| jog0 | git the diff stats between Nova Diablo and Essex (git diff --stat stable/diablo milestone-proposed): | 00:04 |
|---|---|---|
| jog0 | 1067 files changed, 265826 insertions(+), 118319 deletions(-) | 00:04 |
| *** rkukura has joined #openstack-dev | 00:05 | |
| *** jdg has quit IRC | 00:05 | |
| *** bsza has joined #openstack-dev | 00:06 | |
| adam_g | /winzu/away gone | 00:06 |
| adam_g | winzu! | 00:06 |
| *** heckj has joined #openstack-dev | 00:07 | |
| heckj | soren: you around? | 00:07 |
| *** dachary has quit IRC | 00:07 | |
| *** heckj has quit IRC | 00:07 | |
| *** davidkranz_ has joined #openstack-dev | 00:09 | |
| *** gyee has quit IRC | 00:10 | |
| *** ayoung has joined #openstack-dev | 00:14 | |
| *** bsza has quit IRC | 00:18 | |
| *** LarsErikP has quit IRC | 00:19 | |
| *** bhall has quit IRC | 00:22 | |
| *** maplebed has quit IRC | 00:27 | |
| *** dtroyer has quit IRC | 00:29 | |
| *** spiffxp has quit IRC | 00:30 | |
| *** davidkranz_ has quit IRC | 00:46 | |
| *** vizsla has joined #openstack-dev | 00:49 | |
| *** troytoman-away is now known as troytoman | 00:50 | |
| *** johnpostlethwait has quit IRC | 00:52 | |
| *** johnpostlethwait has joined #openstack-dev | 00:53 | |
| *** devananda has quit IRC | 00:55 | |
| *** vincentricci has quit IRC | 00:59 | |
| *** pixelbeat has quit IRC | 01:02 | |
| *** vincentricci has joined #openstack-dev | 01:14 | |
| *** vizsla has quit IRC | 01:14 | |
| *** hub_cap has joined #openstack-dev | 01:17 | |
| *** vincentricci has quit IRC | 01:19 | |
| *** dalang has quit IRC | 01:23 | |
| *** dtroyer has joined #openstack-dev | 01:25 | |
| *** shevek_ has quit IRC | 01:27 | |
| *** ywu has joined #openstack-dev | 01:27 | |
| *** Mandell has joined #openstack-dev | 01:28 | |
| *** dtroyer has quit IRC | 01:28 | |
| *** davlaps has quit IRC | 01:31 | |
| *** johnpostlethwait has quit IRC | 01:35 | |
| *** agonella has quit IRC | 01:42 | |
| *** novas0x2a|laptop has quit IRC | 01:43 | |
| *** Ryan_Lane has quit IRC | 01:45 | |
| *** vizsla has joined #openstack-dev | 01:46 | |
| *** dalang has joined #openstack-dev | 02:02 | |
| *** mnewby has quit IRC | 02:02 | |
| *** flaviamissi has joined #openstack-dev | 02:06 | |
| *** sandywalsh has quit IRC | 02:07 | |
| *** anotherjesse_zz is now known as anotherjesse | 02:09 | |
| *** dalang has quit IRC | 02:10 | |
| *** flaviamissi has quit IRC | 02:11 | |
| *** flaviamissi has joined #openstack-dev | 02:12 | |
| *** mdrnstm has quit IRC | 02:12 | |
| *** agonella has joined #openstack-dev | 02:12 | |
| *** agonella has quit IRC | 02:14 | |
| *** jdurgin has quit IRC | 02:18 | |
| *** dtroyer has joined #openstack-dev | 02:21 | |
| *** halfss_ has joined #openstack-dev | 02:27 | |
| *** anotherjesse is now known as anotherjesse_zz | 02:28 | |
| *** rods has quit IRC | 02:28 | |
| *** halfss__ has joined #openstack-dev | 02:29 | |
| *** phorgh has quit IRC | 02:29 | |
| *** halfss_ has quit IRC | 02:33 | |
| *** reed has quit IRC | 02:33 | |
| *** anotherjesse_zz is now known as anotherjesse | 02:33 | |
| *** Mandell has quit IRC | 02:39 | |
| *** mdomsch has joined #openstack-dev | 02:42 | |
| *** mikal has quit IRC | 02:44 | |
| *** mikal has joined #openstack-dev | 02:48 | |
| *** troytoman is now known as troytoman-away | 02:54 | |
| rmk | Are all the milestone-proposed branches now essex RC1? | 02:55 |
| *** sandywalsh has joined #openstack-dev | 02:59 | |
| *** ywu has quit IRC | 03:00 | |
| *** anotherjesse is now known as anotherjesse_zz | 03:03 | |
| *** mdrnstm has joined #openstack-dev | 03:08 | |
| *** paulstallworth has joined #openstack-dev | 03:11 | |
| *** dalang has joined #openstack-dev | 03:17 | |
| *** jmckenty has joined #openstack-dev | 03:18 | |
| *** spiffxp has joined #openstack-dev | 03:18 | |
| *** sandywalsh has quit IRC | 03:23 | |
| *** danwent has joined #openstack-dev | 03:25 | |
| *** pixelbeat has joined #openstack-dev | 03:27 | |
| *** hub_cap has quit IRC | 03:28 | |
| *** pixelbeat has quit IRC | 03:33 | |
| *** sandywalsh has joined #openstack-dev | 03:36 | |
| *** Mandell has joined #openstack-dev | 03:38 | |
| *** vizsla has quit IRC | 03:51 | |
| *** dolphm has quit IRC | 04:01 | |
| *** zaitcev has quit IRC | 04:05 | |
| *** dalang has quit IRC | 04:08 | |
| *** dalang has joined #openstack-dev | 04:14 | |
| *** mnewby has joined #openstack-dev | 04:16 | |
| *** dalang has quit IRC | 04:19 | |
| *** dolphm has joined #openstack-dev | 04:19 | |
| *** mnewby has quit IRC | 04:20 | |
| *** mnewby has joined #openstack-dev | 04:26 | |
| *** dolphm has quit IRC | 04:34 | |
| *** dolphm has joined #openstack-dev | 04:36 | |
| *** reed has joined #openstack-dev | 04:39 | |
| *** dolphm has quit IRC | 04:39 | |
| *** jmckenty has quit IRC | 04:51 | |
| *** refactor has joined #openstack-dev | 04:52 | |
| *** dtroyer has quit IRC | 04:53 | |
| *** paulstallworth has left #openstack-dev | 05:03 | |
| *** hattwick has quit IRC | 05:03 | |
| mnewby | Calling all gerrit ninjas. Is there a way to submit a change that depends on another, already submitted change? | 05:27 |
| *** deshantm has quit IRC | 05:36 | |
| *** Mandell has quit IRC | 05:42 | |
| *** flaviamissi has quit IRC | 05:51 | |
| *** reed has quit IRC | 05:59 | |
| *** danwent has quit IRC | 06:00 | |
| *** Mandell has joined #openstack-dev | 06:03 | |
| markmc | mnewby, yes, your commit should just have the other change's commit as its parent | 06:12 |
| *** danwent has joined #openstack-dev | 06:12 | |
| markmc | mnewby, if you tell me which change, I can help | 06:12 |
| mnewby | Here is the first change: https://review.openstack.org/#change,5595 | 06:12 |
| *** Mandell has quit IRC | 06:13 | |
| mnewby | I branched off the branch for the first change, made another commit, but when I tried to submit it, git review warned that I was going to submit two commits (the first change and the second). | 06:13 |
| *** Mandell has joined #openstack-dev | 06:13 | |
| mnewby | Is it safe to do that - the first commit will be a no-op? | 06:13 |
| markmc | what does 'git log --oneline origin/master..' show? | 06:14 |
| *** danwent has quit IRC | 06:14 | |
| mnewby | nada | 06:15 |
| markmc | it should at least show this: | 06:17 |
| markmc | $ git log --oneline origin/master..0fa2dd7e | 06:17 |
| markmc | 0fa2dd7 Improve swift_auth test coverage + Minor fixes | 06:17 |
| markmc | but I was expecting there to be a second commit on your branch too? | 06:17 |
| mnewby | There is definitely a second commit. Both have change-ids. | 06:17 |
| *** Mandell_ has joined #openstack-dev | 06:17 | |
| *** Mandell has quit IRC | 06:18 | |
| mnewby | I'm not familiar with the oneline syntax, it doesn't seem to work. | 06:18 |
| mnewby | Ah, I get it. | 06:18 |
| mnewby | git log --pretty=oneline works. | 06:18 |
| markmc | ah, ok - older version of git | 06:18 |
| mnewby | My commit log shows the commit for the review I pointed you at and the new commit for the dependent change. | 06:19 |
| markmc | cool, then just "Type yes to confirm" :) | 06:20 |
| mnewby | Ok, cool :) | 06:20 |
| * markmc doesn't use git-review | 06:20 | |
| markmc | didn't know about the warning | 06:20 |
| * markmc usually does 'git push gerrit HEAD:refs/for/master' | 06:20 | |
| markmc | which doesn't give a warning | 06:20 |
| mnewby | Ah, worked like a charm. danke! | 06:21 |
| markmc | mnewby, did "The outstanding commits are:" show anything in that warning? | 06:21 |
| mnewby | Yes, the most recent commit and the one that I'd already submitted. | 06:21 |
| markmc | huh, interesting | 06:22 |
| markmc | git-review is using 'git log --oneline' too | 06:22 |
| * markmc shrugs | 06:22 | |
| mnewby | Weird. | 06:22 |
| mnewby | Oh, that does work. | 06:22 |
| mnewby | I was confused by the line-wrapping in my irc client. | 06:22 |
| mnewby | i was executing 'git log -- oneline' | 06:22 |
| mnewby | I didn't realize it was an option - I know some git options require the -- separator, but I'm still green wrt git. | 06:23 |
| markmc | ah, ok | 06:23 |
| markmc | it interpreted that as 'show me the log for the file called oneline' | 06:23 |
| *** bepernoot has joined #openstack-dev | 06:33 | |
| *** Ash has quit IRC | 06:35 | |
| *** bepernoot has quit IRC | 06:55 | |
| *** mnewby has quit IRC | 07:10 | |
| *** dachary has joined #openstack-dev | 07:10 | |
| *** ghe_ is now known as GheRivero | 07:25 | |
| *** openstackjenkins has quit IRC | 07:30 | |
| *** openstackjenkins has joined #openstack-dev | 07:31 | |
| *** apevec has joined #openstack-dev | 07:32 | |
| *** openstackjenkins has quit IRC | 07:36 | |
| *** openstackjenkins has joined #openstack-dev | 07:37 | |
| *** GheRivero has quit IRC | 07:38 | |
| *** ghe_ has joined #openstack-dev | 07:38 | |
| *** LarsErikP has joined #openstack-dev | 07:39 | |
| *** ChanServ sets mode: +o mtaylor | 07:54 | |
| *** mtaylor sets mode: -b GheRivero!*@* | 07:54 | |
| *** ohnoimdead has quit IRC | 07:54 | |
| *** zigo has joined #openstack-dev | 07:54 | |
| *** hattwick has joined #openstack-dev | 07:59 | |
| *** bepernoot has joined #openstack-dev | 08:03 | |
| *** littleidea has joined #openstack-dev | 08:03 | |
| *** Mandell_ has quit IRC | 08:09 | |
| *** reidrac has joined #openstack-dev | 08:21 | |
| *** milner has quit IRC | 08:22 | |
| *** ghe_ is now known as GheRivero | 08:25 | |
| *** hugokuo has joined #openstack-dev | 08:28 | |
| ttx | rmk: no. Missing keystone | 08:32 |
| *** milner has joined #openstack-dev | 08:36 | |
| *** shevek_ has joined #openstack-dev | 08:36 | |
| LarsErikP | question: when horizon gives me alot of HTTP400 when trying to recive usage info, sec_groups and so on... Where to look? | 08:38 |
| LarsErikP | and, do I need one service and one endpoint per compute-node in keystone when running multihost? | 08:39 |
| *** darraghb has joined #openstack-dev | 08:41 | |
| *** ohnoimdead has joined #openstack-dev | 08:54 | |
| *** dachary has quit IRC | 09:00 | |
| *** oneiroi has joined #openstack-dev | 09:03 | |
| *** oneiroi has joined #openstack-dev | 09:03 | |
| *** derekh has joined #openstack-dev | 09:06 | |
| *** oneiroi has quit IRC | 09:11 | |
| *** oneiroi has joined #openstack-dev | 09:12 | |
| *** dachary has joined #openstack-dev | 09:19 | |
| *** pixelbeat has joined #openstack-dev | 09:38 | |
| *** dneary has joined #openstack-dev | 09:40 | |
| *** dneary has quit IRC | 09:40 | |
| *** dneary has joined #openstack-dev | 09:40 | |
| *** zul has quit IRC | 09:40 | |
| *** zul has joined #openstack-dev | 09:42 | |
| *** vizsla has joined #openstack-dev | 09:47 | |
| *** rkukura has quit IRC | 09:51 | |
| *** vizsla has quit IRC | 09:52 | |
| *** rkukura has joined #openstack-dev | 09:53 | |
| *** refactor has quit IRC | 10:13 | |
| *** paulormg has joined #openstack-dev | 10:20 | |
| *** tryggvil_ has joined #openstack-dev | 10:21 | |
| *** markmc_ has joined #openstack-dev | 10:25 | |
| *** fc__ has joined #openstack-dev | 10:27 | |
| dachary | fc__: do you think https://github.com/puppetlabs/puppetlabs-swift can be used as is on Debian GNU/Linux ? | 10:29 |
| fc__ | dachary: don't know yet, I have to review it more in depth than I did | 10:31 |
| *** halfss__ has quit IRC | 10:32 | |
| dachary | fc__: ok. I came to realize that the target of puppetlabs-swift is diablo, meaning you'll probably need to adapt for essex | 10:50 |
| *** rods has joined #openstack-dev | 10:54 | |
| *** hugokuo has quit IRC | 11:00 | |
| *** mdrnstm has quit IRC | 11:04 | |
| *** markmc_ has quit IRC | 11:07 | |
| *** dneary has quit IRC | 11:25 | |
| *** sandywalsh has quit IRC | 11:25 | |
| *** vizsla has joined #openstack-dev | 11:34 | |
| *** sandywalsh has joined #openstack-dev | 11:38 | |
| *** TREllis has quit IRC | 11:43 | |
| *** asalkeld has quit IRC | 11:53 | |
| *** maploin has joined #openstack-dev | 12:13 | |
| *** maploin has quit IRC | 12:13 | |
| *** maploin has joined #openstack-dev | 12:13 | |
| *** vizsla has quit IRC | 12:13 | |
| *** flaviamissi has joined #openstack-dev | 12:20 | |
| *** alaski has joined #openstack-dev | 12:24 | |
| *** dneary has joined #openstack-dev | 12:24 | |
| *** dneary has quit IRC | 12:24 | |
| *** dneary has joined #openstack-dev | 12:24 | |
| *** lts has joined #openstack-dev | 12:25 | |
| *** kbringard has joined #openstack-dev | 12:26 | |
| *** kbringard1 has joined #openstack-dev | 12:30 | |
| *** roge has joined #openstack-dev | 12:30 | |
| *** kbringard has quit IRC | 12:31 | |
| *** roge has quit IRC | 12:35 | |
| *** roge has joined #openstack-dev | 12:37 | |
| *** ccorrigan has quit IRC | 12:38 | |
| *** kbringard1 has quit IRC | 12:43 | |
| *** mdomsch has quit IRC | 12:45 | |
| mtaylor | anybody know if there's a way to tell sqlalchmey to pretend that the underlying database doesn't support foreign keys and to not create them? | 12:50 |
| *** kbringard has joined #openstack-dev | 12:51 | |
| *** dprince has joined #openstack-dev | 13:05 | |
| *** SusanneB_ has quit IRC | 13:05 | |
| oneiroi | mtaylor: http://stackoverflow.com/questions/9410543/will-sqlalchemy-automatically-create-index-for-foreign-key _may_ help though it's not an accepted answer | 13:06 |
| mtaylor | oneiroi: thanks | 13:07 |
| mtaylor | oneiroi: oh, it's not the index on the foreign key I don't want | 13:08 |
| mtaylor | oneiroi: I don't want it to create the foreign key in the first place | 13:08 |
| oneiroi | hmmm | 13:09 |
| mtaylor | looking at the code, I think I'd have to write an overloaded DDLCompiler class | 13:10 |
| jaypipes | mtaylor: it won't if your metadata (model) doesn't have a back_ref IIRC | 13:10 |
| jaypipes | mtaylor: http://docs.sqlalchemy.org/en/latest/core/schema.html#defining-constraints-and-indexes | 13:10 |
| mtaylor | jaypipes: morning! I was just about to ping you for something else | 13:10 |
| jaypipes | mtaylor: really? I was just about to go get some coffee ;) | 13:11 |
| mtaylor | jaypipes: hrm. I should go get some too | 13:11 |
| mtaylor | jaypipes: but in this case... | 13:11 |
| mtaylor | jaypipes: I'm looking at Patrick's db performance things ... and honestly it would be really worthwhile (and not hard) to give sqlalchemy an option to not add foreign keys to the constraints even if it detects innodb | 13:12 |
| mtaylor | jaypipes: I don't want to change the model :) | 13:13 |
| *** refactor has joined #openstack-dev | 13:14 | |
| *** crobinso has joined #openstack-dev | 13:14 | |
| *** hub_cap has joined #openstack-dev | 13:21 | |
| *** hub_cap has quit IRC | 13:22 | |
| jaypipes | mtaylor: k, gimme a while to get to CaptTofu's stuff there... got that Road to Folsom meeting this mornign... | 13:22 |
| jaypipes | mtaylor: forget that... just saw Matt cancelled the meeting. | 13:23 |
| jaypipes | 12 minutes ago. | 13:24 |
| *** ayoung has quit IRC | 13:24 | |
| jaypipes | mtaylor: so I will look into those things as soon as I get coffee. :) | 13:24 |
| jaypipes | mtaylor: ah... so here's the thing... | 13:26 |
| *** deshantm has joined #openstack-dev | 13:27 | |
| jaypipes | mtaylor: every one of those queries is not applicable in Essex... :) | 13:28 |
| jaypipes | mtaylor: anyway, off for coffee. we can chat with brian later about this. | 13:28 |
| *** troytoman-away is now known as troytoman | 13:29 | |
| mtaylor | jaypipes: awesome | 13:31 |
| *** ayoung has joined #openstack-dev | 13:37 | |
| *** tryggvil_ has quit IRC | 13:42 | |
| *** mdomsch has joined #openstack-dev | 13:48 | |
| *** andrewsmedina has joined #openstack-dev | 13:53 | |
| *** rkukura has quit IRC | 13:55 | |
| *** AlanClark has joined #openstack-dev | 13:56 | |
| *** troytoman is now known as troytoman-away | 14:00 | |
| *** andrewsmedina has quit IRC | 14:00 | |
| *** alaski has quit IRC | 14:01 | |
| *** alaski has joined #openstack-dev | 14:01 | |
| *** dtroyer has joined #openstack-dev | 14:04 | |
| *** dtroyer has left #openstack-dev | 14:06 | |
| *** dtroyer_zzz is now known as dtroyer | 14:06 | |
| *** andrewsmedina has joined #openstack-dev | 14:07 | |
| jaypipes | mtaylor: alrighty Mr. Taylor. coffee and breakfast'd up. | 14:15 |
| *** dolphm has joined #openstack-dev | 14:16 | |
| *** dolphm_ has joined #openstack-dev | 14:16 | |
| jaypipes | mtaylor: so the trick with these slow database queries is that all of them are related (pun intended) to the now-gone networks table | 14:17 |
| *** littleidea has quit IRC | 14:17 | |
| jaypipes | mtaylor: the network data is now managed by Quantum, and so all of these queries are actually now messages sent to the network service and cached in the network_api for a period of time | 14:18 |
| jaypipes | mtaylor: Quantum and melange I should say... | 14:18 |
| jaypipes | mtaylor: the work that jkoelker and tr3buchet did in decoupling the network DB calls from the network manager is central to these database queries. And, while it might be fine to add an index here or there to the Diablo install of Nova, these queries don't appear in Essex... | 14:19 |
| jkoelker | well they may appear in essex | 14:20 |
| *** dolphm has quit IRC | 14:21 | |
| jkoelker | it depends on if you are using a legacy network manager or the quantum manager ;( | 14:21 |
| jkoelker | big hozers | 14:21 |
| *** dspano has joined #openstack-dev | 14:22 | |
| jaypipes | jkoelker: ok, yes, thx for that clarification... mtaylor ^^ | 14:25 |
| jaypipes | jkoelker: wasn't the FK from network to virtual_interface and network to fixed_ip removed in Essex as well? | 14:26 |
| jkoelker | it was in preperation | 14:26 |
| jkoelker | we didn't get as far as we wanted | 14:26 |
| jkoelker | we have a branch baking now that is totally decoupled from the db | 14:27 |
| jkoelker | but we didn't get it ready in time to merge prop it prior to feature freeze | 14:27 |
| jaypipes | jkoelker: could you elaborate on that? what has been done already and what hasn't? | 14:27 |
| jkoelker | so all the network functions from compute/api go over rabbit to the network manager now | 14:27 |
| jaypipes | right, I remember reviewing that series of patchsets | 14:28 |
| jkoelker | and we started on moving all the response data from the network manager over to using "the network model" | 14:28 |
| jkoelker | so far i think allocate_for_instance and get_instance_nw_info are the only ones that use it | 14:28 |
| jaypipes | k, and the network model is what has the caching layer on it, right? | 14:28 |
| jkoelker | correct, it has the ability to serialize itself to and from json easily | 14:29 |
| jaypipes | k, that is what I remembered. good. | 14:29 |
| jkoelker | and we stuff it into the instance_info_cache whenever we generate it | 14:29 |
| jaypipes | yeah. | 14:29 |
| jaypipes | jkoelker: thx mate :) | 14:29 |
| jaypipes | eglynn__: ping | 14:30 |
| jkoelker | cools, if you want to checkout out no-db network manager branch its at https://github.com/jkoelker/nova/tree/the_deuce | 14:30 |
| jkoelker | but its totally not ready for prime time yet | 14:31 |
| jkoelker | it only does "flat network" | 14:31 |
| *** stuntmachine has joined #openstack-dev | 14:32 | |
| jaypipes | jkoelker: k, will do. cheers! | 14:33 |
| *** troytoman-away is now known as troytoman | 14:35 | |
| dolphm_ | countdown page dissapearred? http://wiki.openstack.org/releasestatus/rc1.html | 14:36 |
| annegentle | ttx: ^^ | 14:37 |
| ttx | dolphm_: now at http://wiki.openstack.org/releasestatus/rc.html | 14:37 |
| dolphm_ | ttx: thanks | 14:38 |
| dolphm_ | ttx: also, the "Release Schedule" link on http://wiki.openstack.org/releasestatus/ points to Diablo | 14:39 |
| ttx | dolphm_: ack will fix | 14:40 |
| mtaylor | jaypipes, jkoelker thanks | 14:40 |
| *** tryggvil_ has joined #openstack-dev | 14:40 | |
| eglynn__ | jaypipes: pong | 14:41 |
| *** rkukura has joined #openstack-dev | 14:41 | |
| mtaylor | jaypipes: just to be pedantic though - I still want to add an option to sqlalchemy to tell it not to create the bloody things | 14:41 |
| mtaylor | jaypipes: because in a sqlalchemy context they're only useful actually existing in the db for reflection purposes - and if the schema is fully managed by sqlalchemy, they are operational overhead | 14:42 |
| *** shall_ has joined #openstack-dev | 14:42 | |
| LinuxJedi | mtaylor: mostly because MySQL doesn't do them properly ;) | 14:42 |
| *** dtroyer is now known as dtroyer_zzz | 14:43 | |
| annegentle | can I ever unabandon someone else's change? | 14:45 |
| annegentle | https://review.openstack.org/#change,4965 just needs a rebase, I could do it | 14:45 |
| *** alaski has quit IRC | 14:45 | |
| mtaylor | LinuxJedi: yes | 14:46 |
| mtaylor | also, they're a terrible idea for scaling, but I digress | 14:46 |
| *** zzed has joined #openstack-dev | 14:47 | |
| *** alaski has joined #openstack-dev | 14:47 | |
| jaypipes | LinuxJedi: MySQL doesn't do them at all. Innodb does ;) | 14:48 |
| eglynn__ | annegentle: one way I guess would be to apply the patch locally, rebase, then push back up to the gerrit with a different Change-Id (making it a fresh review) but keeping the original Author field intact | 14:48 |
| LinuxJedi | jaypipes: indeed :) | 14:48 |
| mtaylor | jaypipes: you say potato, I say salmiakki | 14:48 |
| jaypipes | LinuxJedi: and good morning :) | 14:48 |
| LinuxJedi | hehe :) | 14:48 |
| LinuxJedi | hey jaypipes | 14:48 |
| eglynn__ | jaypipes: missed your ping a few minutes ago | 14:48 |
| jaypipes | eglynn__: heya, I'm wondering if we can get jeblair to manually test your S3 branch(es) to see if that fixes the issue? | 14:48 |
| eglynn__ | jaypipes: yeah that's an idea, they're not very instrusive though, could easily revert again if it doesn't resolve the issue | 14:49 |
| eglynn__ | jaypipes: also the first patch fixes a kosher bug | 14:50 |
| eglynn__ | https://bugs.launchpad.net/glance/+bug/961050 | 14:50 |
| uvirtbot | Launchpad bug 961050 in glance "Creating bucket via region-specific S3 endpoint fails" [Medium,In progress] | 14:50 |
| *** danwent has joined #openstack-dev | 14:50 | |
| eglynn__ | (so I think we'd want https://review.openstack.org/5607 to go in regardless) | 14:51 |
| jaypipes | eglynn__: k, let me review that one separately rthen. | 14:52 |
| eglynn__ | jaypipes: cool | 14:52 |
| *** mdomsch has quit IRC | 14:53 | |
| *** n0ano has quit IRC | 14:53 | |
| *** GheRivero_ has joined #openstack-dev | 14:54 | |
| *** dtroyer_zzz is now known as dtroyer | 14:55 | |
| ttx | dolphm, heckj, bcwaldon: please have a look at https://review.openstack.org/#change,5628 -- will be needed to unfuck python-keystoneclient tarball generation before we do RC1 | 14:56 |
| jaypipes | mtaylor: and as for sqlalchemy and having it not auto-create foreign key constraints, I can't seem to find any way to do that, no. Looks like you would have to manually remove the FK from the database. | 15:00 |
| dolphm_ | ttx: looking | 15:01 |
| *** mattray has joined #openstack-dev | 15:03 | |
| eglynn__ | jaypipes: one quick'n'dirty way of validating https://review.openstack.org/5590 *without* merging would be to deliberately include a failing test | 15:03 |
| eglynn__ | self.assertFalse(True) or whatever | 15:03 |
| eglynn__ | jaypipes: then approve the patch, guaranteed that it will not be merged to trunk | 15:03 |
| eglynn__ | jaypipes: I could then re-trigger the s3-enabled Jenkins builds with that patch any number of times to validate that the intermittent failure no longer occurs | 15:03 |
| eglynn__ | jaypipes: a tad hacky, but ... | 15:04 |
| jaypipes | eglynn__: hmm, evil, but could work :) | 15:04 |
| eglynn__ | jaypipes: cool, I'll add a failing test so ... | 15:04 |
| jaypipes | eglynn__: yeah, sounds like a plan. do that and when jeblair gets in, have him set those env vars in the jenkins jobs? | 15:05 |
| dolphm_ | ttx: mtaylor: +2'd | 15:05 |
| eglynn__ | jaypipes: jeblair has already set up some alternative (non-gating) Jenkins jobs for me with the s3 config enabled | 15:06 |
| jaypipes | ah, that's right.. go ahead and approve it yourself then.. | 15:08 |
| jaypipes | when yo add that assertFalse | 15:08 |
| eglynn__ | eglynn: cool, will do ... one nit though, it depends on https://review.openstack.org/5607 so that would have to go in first | 15:09 |
| eglynn__ | jaypipes: cool, will do ... one nit though, it depends on https://review.openstack.org/5607 so that would have to go in first | 15:09 |
| jaypipes | eglynn__: k, lemme get on that right quick. | 15:10 |
| eglynn__ | jaypipes: cool, thanks! | 15:10 |
| *** alaski has quit IRC | 15:11 | |
| *** alaski has joined #openstack-dev | 15:11 | |
| sdague | anyone familiar with the ldap backend in keystone? It looks like there is an assumption that you have write access to the LDAP directory. Anyone doing any work in making LDAP only used on User (for corp directory), and the rest in the SQL backend? | 15:12 |
| jaypipes | sdague: might want to ping ayoung or Ryan Lane (can't remember Ryan's IRC nick...) | 15:13 |
| ayoung | sdague, that would be me | 15:13 |
| sdague | ayoung: ok great. I saw the one launchpad issue registered about it https://bugs.launchpad.net/keystone/+bug/928065, but it wasn't clear if there was actual motion there | 15:13 |
| uvirtbot | Launchpad bug 928065 in keystone "implement SQL+LDAP backend for Identity" [Medium,Confirmed] | 15:13 |
| ayoung | sdague, so I am not sure if it will be SQL back end for the other, or just split up Identity to allow mix and match | 15:14 |
| ayoung | I suspect that saying the Identity provider is composed of a user provider and a tenants/roles provider, and then specifying which to use for each will be the easier way to go | 15:15 |
| sdague | ayoung: right, anyway to split it up. Mostly I just need to take out the assumption that you can write back info in the LDAP server, and limit it to only User auth, because in many environments, you can't. | 15:15 |
| *** joesavak has joined #openstack-dev | 15:15 | |
| ayoung | sdague, yea, I am aware. | 15:15 |
| ayoung | sdague, one approach would be to deal with that at the LDAP level | 15:15 |
| ayoung | set up an OpenLDAP server that defers to, say, Active Directory for the user lists | 15:16 |
| ayoung | That should be do-able now | 15:16 |
| sdague | ayoung: yes, that's probably true. | 15:17 |
| ayoung | You won't be able to add users, but assigning them to roles and tenants should work | 15:17 |
| ayoung | sdague, if you have the motivation and setup to test it, please be sure to share your findings | 15:18 |
| sdague | ayoung: right, that's all I care about. Though I'd like to work on doing it the right way in keystone as well. Just to make it easier to setup. | 15:18 |
| sdague | ayoung: sure. | 15:18 |
| ayoung | sdague, define easier? | 15:18 |
| ayoung | heh | 15:18 |
| jaypipes | eglynn__: k, review on https://review.openstack.org/#change,5607 done | 15:19 |
| sdague | for instance, when I setup jenkins in our directory environment, I just have to provide suffix, filter, and ldap server, and it auths anyone in the company, though the users start off with basically no permissions. Then I can add roles to them in it's UI. | 15:20 |
| eglynn__ | jaypipes: thanks, will push an updated patch set in a couple minutes ... | 15:20 |
| jaypipes | eglynn__: cheers mate | 15:20 |
| sdague | just comparing to existing open source tools in how they drop into the directory without adding an openldap server :) | 15:20 |
| *** danwent has quit IRC | 15:21 | |
| *** Mandell has joined #openstack-dev | 15:23 | |
| eglynn__ | jaypipes: https://review.openstack.org/5607 updated | 15:30 |
| eglynn__ | bbiab ... | 15:30 |
| jaypipes | eglynn__: approved. | 15:32 |
| ttx | mtaylor: was talking to dtroyer about having a devstack branch that can be used to test Essex RCs | 15:35 |
| ttx | mtaylor: since master devstack points to master (Folsom) by default | 15:36 |
| ttx | mtaylor: could we add that to openstack-dev/devstack? | 15:36 |
| ttx | (I want to make sure it's easy and discoverable to test Essex RCs using devstack) | 15:37 |
| ttx | jeblair: ^ | 15:39 |
| *** Mandell has quit IRC | 15:40 | |
| eglynn__ | jaypipes: thanks | 15:48 |
| *** shevek_ has quit IRC | 15:49 | |
| jaypipes | ttx: you can already do that with devstack... just set the branches appropriately in the stackrc to point at milestone-proposed instead of the default of master. | 15:49 |
| jaypipes | ttx: now, I could propose a devstack branch called milestone-proposed that had those branches changed in stackrc? | 15:50 |
| ttx | jaypipes: sure, maybe only a good ML post reminding people to do that is enough | 15:50 |
| ttx | jaypipes: I want it to be as obvious as possible, so that people test the RC rather than Folsom :) | 15:50 |
| jaypipes | ttx: happy to do either one :) | 15:50 |
| jaypipes | ttx: I can do a blog with devstack + milestone-poposed + tempest. | 15:51 |
| dtroyer | jaypipes: also, this is a warmup to the essex branch of devstack that isn't going to evolve and not be able to run the essex releases…. | 15:51 |
| ttx | jaypipes: sounds good, I wanted to make sure mtaylor was supportive of the idea to push such a branch under the repo | 15:52 |
| ttx | also coordinate with dtroyer so as to not duplicate the work :) | 15:52 |
| jaypipes | dtroyer: yup, sounds good. | 15:52 |
| ttx | jaypipes: blog with devstack + milestone-poposed + tempest sounds good | 15:53 |
| ttx | I can relay it on the ML and Twitter and my own blogpost | 15:53 |
| jaypipes | ttx: I'll work on that today. | 15:54 |
| ttx | jaypipes: awesome | 15:54 |
| *** sandywalsh has quit IRC | 15:56 | |
| *** joesavak has quit IRC | 15:59 | |
| *** rnirmal has joined #openstack-dev | 16:04 | |
| *** joesavak has joined #openstack-dev | 16:04 | |
| *** danwent has joined #openstack-dev | 16:05 | |
| *** maplebed has joined #openstack-dev | 16:05 | |
| *** jsavak has joined #openstack-dev | 16:06 | |
| *** deshantm_ has joined #openstack-dev | 16:08 | |
| *** fc__ has quit IRC | 16:08 | |
| *** joesavak has quit IRC | 16:09 | |
| *** sandywalsh has joined #openstack-dev | 16:11 | |
| jeblair | ttx, dtroyer, jaypipes: so it sounds like getting people to test RCs via jay's blog post is the way you're going, i think that's best. | 16:11 |
| *** deshantm has quit IRC | 16:12 | |
| jeblair | we _can_ create a milestone-proposed branch of devstack, if that's needed. sounds a little heavyweight though since devstack is already designed for this | 16:12 |
| annegentle | jeblair: can I have a link to Jay's blog post? | 16:12 |
| jeblair | once essex is released, i assume we'll be making stable/essex branches of everything, including devstack. | 16:13 |
| * annegentle one day will build a killer doc system on Wordpress | 16:13 | |
| jeblair | annegentle: i think he's writing it today | 16:13 |
| annegentle | jeblair: the devstack README says you could get essex3 but that branch doesn't exist. | 16:13 |
| annegentle | jeblair: https://github.com/openstack-dev/devstack <-- that one | 16:14 |
| jaypipes | annegentle: writing it today. | 16:14 |
| annegentle | jaypipes: go go go :) | 16:14 |
| jaypipes | now that dtroyer approved vish's ratelimit disable patch, things will be much nicer :) | 16:15 |
| jeblair | annegentle: weird, i don't think that's correct. :/ | 16:15 |
| dtroyer | jeblair: yes, that's the plan (stable/essex branch of devstack) | 16:16 |
| annegentle | jeblair: was it ever? | 16:16 |
| dtroyer | annegentle: I'll fix that up…it was meant to be an example | 16:17 |
| annegentle | dtroyer: ok, thanks | 16:17 |
| *** vincentricci has joined #openstack-dev | 16:19 | |
| *** dalang has joined #openstack-dev | 16:20 | |
| jeblair | annegentle: i'd like to file a bug about my keystone docs question from yesterday. what's the best way to do that? file it on keystone? should i target any other projects (openstack-manuals or anything)? | 16:21 |
| jeblair | https://bugs.launchpad.net/openstack-manuals/+bug/933547 | 16:22 |
| uvirtbot | Launchpad bug 933547 in openstack-manuals "api-site FlavorRef and ImageRef are not documented as parameters when creating a server" [High,Confirmed] | 16:22 |
| jeblair | annegentle: ^ that looks similar, actually; maybe it should just be a bug on openstack-manuals? | 16:23 |
| annegentle | jeblair: yep, for api docs you can log against openstack-manuals | 16:23 |
| jeblair | annegentle: cool, thanks | 16:23 |
| annegentle | jeblair: I never made a separate identity-api launchpad project (nor compute-api, nor object-api, nor netconn-api, nor image-api, you get the picture) | 16:24 |
| jeblair | gotcha | 16:24 |
| annegentle | jeblair: I suppose we could but it's ok to track in openstack-manuals | 16:24 |
| annegentle | jeblair: tag with identity-api | 16:24 |
| *** deshantm_ is now known as deshantm | 16:24 | |
| annegentle | jeblair: and assign to heckj :) | 16:26 |
| *** Mandell has joined #openstack-dev | 16:26 | |
| *** danwent has quit IRC | 16:26 | |
| annegentle | jeblair: ok maybe I should ask him before assigning it, but basically the PTL has to manage their APIs docs | 16:26 |
| *** nati has joined #openstack-dev | 16:26 | |
| jeblair | annegentle: he can always reassign it. :) | 16:27 |
| *** bepernoot has quit IRC | 16:28 | |
| *** heckj has joined #openstack-dev | 16:28 | |
| *** popux has joined #openstack-dev | 16:28 | |
| annegentle | jeblair: tru dat | 16:29 |
| *** reed has joined #openstack-dev | 16:29 | |
| chmouel | vishy: is posting to /{ec2,s3}tokens on keystone admin-url/port need a X-Auth-Token/Username-Password ? it doesn't in the code but was wondering if it should | 16:30 |
| jeblair | annegentle: how should i reference the existing doc? a URL, or section number (3.2.1.1), or...? | 16:30 |
| annegentle | jeblair: I prefer URLs as they shouldn't change | 16:30 |
| annegentle | jeblair: in API docs the section numbers are output at build time, so they could change with insertions/deletions | 16:31 |
| *** jdurgin has joined #openstack-dev | 16:31 | |
| vishy | chmouel: hmm, it probably should | 16:31 |
| jeblair | http://docs.openstack.org/api/openstack-identity-service/2.0/content/POST_authenticate_v2.0_tokens_Admin_API_Service_Developer_Operations-d1e1356.html | 16:31 |
| jeblair | annegentle: okay, but that ^ url won't change? | 16:31 |
| vishy | chmouel: although I don't know if it matters | 16:31 |
| jeblair | (it's got some autogenerated looking stuff on it...) | 16:31 |
| ttx | heckj: yo | 16:31 |
| annegentle | jeblair: it shouldn't because the xml source contains an ID that creates the html file name | 16:31 |
| heckj | ttx: mornin - just checkin in. Looking at our state right now | 16:32 |
| ttx | heckj: just added as a blocker: bug 961354 | 16:32 |
| uvirtbot | Launchpad bug 961354 in keystone "python-keystoneclient job fails to build a tarball" [Critical,In progress] https://launchpad.net/bugs/961354 | 16:32 |
| *** reidrac has quit IRC | 16:32 | |
| jeblair | annegentle: groovy, thanks. | 16:32 |
| ttx | heckj: needs https://review.openstack.org/#change,5628 approved to fix | 16:32 |
| *** johan_-_ has quit IRC | 16:32 | |
| LinuxJedi | ok, so the review site and jenkins site have had a makeover today. Please let me know (or file a CI bug) if anything looks out of place | 16:32 |
| heckj | soren: ping, you around? | 16:33 |
| heckj | holy crap, what happened to gerrit? | 16:33 |
| ttx | heckj: otherwise we could consider we are good to go if we dismiss the create/delete users bug from soren | 16:33 |
| chmouel | vishy: okay thanks will remove that from s3_token | 16:33 |
| cp16net | markmc: you around? | 16:33 |
| LinuxJedi | heckj: what I just said ;) | 16:33 |
| * jk0 is a fan of the new gerrit style | 16:34 | |
| * ttx got used to lightgreen | 16:34 | |
| chmouel | LinuxJedi: nice :) | 16:34 |
| chmouel | LinuxJedi: it's pretty fast | 16:34 |
| *** jerdfelt has joined #openstack-dev | 16:34 | |
| ttx | and slimyyellow | 16:34 |
| LinuxJedi | nice, but the logo is missing :/ | 16:34 |
| jeblair | LinuxJedi: yeah, i was about to mention that. :) | 16:34 |
| *** PotHix has joined #openstack-dev | 16:34 | |
| LinuxJedi | jeblair: that could be because puppet bailed half way through due to that thing that just got pushed through | 16:35 |
| ttx | LinuxJedi: some things appear a bit too large IMO, like names | 16:35 |
| ttx | LinuxJedi: will file bugs | 16:35 |
| LinuxJedi | ttx: awesome, thanks :) | 16:35 |
| jk0 | LinuxJedi: any chance you could change the red links to #600? | 16:35 |
| jeblair | General reminder: this is all in git, so anyone can hack on it and submit patches. :) | 16:35 |
| ttx | yeah, we could theme it on specific days like GOOG does | 16:36 |
| jk0 | #600 is slightly darker and easieron the eyes | 16:36 |
| jeblair | https://github.com/openstack/openstack-ci-puppet/tree/master/modules/gerrit/files | 16:36 |
| jeblair | jk0: i agree, darker red would be nice | 16:37 |
| jeblair | but there's a logo that i think he's matching, so could be complicated. :) | 16:37 |
| LinuxJedi | jeblair: grr... review puppet doesn't support pip either :( | 16:37 |
| jk0 | #600 should match the logo if memory serves. it's my goto color :) | 16:37 |
| *** buterfly has joined #openstack-dev | 16:38 | |
| annegentle | I think OpenStack red is #CF2F19 | 16:39 |
| *** shall_ has quit IRC | 16:39 | |
| annegentle | or #BC1518 hm. | 16:40 |
| annegentle | The CF one looks too orange to me | 16:40 |
| *** buterfly has quit IRC | 16:40 | |
| *** spiffxp has quit IRC | 16:42 | |
| *** devananda has joined #openstack-dev | 16:42 | |
| *** jerdfelt1 has joined #openstack-dev | 16:43 | |
| *** jerdfelt has quit IRC | 16:43 | |
| *** jerdfelt1 is now known as jerdfelt | 16:43 | |
| ttx | LinuxJedi: bug 961367 | 16:45 |
| uvirtbot | Launchpad bug 961367 in openstack-ci "Some things appear too large in gerrit new style" [Undecided,New] https://launchpad.net/bugs/961367 | 16:45 |
| LinuxJedi | ttx: excellent, thanks. Will try and work on it today (might be tomorrow morning now) | 16:46 |
| ttx | LinuxJedi: not really a top urgency :) | 16:47 |
| heckj | mtaylor: what happened to test-requires in https://review.openstack.org/#change,5628? | 16:47 |
| dolphm_ | ahh fancy UI's popping up everywhere | 16:49 |
| heckj | monty: did "tox" or "run_tests.sh" work for you with https://review.openstack.org/#change,5628, I can't get either to run through completion | 16:50 |
| LinuxJedi | well, we figured if people need to look at it all day it should be a bit more pleasing to the eye, and more consistent with the Openstack website | 16:50 |
| heckj | mtaylor: ^^ | 16:50 |
| LinuxJedi | jk0: the red links can be changed, I just used the red from Openstack's site | 16:50 |
| jakedahn | who styled review.github.com / | 16:50 |
| jakedahn | er review.openstack.org | 16:50 |
| dolphm_ | LinuxJedi? | 16:51 |
| apevec | dolphm_, https://review.openstack.org/#patch,unified,5358,8,keystone/catalog/core.py | 16:51 |
| apevec | where is catalog_api.service_exists ? | 16:51 |
| jakedahn | ah just read teh scrollback - but I must say it is certainly not pleasing on the eye - and I think it should be reverted to the original gerrit theme | 16:52 |
| *** spiffxp has joined #openstack-dev | 16:52 | |
| apevec | yeah too much red | 16:52 |
| jakedahn | and the table alternating rows is not as easily apparent as it was before | 16:52 |
| jakedahn | and the review menu has moved, which sucks for a tool that people are used to | 16:53 |
| jk0 | jakedahn: tell us what you really think :) | 16:53 |
| dragondm | Also, the checkmark next to seen files had gone away :( | 16:53 |
| jakedahn | jk0: AHHHHHHHHHHHHHHHHHHHHHH | 16:53 |
| jakedahn | jk0: btw moving to sf? | 16:53 |
| jk0 | nah, staying here in lovely WI :) | 16:53 |
| jakedahn | 'lovely' :) | 16:54 |
| *** gyee has joined #openstack-dev | 16:54 | |
| jk0 | it's beautiful this time of year: http://jk0.org/tmp/hazelgreen-2.jpg (taken yesterday) | 16:54 |
| jakedahn | jk0: yeah I'm from minnesota, spring and fall rock, as for winter and summer... | 16:55 |
| jk0 | yeah | 16:55 |
| dprince | So. I used to theme SmokeStack in 'OpenStack red'... but I found that all that red just made people really mad all the time... at least I thought that was why people were always mad at me. | 16:55 |
| jakedahn | yeah I'm going to start raging | 16:55 |
| jeblair | dprince: based on that, i asked LinuxJedi to use blue for the jenkins theme | 16:56 |
| jeblair | https://jenkins.openstack.org | 16:56 |
| jk0 | I like that shade of blue | 16:56 |
| * dprince can't afford bodyguards! | 16:56 | |
| jakedahn | it works on jenkins fairly well | 16:56 |
| * LinuxJedi will is considering switching to blue on Gerrit when I update that skin | 16:56 | |
| jakedahn | also where do i push fixes to for gerrit? | 16:57 |
| jakedahn | is it https://github.com/openstack/openstack-ci-puppet/tree/master/modules/gerrit/files ? | 16:57 |
| jeblair | jakedahn: yes | 16:57 |
| jeblair | it's managed just like the rest of the projects, 'git review' and everything | 16:57 |
| jakedahn | there's a very large chance that I'm going to make a pass at it later today | 16:58 |
| *** sandywalsh has quit IRC | 16:58 | |
| jeblair | LinuxJedi, jakedahn: https://bugs.launchpad.net/openstack-ci/+bug/961394 | 16:59 |
| uvirtbot | Launchpad bug 961394 in openstack-ci "Reviewed checkmack in gerrit is missing" [High,New] | 16:59 |
| jeblair | heh, checkmark even. :) | 17:00 |
| *** vincentricci has quit IRC | 17:00 | |
| *** vincentricci has joined #openstack-dev | 17:00 | |
| *** devananda has quit IRC | 17:00 | |
| dragondm | ^ Yah, good. was abt to file same bug. | 17:00 |
| uvirtbot | dragondm: Error: "Yah," is not a valid command. | 17:00 |
| * dragondm thwaps bot. | 17:00 | |
| LinuxJedi | jeblair: damn, I'll jump on that shortly | 17:01 |
| * jaypipes hugs LinuxJedi, jeblair and mtaylor for the awesome skin overhaul in Gerrit. | 17:01 | |
| jaypipes | my eyes no longer bleed. | 17:01 |
| jeblair | LinuxJedi did all the work. We just reviewed it. :) | 17:02 |
| jeblair | (and as it turns out, we missed some things in the review. :) | 17:03 |
| LinuxJedi | also filed a bug about the link color, will fix that in the same patch | 17:03 |
| jakedahn | speaking of unifying opensatcky design - on the last docday i put this together: http://jakedahn.com/os-example/ and pushed the changes to horizon as a first step - but the jenkins builder sort of hated the changes and it now looks like this http://horizon.openstack.org/ | 17:05 |
| jakedahn | jeblair: is there any chance you could take a look at the jenkins job / sphinx builder and figure out why its not working? | 17:06 |
| *** nati has quit IRC | 17:06 | |
| jeblair | jakedahn: sure i'll take a look. i like yours better. :) | 17:06 |
| dolphm_ | http://i.imgur.com/9hpbi.png new H1 font, better cell padding, subdued link colors, margin around search box, more padding on search button | 17:06 |
| jakedahn | dolphm_: shipit! | 17:07 |
| *** devananda has joined #openstack-dev | 17:09 | |
| jeblair | jakedahn: nature.css doesn't seem to be in the horizon repo. | 17:10 |
| *** Mandell has quit IRC | 17:11 | |
| *** jakedahn is now known as jakedahn_zz | 17:11 | |
| *** maploin has quit IRC | 17:13 | |
| *** sandywalsh has joined #openstack-dev | 17:13 | |
| dolphm_ | apevec: you're right, i'm probably crazy (or something funky happened merging) -- i'm still looking at turning 500's into 404's, so this is very much on my radar | 17:14 |
| apevec | dolphm_, ok, btw that renders sql catalog backend unusable since I can't create endpoints | 17:14 |
| *** devananda has quit IRC | 17:15 | |
| dolphm_ | can you file a bug, and i'll put in a quick fix for just that? i'd deem that rc1 critical | 17:15 |
| apevec | unfortunately both devstack and smokestack use templated backedn only | 17:15 |
| apevec | yep, I'd vote rc1 | 17:15 |
| *** devananda has joined #openstack-dev | 17:16 | |
| dolphm_ | apevec: link me the bug when you have it, i'm putting together a fix | 17:16 |
| apevec | yep, filing now | 17:16 |
| *** oneiroi has quit IRC | 17:17 | |
| apevec | dolphm_, bug 961412 | 17:21 |
| uvirtbot | Launchpad bug 961412 in keystone "endpoint-create fails" [Undecided,New] https://launchpad.net/bugs/961412 | 17:21 |
| dolphm_ | apevec: thanks | 17:21 |
| *** bepernoot has joined #openstack-dev | 17:21 | |
| apevec | looks like setting milestone is privileged operation | 17:22 |
| dolphm_ | apevec: got it | 17:22 |
| apevec | ok, thanks! | 17:22 |
| *** bepernoot has quit IRC | 17:27 | |
| *** popux has quit IRC | 17:29 | |
| heckj | dolphm_ thanks for marking that in | 17:32 |
| *** Ryan_Lane has joined #openstack-dev | 17:32 | |
| *** Ryan_Lane has joined #openstack-dev | 17:32 | |
| heckj | I think we've got another one that needs to be RC1 as well: https://bugs.launchpad.net/keystone/+bug/958135 | 17:34 |
| uvirtbot | Launchpad bug 958135 in keystone "UnboundLocalError: local variable 'metadata_ref' referenced before assignment" [High,Triaged] | 17:34 |
| heckj | chmouel: ping | 17:35 |
| chmouel | heckj: pong | 17:42 |
| chmouel | heckj: was trying to read bug 958135 | 17:42 |
| uvirtbot | Launchpad bug 958135 in keystone "UnboundLocalError: local variable 'metadata_ref' referenced before assignment" [Critical,Triaged] https://launchpad.net/bugs/958135 | 17:42 |
| heckj | chmouel: just sent you some email - need your assistance with taht bug - yeah | 17:42 |
| heckj | The environment setup that Kevin generated wasn't super clear, but the traceback was what made me specifically concerned. Can you take a stab at it? | 17:43 |
| chmouel | heckj: cool I probably would need to get some more info from kevin but i can do that on GMT time so that should be good | 17:43 |
| heckj | sweet, thank you! | 17:43 |
| chmouel | no problem | 17:44 |
| LinuxJedi | ok guys, gerrit style has been updated slightly to fix a few things | 17:45 |
| *** novas0x2a|laptop has joined #openstack-dev | 17:47 | |
| dolphm_ | heckj: before this week, we had almost zero coverage for catching what should be 404's -- looks like another case of that in that metadata_ref bug | 17:48 |
| *** Ryan_Lane has quit IRC | 17:48 | |
| *** zigo has quit IRC | 17:49 | |
| *** devananda has quit IRC | 17:52 | |
| *** dneary has quit IRC | 17:55 | |
| *** devananda has joined #openstack-dev | 17:57 | |
| *** alaski has quit IRC | 17:59 | |
| *** alaski has joined #openstack-dev | 18:00 | |
| *** joesavak has joined #openstack-dev | 18:03 | |
| *** jsavak has quit IRC | 18:04 | |
| *** danwent has joined #openstack-dev | 18:06 | |
| LinuxJedi | dprince: ok, so gerrit is using the same blue as Jenkins now for the links. ttx: the fonts sizes should be a little better now. dragondm: checkmark fixed | 18:06 |
| *** alaski_ has joined #openstack-dev | 18:07 | |
| * LinuxJedi also made a tweak to Jenkins breadcrumbs to make scrolling work better | 18:07 | |
| LinuxJedi | and now I'm going to rest for a bit :) | 18:07 |
| *** joesavak has quit IRC | 18:08 | |
| *** johnpostlethwait has joined #openstack-dev | 18:08 | |
| *** darraghb has quit IRC | 18:08 | |
| *** danwent has quit IRC | 18:08 | |
| dprince | LinuxJedi: excellent. My blood pressure is dropping... slowly. | 18:08 |
| LinuxJedi | lol :) | 18:08 |
| *** alaski has quit IRC | 18:09 | |
| *** derekh has quit IRC | 18:11 | |
| *** jaypipes has quit IRC | 18:13 | |
| *** rkukura has quit IRC | 18:13 | |
| dolphm_ | apevec: heckj: endpoint crud fix https://review.openstack.org/#change,5636 | 18:14 |
| *** GheRivero_ has quit IRC | 18:14 | |
| *** jaypipes has joined #openstack-dev | 18:19 | |
| *** bhall has joined #openstack-dev | 18:19 | |
| *** bhall has quit IRC | 18:19 | |
| *** bhall has joined #openstack-dev | 18:19 | |
| *** alaski_ has quit IRC | 18:19 | |
| *** alaski has joined #openstack-dev | 18:20 | |
| *** bepernoot has joined #openstack-dev | 18:25 | |
| *** bepernoot has quit IRC | 18:27 | |
| *** dolphm_ has quit IRC | 18:28 | |
| *** heckj has quit IRC | 18:29 | |
| *** refactor has quit IRC | 18:29 | |
| eglynn__ | jaypipes: https://review.openstack.org/5590 is looking good for that S3 issue bug #958568 | 18:32 |
| uvirtbot | Launchpad bug 958568 in glance "NotAuthorized: You are not authorized to complete this action" [Undecided,New] https://launchpad.net/bugs/958568 | 18:32 |
| eglynn__ | jaypipes: re-triggered Jenkins builds 10 times concurrently in py26 & py27 and no sign of the old 404 | 18:32 |
| eglynn__ | jaypipes: I'm gonna push an updated patch set without the deliberately failing test for proper approval this time | 18:32 |
| eglynn__ | ... pushed | 18:32 |
| eglynn__ | wrong bug, meant #955527 | 18:33 |
| *** mnewby has joined #openstack-dev | 18:37 | |
| jaypipes | eglynn__: well I'll be... | 18:37 |
| jaypipes | eglynn__: crazy, huh? | 18:37 |
| eglynn__ | jaypipes: yep, should have applied Occam's razor much earlier ... | 18:37 |
| jaypipes | lol indeed :) | 18:38 |
| eglynn__ | better slow than never I guess ... | 18:38 |
| *** dachary has quit IRC | 18:38 | |
| jaypipes | eglynn__: :) | 18:38 |
| eglynn__ | at least we got that glance-logcapture out of it, I've been wanting something of that ilk for a good while ... | 18:39 |
| *** n0ano has joined #openstack-dev | 18:40 | |
| *** dolphm has joined #openstack-dev | 18:40 | |
| *** jakedahn_zz is now known as jakedahn | 18:41 | |
| jaypipes | eglynn__: indeed :) | 18:45 |
| *** villep has quit IRC | 18:48 | |
| *** fesnel has quit IRC | 18:48 | |
| *** fesnel has joined #openstack-dev | 18:48 | |
| *** villep has joined #openstack-dev | 18:50 | |
| *** deshantm has quit IRC | 18:50 | |
| *** litejk has joined #openstack-dev | 18:51 | |
| *** litejk has left #openstack-dev | 18:51 | |
| *** dolphm has quit IRC | 18:52 | |
| *** heckj has joined #openstack-dev | 18:53 | |
| *** Kiall has quit IRC | 18:54 | |
| *** dachary has joined #openstack-dev | 19:00 | |
| *** Kiall has joined #openstack-dev | 19:01 | |
| *** Ryan_Lane has joined #openstack-dev | 19:04 | |
| *** novas0x2a|laptop has quit IRC | 19:06 | |
| *** novas0x2a|laptop has joined #openstack-dev | 19:06 | |
| *** dtroyer is now known as dtroyer_zzz | 19:07 | |
| *** oneiroi has joined #openstack-dev | 19:08 | |
| *** devananda has quit IRC | 19:10 | |
| *** devananda has joined #openstack-dev | 19:14 | |
| *** dbuzz has joined #openstack-dev | 19:15 | |
| *** danwent has joined #openstack-dev | 19:18 | |
| *** dbuzz has quit IRC | 19:18 | |
| *** dbuzz has joined #openstack-dev | 19:18 | |
| *** dtroyer_zzz is now known as dtroyer | 19:19 | |
| *** oneiroi has quit IRC | 19:22 | |
| *** dbuzz is now known as oneiroi | 19:22 | |
| *** danwent has quit IRC | 19:24 | |
| *** oneiroi is now known as Oneiroi | 19:24 | |
| *** dolphm has joined #openstack-dev | 19:25 | |
| *** rkukura has joined #openstack-dev | 19:26 | |
| jk0 | LinuxJedi: any chance that <h1>Code Review</h1> header can be hidden? | 19:26 |
| jk0 | sticks out like a sore thumb | 19:26 |
| sdague | dtroyer: on the stack.sh apache patch, should I just respin based on your suggestions, or do you have a commit somewhere that can just be cherry picked in? | 19:29 |
| *** bepernoot has joined #openstack-dev | 19:30 | |
| soren | heckj: I'll be around in a couple of hours if you still need me. | 19:31 |
| *** shevek_ has joined #openstack-dev | 19:31 | |
| *** bepernoot has quit IRC | 19:33 | |
| dtroyer | sdague: it is part of https://review.openstack.org/4364 which is a good-sized change to handle Fedora differences. But you can see what I mean by the name changes | 19:35 |
| dtroyer | I just pushed a rebase up so it should be current again | 19:35 |
| ayoung | termie, heckj, I have (I think) re applied the patch for Client Certs. What would be the right path to go around making tests for it? I assume I would want to run a few of the current tests, but to indicate that they should use SSL. | 19:37 |
| notmyname | any chance florian haas is in here? | 19:37 |
| sdague | dtroyer: ok, you want to take over the launchpad bug? and just have it go in as part of the bigger changes? I started working on that just because it was an easy unclaimed launchpad bug. | 19:38 |
| *** bcwaldon has quit IRC | 19:39 | |
| *** mdrnstm has joined #openstack-dev | 19:40 | |
| dtroyer | sdague: yeah, I should have claimed the bug…sorry for the confusion | 19:40 |
| sdague | no worries | 19:40 |
| jk0 | .c | 19:40 |
| comstud | how can I change the gerrit colors back? | 19:40 |
| comstud | :) | 19:40 |
| *** bcwaldon has joined #openstack-dev | 19:41 | |
| heckj | ayoung: ideally you could subclass the keystoneclient test cases (assuming I've got it right) and use that suite, which works the whole API. | 19:41 |
| ayoung | heckj, that makes sense. I'll give it a try | 19:41 |
| heckj | kk | 19:42 |
| *** bcwaldon has quit IRC | 19:46 | |
| *** asalkeld has joined #openstack-dev | 19:48 | |
| LinuxJedi | jk0, comstud: file bugs or edit the files in the repo | 19:51 |
| jk0 | "contributions welcome"? :) | 19:51 |
| LinuxJedi | yep, it is an openstack project after all, and goes through gerrit just like everything else | 19:52 |
| *** GheRivero_ has joined #openstack-dev | 19:52 | |
| LinuxJedi | comstud: although I do question your sanity if you liked the original Gerrit colour scheme | 19:56 |
| comstud | well | 19:56 |
| comstud | i'm partially colorblind... and the new background colors almost looks completely white to me | 19:57 |
| comstud | which is not easy on the eyes | 19:57 |
| LinuxJedi | comstud: ah! :) | 19:57 |
| comstud | looks like...white... and maybe... a very very light aqua ? | 19:57 |
| comstud | i dunno, but it all blends together for me | 19:57 |
| LinuxJedi | comstud: pretty much, yes. Same as the main openstack site | 19:57 |
| comstud | ah | 19:57 |
| *** vincentricci has quit IRC | 20:01 | |
| mtaylor | heckj: weird ... I don't understand why that patch isn't working for you | 20:03 |
| heckj | mtaylor: yeah, confused me too. | 20:04 |
| heckj | pip-requires doesn't have any of the project deps anymore in that patch - that expected? Moved elsewhere that I didn't see? | 20:04 |
| mtaylor | heckj: it has all the ones it did before the patch? | 20:05 |
| mtaylor | heckj: (this is keystoneclient, not keystone) | 20:05 |
| heckj | uh, hm. | 20:05 |
| heckj | It might be that I'm a moron. | 20:05 |
| heckj | just a sec, checking | 20:05 |
| *** bcwaldon has joined #openstack-dev | 20:06 | |
| heckj | mtaylor: yep, it's official. I'm a moron. | 20:06 |
| heckj | I applied that to keystone, which obviously fucked things up - not keystoneclient | 20:06 |
| LinuxJedi | heckj: we all do it at some point, it is an indicator that you need more coffee ;) | 20:07 |
| heckj | LinuxJedi - just started in on a new latte, as it happens... | 20:07 |
| LinuxJedi | or beer if you are trying to reach the Balmer Peak | 20:07 |
| LinuxJedi | http://xkcd.com/323/ | 20:08 |
| heckj | LinuxJedi: I prefer scotch or bourbon for that cycle. | 20:09 |
| * LinuxJedi tried a Vanilla Latte with some Oreos for dunking earlier, actually rather nice | 20:09 | |
| LinuxJedi | heckj: lol :) | 20:09 |
| mtaylor | heckj: I can haz positive review now? | 20:09 |
| heckj | mtaylor: just flipped the bit | 20:09 |
| mtaylor | heckj: yay! | 20:09 |
| *** bcwaldon has quit IRC | 20:10 | |
| *** bcwaldon has joined #openstack-dev | 20:11 | |
| *** shevek_ has quit IRC | 20:12 | |
| andrewbogott | Hm... Rick Harris (irc handle unknown) has just advised me to name a method 'list' in a code review. Doesn't that violate the rule against using the same name as a builtin? | 20:19 |
| *** vricci has joined #openstack-dev | 20:19 | |
| *** bcwaldon has quit IRC | 20:19 | |
| mtaylor | andrewbogott: um, yeah - making a method called list seems like a bad idea | 20:20 |
| andrewbogott | hm | 20:20 |
| andrewbogott | Ok, well, I'll leave it as is then :) | 20:20 |
| *** bcwaldon has joined #openstack-dev | 20:21 | |
| *** vricci is now known as vincentricci | 20:29 | |
| *** dprince has quit IRC | 20:30 | |
| *** bcwaldon has quit IRC | 20:34 | |
| *** Glacee has joined #openstack-dev | 20:37 | |
| andrewbogott | Vek: Does that guideline about string encoding strike you as common knowledge or obscure? Wondering if I should/need to schedule a summit session to proselytize. | 20:43 |
| jeblair | mtaylor, andrewbogott: not being contrary, just curious: is that really intended to apply to _methods_, which will always be called as "foo.list()"? I realize the HACKING rule says "anything", but there are many "def list(self)"'s already in novaclient. | 20:56 |
| andrewbogott | jeblair: I don't feel all that strongly, although it annoys me when methods get lit up in the wrong color by my syntax highlighter. | 20:58 |
| *** crobinso has quit IRC | 20:58 | |
| jeblair | point. :) | 20:58 |
| *** jdg has joined #openstack-dev | 20:59 | |
| *** bcwaldon has joined #openstack-dev | 21:02 | |
| *** anotherjesse_zz is now known as anotherjesse | 21:04 | |
| *** anotherjesse is now known as anotherjesse_zz | 21:07 | |
| *** littleidea has joined #openstack-dev | 21:07 | |
| *** anotherjesse_zz is now known as anotherjesse | 21:08 | |
| *** lts has quit IRC | 21:10 | |
| *** bepernoot has joined #openstack-dev | 21:10 | |
| *** alaski has quit IRC | 21:10 | |
| *** andrewsmedina has quit IRC | 21:11 | |
| *** andrewsmedina has joined #openstack-dev | 21:13 | |
| *** novas0x2a|lapto1 has joined #openstack-dev | 21:14 | |
| *** bcwaldon has quit IRC | 21:14 | |
| *** novas0x2a|laptop has quit IRC | 21:14 | |
| ayoung | heckj, so to make the connections work, it is going to take changes to keystoneclient, too. Which I guess shouldn't be that surprising. Not sure how to get the certififace and key passed from the config files through to the client object | 21:15 |
| *** bepernoot has quit IRC | 21:17 | |
| notmyname | ttx: swift 1.4.8 is good to go | 21:18 |
| *** jdg has quit IRC | 21:19 | |
| *** stuntmachine has quit IRC | 21:19 | |
| *** jdg has joined #openstack-dev | 21:20 | |
| *** Oneiroi is now known as Oneiroi|Zzzz | 21:20 | |
| *** pmyers has quit IRC | 21:22 | |
| *** pmyers has joined #openstack-dev | 21:22 | |
| *** danwent has joined #openstack-dev | 21:25 | |
| ttx | notmyname: will do, in my tomorrow morning | 21:28 |
| notmyname | ttx: thanks | 21:28 |
| notmyname | ttx: anything special on my end this time (since it's the essex release)? | 21:28 |
| ttx | notmyname: not really. I'll push Final=True to the milestone-proposed branch, but that's business as usual | 21:29 |
| notmyname | ok, thanks | 21:29 |
| ttx | LinuxJedi: looks good ! | 21:29 |
| LinuxJedi | ttx: great :) | 21:31 |
| *** mattstep has quit IRC | 21:35 | |
| *** mattstep has joined #openstack-dev | 21:36 | |
| *** markmc has quit IRC | 21:38 | |
| *** lloydde has joined #openstack-dev | 21:39 | |
| *** GheRivero_ has quit IRC | 21:41 | |
| *** andrewsmedina has quit IRC | 21:42 | |
| dolphm | why is ubuntu packaging the keystone SQL catalog backend instead of the flat-file TemplatedCatalog? i thought sql.Catalog was supposed to be undocumented for essex? it's not even tested afaik | 21:47 |
| apevec | dolphm, you would recommend templated for production? | 21:49 |
| dolphm | Madkiss: i thought that was our official recommendation, yes | 21:49 |
| apevec | I'm using sql catalog backend in fedora | 21:50 |
| apevec | seemed more "production"ish | 21:50 |
| dolphm | apevec: the flat file came first, and is much more thoroughly tested at this point -- that's about all i can say | 21:51 |
| dolphm | apevec: it's a much simpler implementation | 21:51 |
| anotherjesse | apevec: plus the sql catalog makes a lot of assumptions about the relationships between users/tenants/endpoints | 21:51 |
| adam_g | dolphm: we're testing the sql backend quite thooroughly | 21:51 |
| anotherjesse | that might be ok for smaller clouds, but isn't necessarily true for larger deploys | 21:51 |
| *** bcwaldon has joined #openstack-dev | 21:51 | |
| anotherjesse | we see situations where different tenants are given different endpoints based on their roles, ... | 21:52 |
| dolphm | adam_g: link? | 21:52 |
| anotherjesse | (dev vs prod) | 21:52 |
| *** dspano has quit IRC | 21:52 | |
| apevec | but last I tried it, templated wasn't persisting any data, so you had to re-run keystone service-create iirc ? | 21:52 |
| adam_g | anotherjesse: and those relationships are expressed in the templated file? | 21:52 |
| apevec | after every keystone restart | 21:52 |
| *** bcwaldon has quit IRC | 21:53 | |
| adam_g | dolphm: https://jenkins.qa.ubuntu.com/view/Precise%20OpenStack%20Testing/ | 21:53 |
| *** bcwaldon has joined #openstack-dev | 21:53 | |
| jeblair | bcwaldon: https://answers.launchpad.net/nova/+question/185110 | 21:54 |
| jeblair | bcwaldon: is that still the latest thinking on that issue? | 21:54 |
| jeblair | bcwaldon: becaues the docs suggest that one would expect to see two networks public/private returned | 21:55 |
| jeblair | http://docs.openstack.org/api/openstack-compute/2/content/Get_Server_Details-d1e2623.html | 21:55 |
| *** dolphm has quit IRC | 21:55 | |
| *** apevec has quit IRC | 22:00 | |
| adam_g | anotherjesse: im curious, the ability to model those relationships is a feature TBD at the summit and planned for F? or is this something expected of the templated on essex? | 22:08 |
| *** flaviamissi has quit IRC | 22:11 | |
| *** dtroyer is now known as dtroyer_zzz | 22:14 | |
| bcwaldon | jeblair: the docs use public/private as examples | 22:17 |
| bcwaldon | jeblair: its not a part of the api itself | 22:18 |
| jeblair | bcwaldon: okay that makes sense... | 22:18 |
| jeblair | bcwaldon: the issue i'm concerned about is how should people expect to get that information? libcloud expects two different networks, and i know of at least one cloud provider that uses two different networks. :) | 22:19 |
| jeblair | but i also know of another provider that has the private and public ip on the same network | 22:19 |
| jeblair | so it looks like the question that was in launchpad... | 22:19 |
| jeblair | bcwaldon: http://paste.openstack.org/show/11760/ | 22:20 |
| jeblair | there's what i get when i talk to two different openstack providers | 22:20 |
| cloudfly | honestly nova will need to scale to handle n numbers of WAN | 22:20 |
| cloudfly | if only to handle security profiles | 22:20 |
| jeblair | so how should we expect people to write code to handle that? | 22:20 |
| jeblair | a little provider-specific routine to figure out which ip address is public? | 22:21 |
| cloudfly | i think it will ultimately need to be a function of the API for talking to the network controller | 22:21 |
| bcwaldon | jeblair: you might be able to use the floating ips extension | 22:21 |
| jeblair | or is there a best practice for how providers should set that up? ie, openstack providers are expected to have two networks like that... | 22:21 |
| jeblair | bcwaldon: yes, i imagine that would work | 22:22 |
| bcwaldon | jeblair: I think it would make sense to deploy this way, to have a specific network that is public | 22:22 |
| bcwaldon | jeblair: but if your fixed ips are always private and floating always public, that model wont work | 22:22 |
| *** rods has quit IRC | 22:25 | |
| jeblair | bcwaldon: so it sounds like it boils down to private and public are abstract concepts, and nova isn't going to dictate that, so you'll need some provider-specific knowledge to work it out. | 22:25 |
| jeblair | bcwaldon: but it sure would be handy if most public cloud providers had two networks they called private and public. :) | 22:25 |
| bcwaldon | jeblair: yep | 22:25 |
| cloudfly | i disagree | 22:26 |
| cloudfly | it's bad terminology | 22:26 |
| *** kbringard has quit IRC | 22:27 | |
| jeblair | cloudfly: alternate suggestion as to how an end-user should be expected to find out what address to shell into? | 22:27 |
| jeblair | that might conceivably work across several providers... | 22:28 |
| *** troytoman is now known as troytoman-away | 22:28 | |
| cloudfly | give your public floating ip pool internal ips | 22:29 |
| cloudfly | put it behind a gateway | 22:29 |
| cloudfly | and map the address spaces wholesale via nat | 22:29 |
| cloudfly | and provide mesh vpn between all cross network paths | 22:29 |
| cloudfly | that should allow for layer 3 linkage in the ip space. | 22:29 |
| cloudfly | but your mileage may vary | 22:29 |
| cloudfly | alternatively GRE tunnels | 22:32 |
| cloudfly | avoid the nasty crypto overhead | 22:32 |
| *** mattstep has quit IRC | 22:39 | |
| *** mattstep has joined #openstack-dev | 22:40 | |
| *** lloydde has quit IRC | 22:45 | |
| *** mattray has quit IRC | 22:50 | |
| *** AlanClark has quit IRC | 22:50 | |
| *** Glacee has quit IRC | 22:51 | |
| *** rnirmal has quit IRC | 22:52 | |
| *** heckj has quit IRC | 22:53 | |
| *** heckj has joined #openstack-dev | 23:00 | |
| *** zzed has quit IRC | 23:00 | |
| *** sandywalsh has quit IRC | 23:00 | |
| *** heckj has quit IRC | 23:02 | |
| *** bhall has quit IRC | 23:02 | |
| *** sandywalsh has joined #openstack-dev | 23:16 | |
| *** mdomsch has joined #openstack-dev | 23:25 | |
| *** dubsquared has joined #openstack-dev | 23:27 | |
| *** asalkeld is now known as asalkeld_coffee | 23:31 | |
| *** lloydde has joined #openstack-dev | 23:37 | |
| *** flaviamissi has joined #openstack-dev | 23:40 | |
| *** dolphm has joined #openstack-dev | 23:42 | |
| *** asalkeld_coffee is now known as asalkeld | 23:43 | |
| *** nati has joined #openstack-dev | 23:43 | |
| *** jdg has quit IRC | 23:49 | |
| *** lloydde has quit IRC | 23:51 | |
| *** deshantm has joined #openstack-dev | 23:55 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
