
19:25 dachary the comment on resource ordering is explaining the choice made on resource ordering. It was not meant as a tutorial but as a justification.
19:26 dachary mgagne: I got the general message : this is too big ( message, patch, blueprint )
19:28 mgagne dachary: the part about "coping with key management policies" should be documented in the module itself, again, maybe in docs/ or README. IMO, it is useful information for the end users.
19:29 dachary mgagne: ok
19:29 dachary mgagne: is there something else that strikes you as odd when browsing https://review.openstack.org/#/c/56841/14/manifests/mon.pp,unified ?
19:30 mgagne dachary: iweb copyright and original author got removed =)
19:30 dachary I'm very capable of writing bizarre looking things
19:30 dachary mgagne: ahahah :-)
19:31 mgagne dachary: kind of odd for someone that hammered the importance of license and authorship right off the bat ;)
19:32 dachary there was no work to begin with. I dutifully preserve copyright when it is about something that's copyrightable.
19:32 mgagne dachary: tbh, there isn't much left of the original code :P
19:32 dachary I'll add it back.
19:33 mgagne dachary: well, you have a point, there is nothing left
19:33 dachary Although I wrote the code, dmsimard & your reviews are part of the work.
19:33 dmsimard mgagne, dachary: I saw this kind of problem coming with multiple people/organizations working on the project
19:34 dmsimard I suck at licenses so I'm really not into it
19:34 rongze joined #puppet-openstack
19:36 dachary mgagne: I have doubts about the use of a temporary file and they way it is done
19:37 dachary exec { 'rm-keyring':
19:37 dachary command => "/bin/rm ${keyring_path}",
19:37 dachary }
19:37 dachary to remove it
19:40 dachary I also have doubts about the inlined shell script. Although it's really 4 lines ( set variable + mkdir + ceph-mon mkfs + touch || rm ) I'm not sure how confortable people around here would be about that.
19:41 dachary I'm happy about it. Mostly because it's good for debug and short. And I can't see any other way except writing dozens of puppet lines to do the same.
19:42 dachary it's also the first time I implement a resource that supports ensure. I may have done things that are not by the book.
19:44 dachary as you can see I feel pretty insecure about this piece of code. I'm happy because it does pass integration tests. But ...
20:11 rongze joined #puppet-openstack
20:46 rongze joined #puppet-openstack
21:21 tnoor1 joined #puppet-openstack
21:48 rongze joined #puppet-openstack
21:49 michchap_ dachary: yep - passing hashes :)
21:49 dachary michchap_: now I understand what it means ;-)
21:50 michchap_ dachary: I wasn't aware of create_resources...it's a bit neater than passing a list to defined types
21:50 dachary it really is a onliner :-) https://review.openstack.org/#/c/56841/14/manifests/mons.pp,unified
21:51 michchap_ but the yaml stops being very neat, since it now has inline hashes
21:51 michchap_ (I'm OK with that)
21:54 marun joined #puppet-openstack
22:02 francois joined #puppet-openstack
22:32 dmsimard dachary: sorry :P
22:33 dachary dmsimard: what about ?
22:33 dmsimard -1'd mon :(
22:35 dachary If that's all you have against it, it's actually very good news dmsimard  :-)
22:36 dmsimard I don't know, I checked quickly before leaving for the weekend
22:36 dmsimard I saw some discussion on irc but didn't pay much attention, are we changing anything else ?
22:37 dachary dmsimard: that being said, hardcoded path can are common place, it seems.
22:37 dachary for instance nova has only hardcoded paths
22:38 dachary since make install sets the path, there is no variance that would mandate a path, don't you think ?
22:38 dmsimard It's not just for ceph, it's in general
22:39 dmsimard When you're in your shell, you don't type /bin/rm <file> - there is the PATH environmental variable for you that does the work of finding your binary
22:39 dachary I'd like to take example from a puppet module in the openstack set, which one is not setting hard coded execs ?
22:41 dmsimard Don't know if there are openstack modules that do that, I thought it was a standard thing :)
22:42 dmsimard Only reference I can find to defend myself right now is http://www.puppetcookbook.com/posts/set-global-exec-path.html but that's not good :P
22:42 dmsimard Trying to see if there's a documented best practice
22:43 dmsimard http://projects.puppetlabs.com/projects/1/wiki/Puppet_Best_Practice2#sample-sitepp
22:44 dmsimard Don't know what to tell you, could ask the puppet-openstack folks if it should be like this at all for their modules if it's not already like this - I guess it can break things if the path is not set and people are assuming it works without setting one
22:45 dmsimard Not hard coding the path probably makes handling various distributions easier? I don't know
22:47 dachary I changed the path for each command
22:49 dmsimard mgagne left the office already but I'll poke him monday see what he has to say
22:49 dmsimard I'm curious
22:50 rongze joined #puppet-openstack
22:50 dmsimard My brain wants to explode right now, think I'm overdue for a weekend
22:56 dachary :-)
22:57 michchap joined #puppet-openstack
23:48 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-cinder: Add Ceph as backup backend support  https://review.openstack.org/59006
23:52 rongze joined #puppet-openstack

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary