Thursday, 2014-07-31

« Wednesday, 2014-07-30 Index Friday, 2014-08-01 »
*** gokrokve has quit IRC01:03
*** gokrokve has joined #murano01:03
*** gokrokve has quit IRC01:08
*** gokrokve has joined #murano01:34
*** gokrokve has quit IRC01:36
*** gokrokve has joined #murano01:37
*** gokrokve has quit IRC01:41
*** gokrokve has joined #murano01:53
*** gokrokve has quit IRC01:54
*** gokrokve has joined #murano01:54
*** gokrokve_ has joined #murano03:07
*** gokrokve has quit IRC03:11
*** gokrokve_ has quit IRC03:12
*** asalkeld has left #murano03:45
*** gokrokve has joined #murano04:11
*** gokrokve has quit IRC04:21
*** gokrokve has joined #murano04:22
*** gokrokve has quit IRC04:26
*** chandankumar has joined #murano04:45
*** gokrokve has joined #murano04:50
*** serg_melikyan has quit IRC05:37
*** killer_prince has quit IRC05:38
*** killer_prince has joined #murano05:39
*** killer_prince is now known as lazy_prince05:39
*** gokrokve_ has joined #murano05:50
*** k4n0 has joined #murano05:51
*** gokrokve has quit IRC05:52
*** gokrokve_ has quit IRC05:54
openstackgerritStan Lagun proposed a change to stackforge/murano: Unit tests for MuranoPL execution result serializer  https://review.openstack.org/10472108:37
openstackgerritStan Lagun proposed a change to stackforge/murano: Unit tests for MuranoPL execution result serializer  https://review.openstack.org/10472108:39
openstackgerritStan Lagun proposed a change to stackforge/murano: Unit tests for MuranoPL execution result serializer  https://review.openstack.org/10472108:59
openstackgerritStan Lagun proposed a change to stackforge/murano: Unit tests for MuranoPL execution result serializer  https://review.openstack.org/10472109:02
openstackgerritEkaterina Chernova proposed a change to stackforge/murano-dashboard: Fix url_prefix in functional tests  https://review.openstack.org/11065409:21
*** slagun has joined #murano09:44
*** stanlagun has quit IRC09:48
openstackgerritEkaterina Chernova proposed a change to stackforge/murano-dashboard: Add wizard for package uploading  https://review.openstack.org/11089410:26
*** k4n0 has quit IRC12:14
openstackgerritA change was merged to stackforge/murano-dashboard: Fix url_prefix in functional tests  https://review.openstack.org/11065412:39
*** sergmelikyan has quit IRC13:03
*** sergmelikyan has joined #murano13:03
*** SaD has joined #murano13:12
*** lazy_prince has quit IRC13:15
*** sergmelikyan has quit IRC13:15
*** chandankumar has quit IRC13:15
*** akuznetsova has quit IRC13:15
*** akuznetsova has joined #murano13:16
*** killer_prince has joined #murano13:16
*** killer_prince is now known as lazy_prince13:16
*** chandankumar has joined #murano13:16
*** drupalmonkey has joined #murano13:41
*** btully has joined #murano14:05
*** chandankumar has quit IRC14:05
btullyJust a heads up. There was a change merged to Horizon master a few days ago that upgrades Twitter Bootstrap from version 2.x to version 3.x. There are lots of changes in the version upgrade that affect both CSS and JS, so if you pull a recent version of Horizon master into murano-dashboard, you'll likely see a lot of broken layouts and broken functionality (i.e., modals). The workaround, for me, was to explicitly download an older version of14:16
btullyhorizon - horizon-2014.2.b2.tar.gz14:16
*** sjmc7 has joined #murano14:27
sjmc7ruhe - it turns out updating to horizon master was not a good idea :)14:32
sjmc7i'm now backing my counter suggestion of updating to the latest release (i.e. 2014.2)14:32
sjmc7there was a huge change last week that affected nearly all the CSS files, and has broken our dashboard14:33
btullysjmc7 not sure if you see IRC history, but i posted about in here about 20 minutes ago14:33
sjmc7ah, ok :)14:33
btully"Just a heads up. There was a change merged to Horizon master a few days ago that upgrades Twitter Bootstrap from version 2.x to version 3.x. There are lots of changes in the version upgrade that affect both CSS and JS, so if you pull a recent version of Horizon master into murano-dashboard, you'll likely see a lot of broken layouts and broken functionality (i.e., modals). The workaround, for me, was to explicitly download an older version of14:34
btullyhorizon - horizon-2014.2.b2.tar.gz"14:34
sjmc7that'll teach me for being late :)14:34
btullynah. it's good to spread the word14:34
sjmc7is there agreement to peg to 2014.2 ?14:34
btullythanks for the suggestion to revert to 2014.214:34
btullyhaven't heard any response. maybe folks are in meetings or lunch/dinner14:35
ruhesjmc7: btully: so, we just need to update the dependency in tox.ini ?14:39
ruheto the latest milestone14:39
sjmc7i think so. and document it somewhere (since typically horizon's installed separately)14:39
ruheand file a critical bug14:39
sjmc7the alternative is to fix it, and have people use 2014.2 to match horizon 2014.214:40
btullyyes, that's a start. but you might need to explicitly download it again if you've already uprgraded to master. I had to delete horizon along with my pip cache and then run pip install http://tarballs.openstack.org/horizon/horizon-2014.2.b2.tar.gz14:41
sjmc7the problem for us is that we're going to be releasing things in a few weeks14:41
sjmc7but with horizon juno.214:41
ruhesjmc7: btully: is there a bug i could refer to in the commit message? (or should i create one?)14:43
btullythere is this bug that was filed against dashboard14:45
btullyhttps://bugs.launchpad.net/murano/+bug/134989414:45
sjmc7https://bugs.launchpad.net/murano/+bug/1349894, ruhe14:45
sjmc7jinx!14:45
btullyand in that bug there is a comment where I refer to the commit in horizon that broke things14:45
btullyhttp://git.openstack.org/cgit/openstack/horizon/commit/?id=92146772b677e9fce57cc11b4a4a1542a05c23b214:45
katyaferventbtully, I was in the middle of creation the same bugs14:47
openstackgerritRuslan Kamaldinov proposed a change to stackforge/murano-dashboard: Switch to Horizon 2014.2.b2  https://review.openstack.org/11098314:47
sjmc7we need to figure out what we do longer term i guess, when we switch versions14:47
btullyWhile it will be a lot of work, the good news is that there is some documentation on what the changes are in bootstrap 2 and 314:47
btullyThis is a good start http://www.bootply.com/bootstrap-3-migration-guide14:48
sjmc7this means devstack will be broken, for instance14:48
btullyand http://upgrade-bootstrap.bootply.com/14:48
sjmc7so actually... maybe this isn't the right fix14:48
*** sjmc7_ has joined #murano14:50
katyaferventbut we still need to update murano to work with bootstrap 3, right?14:50
sjmc7_urgh, what a pain14:50
sjmc7_yeah, i think we will. otherwise anyone installing from devstack is going to immediately hit a problem14:51
katyaferventand all selenium tests also need to be updated14:51
btullyyup, good call on the selenium tests14:51
sjmc7_ruhe - when was the 2014.2 release?14:52
sjmc7_for murano-dashboard, i mean14:52
ruhesjmc7_: jul2414:52
*** gokrokve has joined #murano14:52
sjmc7_ok. i think katyafervent is right - we will need to fix it14:53
ruhehorizon 2014.2.b2 was release on jul2314:53
ruheyeah, i think i need to abandon my patch14:53
ruheand mark that bug as critical14:53
sjmc7_ok. so for our release, we're going to have to use 2014.2 and cherry-pick anything since then14:53
sjmc7_urgh14:53
*** sjmc7 has quit IRC14:53
*** sjmc7_ is now known as sjmc714:53
ruhesjmc7: we'll need to make sure that patches you're going to cherry-pick do not rely on anything new from horizon master14:56
ruhejust need to be more carefull14:56
sjmc7yeah. i'll identify anything we need. it also means the muranoclient version will be stuck at whatever was released for 2014.214:56
sjmc7which is also a pain14:56
sjmc7ok, so we're agreed we need to do the update14:57
sjmc7ruhe - separate issue; our automated test runner is failing on the db.migration.test_migrations.TestMigrations.test_postgresql_opportunistically test14:57
ruhesjmc7: there is no problem in releasing of new version of client as long as it is compatible with the dashboard 2014.2.b214:58
sjmc7ok14:58
openstackgerritGeorgy Okrokvertskhov proposed a change to stackforge/murano: Adds REST API endpoint for action execution  https://review.openstack.org/9944914:58
ruhesjmc7: can you share corresponding logs?14:58
sjmc7yeah14:58
sjmc7http://paste.openstack.org/show/89393/14:59
ruhethe damn thing hides valuable error messages somewhere14:59
sjmc7http://paste.openstack.org/show/89396/15:00
sjmc7more context15:00
sjmc7'failed to migrate to version'15:00
sjmc7psycopg2 2.5.3 installed15:01
ruhesjmc7: can you update test_migrations_base.py and re-run the test? it's in the bottom of the file, we don't print the error message15:04
sjmc7this is an automated test; i don't have access to the environment15:04
sjmc7we have a CI server somewhere15:04
ruhewould it be triggered if i send a patch on review?15:05
sjmc7umm... i might be able to force it to run, yeah15:06
sjmc7usually it just runs nightly off master15:06
openstackgerritDmitry Teselkin proposed a change to stackforge/murano-deployment: Scripts to run dashboard tests on devstack  https://review.openstack.org/11098815:07
openstackgerritRuslan Kamaldinov proposed a change to stackforge/murano: Added logging for DB migration tests  https://review.openstack.org/11098915:09
ruhesjmc7: we'll get more details if this ^^ changes gets merged today15:09
sjmc7that should do the same thing15:10
sjmc7raise with no argument should reraise the caught exception15:10
sjmc7ah, i see, you added the error15:10
sjmc7the 'as e'   'raise e' isn't necessary15:10
katyaferventIf someone has a better English, please, fix the title and description of this bug https://bugs.launchpad.net/murano/+bug/135090015:10
sjmc7it's like a poem, katyafervent! i like it15:11
katyafervent:)15:11
sjmc7ruhe, ok if i just approve  that review now?15:12
ruhesjmc7: let's just wait for +1 from jenkins15:12
sjmc7yep15:12
ruhesjmc7: about as e'   'raise e'. does it apply to python 2.6?15:12
ruhedoes it behave the same way in python 2.6 and 2.7?15:13
sjmc7i think so.. i can't remember it not ever working15:13
sjmc7no harm doing it15:14
openstackgerritRuslan Kamaldinov proposed a change to stackforge/murano: Fixes to unit tests around policies  https://review.openstack.org/10915115:22
sjmc7thanks15:23
ruheit's just an automatic rebase. i don't know why zuul/gerrit couldn't do that for us15:23
sjmc7being difficult :)15:24
ruhei also want to find out why your patch https://review.openstack.org/#/c/109153/ is still failing on migrations15:25
sjmc7ah, it's the same error15:27
sjmc7as we're getting from our CI15:27
sjmc7so that's good, we can just tweak this patch15:27
ruheso, who's going to fix the horizon-master issue? can we split that work? katyafervent do you volunteer to be one of the heroes?15:29
katyaferventSure :)15:30
katyaferventThat should be just one problem for all our forms15:30
katyaferventI can take this bug https://bugs.launchpad.net/horizon/+bug/125842115:30
btullyi'll volunteer as well15:30
katyaferventbtully,  you can take that one https://bugs.launchpad.net/murano/+bug/135090015:31
sjmc7ruhe - same problem for https://review.openstack.org/#/c/109769/15:31
btullysure thing15:31
katyaferventIf you need more details - ask me15:31
ruhegreat! thank you katyafervent and btully!15:32
sjmc7ruhe - other patches are failing too on the same error15:33
sjmc7https://review.openstack.org/#/c/104721/ is one of stan's15:33
ruhesjmc7: i guess all of them have the same root cause15:33
btullyso how would we manage the commits to prevent anything from getting merged?15:33
ruhesjmc7: i'm working on that15:33
btullydid we agree to stick with 2014.2?15:33
sjmc7no btully15:34
btullyfor next release?15:34
sjmc7er... "we" in hp, yes. the project will stick with master15:34
sjmc7so we'll fix it as though the discussion had enver happened15:34
btullyk15:34
sjmc7and we (hp) will not include the patch in our next release15:34
*** PierreRambaud has quit IRC15:43
openstackgerritEkaterina Chernova proposed a change to stackforge/murano-dashboard: Fix functional tests due to Bootstrap Update  https://review.openstack.org/11100015:46
openstackgerritRuslan Kamaldinov proposed a change to stackforge/murano: Default is_public to false  https://review.openstack.org/10915316:02
sjmc7tests are taking a long time today ruhe16:06
openstackgerritDmitry Teselkin proposed a change to stackforge/murano-deployment: Scripts to run dashboard tests on devstack  https://review.openstack.org/11098816:13
*** gokrokve has quit IRC16:17
ruhesjmc7: it's always like this. zuul queue is almost empty during our day and it gets crowded when folks in US wake up :)16:20
sjmc7:)16:20
*** mpitchiah has joined #murano16:24
*** SaD has quit IRC16:27
*** sergmelikyan has joined #murano16:27
*** lazy_prince has quit IRC16:33
openstackgerritDmitry Teselkin proposed a change to stackforge/murano-deployment: Scripts to run dashboard tests on devstack  https://review.openstack.org/11098816:39
openstackgerritDmitry Teselkin proposed a change to stackforge/murano-deployment: Scripts to run dashboard tests on devstack  https://review.openstack.org/11098816:42
*** killer_prince has joined #murano16:43
*** killer_prince is now known as lazy_prince16:43
*** gokrokve has joined #murano16:43
*** gokrokve has quit IRC17:03
*** gokrokve has joined #murano17:04
btullycan someone explain to me how we inject the Murano navigation into the left sidebar?17:05
sjmc7me? :)17:05
btullyhehe was hoping not to bother you17:06
sjmc7it's all django - look at settings.py where murano-dashboard is added to the list of installed dashboards17:07
btullyHORIZON_CONFIG['dashboards'] += ('murano',)17:07
sjmc7yep, and then in muranodashboard/dashboards.py it registers the murano dashboards17:08
sjmc7some details at http://docs.openstack.org/developer/horizon/topics/tutorial.html17:08
btullysuper, thanks17:08
*** sergmelikyan has quit IRC17:41
*** sergmelikyan has joined #murano17:41
sergmelikyanhttps://github.com/sergmelikyan/io.murano.apps.linux.TwoInstances/commit/8888a96dd5bd3bab010cc3d9f8a2924178eba30a - share with bharath17:41
ruheCONTRIBUTING.rst should have a link to "how to setup an IRC bouncer" :)17:43
openstackgerritEkaterina Chernova proposed a change to stackforge/murano-dashboard: Update to Twitter Botstrap 3  https://review.openstack.org/11102017:45
btullyruhe i just added a new blueprint for your review that I think will be useful when implemented and harmless if not utilized18:17
btullyhttps://blueprints.launchpad.net/murano/+spec/add-ability-to-override-dashboard-name18:17
btullyi have a 1 liner that implements it18:18
*** slagun has quit IRC18:33
*** stanlagun has joined #murano18:50
openstackgerritGeorgy Okrokvertskhov proposed a change to stackforge/murano: Adds REST API endpoint for action execution  https://review.openstack.org/9944918:58
sjmc7ruhe - any luck with the test issue? can i be of assistance?19:33
sjmc7stanlagun - is there a concept of 'protected' in the inheritance model in muranopl?19:50
stanlagunno. All methods are public19:50
sjmc7and all state is private?19:50
sjmc7or rather, immutable?19:50
stanlagunimmutable methods???19:51
sjmc7state. variables19:51
sjmc7and related - the modifications you use to speed up the cycle of modifying a model and re-deploying it, are they documented/easy to copy?19:52
sjmc7it would be great if we could port it all to the CLI19:54
stanlagunthere is no access control available yet. Everything that is persisted in object model can be accessed by any class. The only exception are attributes. You can have getter methods with backing attribute to expose data that can be modified only by the owner class19:54
sjmc7that's what i thought, thanks19:55
stanlagunno that is just a hack of my own. There is a BP to make it official19:55
sjmc7ok. are there big barriers to supporting it via CLI calls?19:56
stanlagunI don't get how this is connects to CLI. Maybe we are talking about different things?19:56
sjmc7sergmelikyan, if you're there - https://blueprints.launchpad.net/murano/+spec/software-config-at-deploy-time19:56
sjmc7what we discussed yesterday19:56
openstackgerritRuslan Kamaldinov proposed a change to stackforge/murano: [WIP] Default is_public to false  https://review.openstack.org/11106319:57
ruhethat's just to find out why the original patch fails on DB migrations19:58
sjmc7why do you always use my patches when you're experimenting? :)19:58
stanlagunWhat I do is patching engine code to work with different class loader and immediately working on some hardcoded object model. CLI talks to API, not engine.19:58
sjmc7stanlagun - yeah, i understand your hack isn't one we can use generally, but it should be possible to upload a package and deploy via the CLI at some stage19:59
stanlagunThis may be difficult because of current API - it was designed for another product and needs to be redesigned. But if you upload the package to database (via API) and can provide object model JSON you can do it already (at least to some degree)20:04
stanlagunYou cannot have "deploy this package" command :)20:05
sjmc7no, i guess not. it would be nice to speed up the package deploy cycle at least for development tho20:07
stanlagunYou can make configuration switch that will make engine load packages from 'meta' folder on disk instead on API. And then symlink your package to it. And then write simple CLI tool that sends object model to engine via RabbitMQ20:10
stanlagunBut much better to have complex solution for that: https://blueprints.launchpad.net/murano/+spec/simulated-execution-mode-murano-engine20:11
stanlagunthe idea is to have "debug mode" switch in engine that will allow sumbitting JSON that contains object model, login/password in keystone, list of directories where engine should search for packages and the level of mocking (mock Heat, mock Agents etc)20:14
stanlagun+CLI tool to submit that JSON20:14
*** asalkeld has joined #murano20:53
*** gokrokve has quit IRC20:54
*** openstackgerrit has quit IRC21:01
*** openstackgerrit has joined #murano21:02
*** openstack has joined #murano21:08
*** asalkeld has left #murano21:25
*** gokrokve has joined #murano22:07
*** tsufiev has quit IRC23:04
*** tsufiev has joined #murano23:09
*** asalkeld has joined #murano23:12
*** asalkeld has left #murano23:12
openstackgerritSteve McLellan proposed a change to stackforge/python-muranoclient: Support keystone v3  https://review.openstack.org/11111723:38
*** sjmc7 has quit IRC23:48
*** lazy_prince has quit IRC23:51
« Wednesday, 2014-07-30 Index Friday, 2014-08-01 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!