Thursday, 2016-07-14

*** toddjohn has quit IRC00:01
*** toddjohn has joined #openstack-mistral00:01
*** toddjohn has quit IRC00:02
*** toddjohn has joined #openstack-mistral00:02
*** gyee has quit IRC00:08
*** toddjohn has quit IRC00:15
*** bobh has joined #openstack-mistral00:24
openstackgerritPanFengyun proposed openstack/mistral: Add Python 3.5 classifier and venv  https://review.openstack.org/34187800:36
*** tonytan4ever has joined #openstack-mistral00:54
*** tonytan4ever has quit IRC00:59
*** bobh has quit IRC01:08
*** tonytan4ever has joined #openstack-mistral02:02
*** tonytan_brb has joined #openstack-mistral02:10
*** tonytan4ever has quit IRC02:13
*** toddjohn has joined #openstack-mistral02:15
*** toddjohn has quit IRC02:20
*** toddjohn has joined #openstack-mistral02:32
*** toddjohn has quit IRC02:34
*** bobh has joined #openstack-mistral03:15
*** bobh has quit IRC03:25
*** bobh has joined #openstack-mistral03:29
*** toddjohn has joined #openstack-mistral03:34
*** bobh has quit IRC03:37
*** toddjohn has quit IRC03:40
*** harlowja has quit IRC04:13
*** toddjohn has joined #openstack-mistral04:37
*** toddjohn has quit IRC04:41
rakhmerovkong: ok04:57
*** mflobo1 has joined #openstack-mistral05:09
*** mflobo has quit IRC05:09
openstackgerritMerged openstack/mistral: Updated from global requirements  https://review.openstack.org/34169305:12
*** Ravikiran_K has joined #openstack-mistral06:10
*** mflobo has joined #openstack-mistral06:15
*** mflobo1 has quit IRC06:15
*** toddjohn has joined #openstack-mistral06:38
*** toddjohn has quit IRC06:43
*** shardy has joined #openstack-mistral07:21
*** Ravikiran_K has quit IRC07:26
*** mgershen has joined #openstack-mistral07:30
*** tonytan_brb has quit IRC07:40
rakhmerovhparekh: does it tell anything to you? http://logs.openstack.org/44/335944/8/gate/gate-python-mistralclient-devstack-dsvm/4c682fc/console.html07:42
rakhmerovsome tempest changes again?07:42
*** d0ugal has joined #openstack-mistral08:00
*** jpich has joined #openstack-mistral08:02
*** jistr is now known as jistr|off08:11
openstackgerritRenat Akhmerov proposed openstack/python-mistralclient: WIP  https://review.openstack.org/34204108:26
*** shardy has quit IRC08:29
hparekhrakhmerov i don't think its tempest issue08:39
rakhmerovhparekh: yeah, seems like. I'm now investigating it08:39
*** toddjohn has joined #openstack-mistral08:40
*** mflobo has quit IRC08:41
rakhmerovhparekh: looks like we have one more issue with Mistral installation08:42
rakhmerovI run the populate command again and output is:08:43
rakhmerovWARNING:mistral.actions.openstack.action_generator.base:Failed to create action: tacker.create_vnfd 'NoneType' object has no attribute 'Client'08:43
rakhmerovWARNING:mistral.actions.openstack.action_generator.base:Failed to create action: tacker.show_extension 'NoneType' object has no attribute 'Client'08:43
rakhmerov/usr/lib/python2.7/site-packages/sqlalchemy/sql/default_comparator.py:153: SAWarning: The IN-predicate on "workflow_definitions_v2.id" was invoked with an empty sequence. This results in a contradiction, which nonetheless can be expensive to evaluate.  Consider alternative strategies for improved performance.08:43
rakhmerov  'strategies for improved performance.' % expr)08:43
rakhmerovbecause of this some of the actions don't get installed08:43
rakhmerovfor example, std.http08:43
rakhmerovdo you have time to investigate it?08:44
*** toddjohn has quit IRC08:44
rakhmerovI asked that person to file a bug, it should be done soon08:44
rakhmerovhparekh: the code was taken from master branch 1-2 days ago08:46
*** mflobo has joined #openstack-mistral09:04
*** shardy has joined #openstack-mistral09:08
rakhmerovhparekh: https://bugs.launchpad.net/mistral/+bug/160299909:14
openstackLaunchpad bug 1602999 in Mistral "std.http cannot be found from action-list " [Critical,New] - Assigned to hardik (hardik-parekh047)09:14
hparekhrakhmerov, Ok09:36
rakhmerovon that func test problem, I found the reason, thinking how to fix it..09:39
openstackgerritRenat Akhmerov proposed openstack/mistral: Remove redundant param in task controller function  https://review.openstack.org/34207609:52
rakhmerovakuznetsova, nmakhotkin, kong, hparekh: guys, please approve https://review.openstack.org/#/c/342076/, it's a stupid bug in task controller that blocks all client development (func tests don't pass)10:00
rakhmerovddeja: here?10:01
rakhmerovI have some work for you :)10:01
rakhmerovping me when available10:01
kongrakhmerov: do you have the error log of your patch?10:05
rakhmerovhttp://logs.openstack.org/44/335944/8/gate/gate-python-mistralclient-devstack-dsvm/4c682fc/console.html10:07
rakhmerovit's a different patch but same cause10:07
kongrakhmerov: I fount bunch of them :-)10:08
rakhmerovI reproduced it locally also10:08
rakhmerovok10:08
kongrakhmerov: i see there is a problem of the code, 'workflow_execution_id' and 'wf_ex_id', which one is valid?10:11
rakhmerovkong, hparekh: wait a sec!!10:11
rakhmerovmy fix is wrong!10:11
kongthey are both in the filters10:11
rakhmerovkong: yes, right10:11
rakhmerovgive me a minute10:11
kongrakhmerov: :-)10:11
rakhmerovplease wait10:11
rakhmerovdon't go10:11
rakhmerov:)10:11
kongI will be here for next 1 hour10:12
rakhmerovwf_ex_id doesn't make sense in filters at all10:12
kongI believe you could still get a coffee before you fix it10:12
rakhmerovso I need to remove it from filters and leave as an explicit parameter10:12
kongrakhmerov: agreed10:12
rakhmerovyes ;) sorry, I've not been in a good shape for a couple of days10:13
konghah10:13
konghparekh: btw, if you are available, please take a look at my event trigger patches10:13
kongI missed n2, the only chance to merge it is n310:14
kongrakhmerov: what time is it in Russia?10:18
rakhmerovkong: that's ok, but I didn't expect you to make a release today :) Was hoping to squeeze more patches into N-210:19
rakhmerovit's now 5.19 pm here10:19
kongseems like you are gonna go home10:20
hparekhkong, yeah sure I will take a look10:23
kongrakhmerov: never mind, we have some valuable features included in n2.10:23
konghparekh: cheers10:24
rakhmerovkong: yes, actually if you look at what we've done in N-2 it's about the same amount of work we did in entire Mitaka cycle :)10:25
kongrakhmerov: indeed10:25
rakhmerovkong: still fixing that problem, it surprisingly doesn't work with my changes..10:25
kongI went over the commit logs before I did the release10:25
kongit's amazing10:26
kongbut i am not that surprised given the amount of participants in IRC channel.10:26
ddejarakhmerov: ping10:36
rakhmerovddeja: yes, hi10:36
ddejahello :)10:36
rakhmerovkong: yes :) We're growing. But we need to make Mistral more stable, people complain very often that something doesn't work after installation10:37
rakhmerovddeja: can you please check mitaka/stable version? People complain that it doesn't work10:37
rakhmerovbecause of rpc problems10:38
ddejarakhmerov: sure10:38
rakhmerovhttp://paste.openstack.org/show/532771/10:38
rakhmerovmay be we need to backport something10:38
ddejaooooh, I know10:39
rakhmerovwhat's that?10:39
ddejathat's stupid bug, but I'm not sure how to fix it...10:39
ddejabasically, on master branch we have directory called rpc (mistral/engine/rpc/) with RPC-related stuff in it10:40
ddejain mitaka, there was no such diractory10:40
rakhmerovand?10:40
ddejabut, there was file rpc.py (mistral/engine/rpc.py)10:40
rakhmerovhm.. still not understanding..10:40
ddejaso, if you firstly have downloaded master version10:40
ddejaand the chechkout to mitaka10:40
rakhmerovoooh!!!10:40
ddejayou will have both empty dir and a file10:40
rakhmerovgosh...10:41
ddejaand python is importing empty dir...10:41
rakhmerovhm...10:41
rakhmerovf...k..10:41
rakhmerovso maybe we just need to mention it in the docs?10:42
rakhmerovto clone mitaka/stable, not just to switch10:42
rakhmerovif there's an installation from master10:42
ddejaor we can add one line to mitaka, to delete rpc dir10:42
ddejain launch.py10:43
rakhmerovyes10:43
rakhmerovthat will work too10:43
rakhmerovwell, yes, ideally it should just work regardless of how it was installed10:43
ddejafor me it is strange that checkout does not delete the rpc directory...10:45
ddejalet me check something, maybe this can be fixed using git10:45
rakhmerovok10:45
rakhmerovkong: I fixed that issue in controller, I found more bugs there )10:45
rakhmerovwill send a patch soon10:45
rakhmerovalso want to refactor it a little bit10:45
rakhmerovthat method _get_task_resources_with_results is not really needed10:46
rakhmerovddeja: and the second thing is comments in you "delivery model" patches10:46
rakhmerovfrom kond )10:46
rakhmerovfrom kong10:46
kongrakhmerov: basicly, there are lack of unit tests10:46
rakhmerovkong: for this particular case yes10:47
ddejarakhmerov: yup, I see them. As I wrote, I'm working on different patches that would be better, so this one can be abandoned :)10:47
rakhmerovwhen we apply these filters10:47
rakhmerovddeja: different patches? What do you mean?10:47
rakhmerovyou're scaring me.. )10:47
ddejaI mean the one that would make use of 'redelivered' flag10:48
rakhmerovaah, I see10:48
rakhmerovok10:48
ddejaso that kombu driver would send all messages in process then ack mode10:48
rakhmerovok10:48
ddejawell, it's not toatly new, I've cherry-picked two of them :)10:49
ddeja(two of old commits)10:49
kongrakhmerov, ddeja, the only reason users want to use kombu driver is for benefit of process-then-ack feature, right?10:49
ddejakong: very short anwser: yes10:50
rakhmerovyes10:50
rakhmerovand I would add this: if users want to avoid other possible issues caused by o.m10:50
rakhmerovunexpected updates etc.10:50
kongddeja: ok, then we should make that deliver mode as default in kombu driver10:51
rakhmerovhm.. I agree, but we probably need to test it more before we make it a default10:51
rakhmerovbecause now there's a bunch of people who use o.m RPC10:52
kongrakhmerov: yes, of course.10:52
rakhmerovthey won't be happy if we jack up :)10:52
openstackgerritRenat Akhmerov proposed openstack/mistral: Fixing filtering in task controller  https://review.openstack.org/34207610:55
rakhmerovkong, hparekh: I fixed the patch, https://review.openstack.org/#/c/342076/11:00
rakhmerovwe need to write tests for filtering, of course, but it's a little tricky thing to do because now we just mock db calls in API tests11:01
rakhmerovso I ask you to approve this patch now and deal with tests later11:01
rakhmerovif units pass, of course11:01
rakhmerovyes, tests passed11:07
rakhmerovhparekh: any luck with that actions problem?11:11
hparekhrakhmerov, Its not reproducing in my environment11:11
hparekhstill trying11:12
rakhmerovit feels to me that because of issues in Tacker actions other actions don't install on 100%11:12
rakhmerovooh, not reproducing...11:12
rakhmerovso you install Mistral from scratch, do mistral-db-manage populate and all actions get installed?11:12
rakhmerovincluding std.http?11:12
hparekhI tried sync_db11:14
hparekhIt works with that script11:14
rakhmerovit should be the same theoretically11:14
hparekhyeah11:14
rakhmerovbut you can also try db manage11:14
hparekhhmmm.11:15
rakhmerovgot something interesting? )11:16
rakhmerovkong: here?11:34
rakhmerovcan you approve https://review.openstack.org/#/c/342076/? Failure is not related to the patch11:34
rakhmerovor akuznetsova, nmakhotkin11:38
hparekhrakhmerov, no luck on reproducing. I tried by uninstalling tacker client also. And one more thing even if tacker registrations fail it cannot block std actions because std actions registered first11:38
rakhmerovhm...11:39
rakhmerovweird11:39
hparekhDo you have detail log if you reproduced it ?11:39
rakhmerovonly what's in the bug description11:39
rakhmerovok, can you please reply in the bug?11:40
hparekhOne more thing regarding mistral on https11:40
rakhmerovyes11:40
hparekhDo you mean mistral api on https is not working ?11:40
hparekhyeah sure I will reply on the bug11:41
rakhmerovhparekh: yes, right11:41
hparekhSo for this mistral should be run as apache wsgi module right ? as mistral is not supporting https directly11:42
rakhmerovit doesn't support it directly?11:43
rakhmerovhm.. well, yes, you're right11:43
rakhmerovbut wait a sec... then I don't understand why it's about Mistral11:44
hparekhNo I don't think we use simple_server api which only supports http request11:44
rakhmerovok, my fault, I should have asked him more about the issue11:44
rakhmerovI asked him to file a bug but seems like he didn't do it yet11:44
rakhmerovlet me ask him again about what they're doing11:44
hparekhYeah I think this should not about mistral11:44
rakhmerovok11:44
hparekhOk no issue11:44
rakhmerovsorry11:44
hparekhHave we already started work on custom apis ?11:45
rakhmerovhparekh: yes, I'm preparing repos now11:48
rakhmerovyou can look at my patches to infra and mistral-extra11:49
rakhmerovI just want to do this first step and then let other people work on it11:49
rakhmerovfolks from Red Hat were interested11:49
rakhmerovI also prepared stubs for packages, modules and methods according to the spec but I'll push them once mistral-lib repo is available11:51
rakhmerovhparekh: how about working on YAQL functions API? :)11:52
rakhmerovit's similar to Actions API in many ways11:52
rakhmerovsame kind of thing but for writing YAQL functions11:52
hparekhrakhmerov, so actions library for YAQL functions11:53
rakhmerovyes11:53
hparekhyes sure I am interested11:53
rakhmerovit will also be part of mistral-lib11:53
rakhmerovjust to clarify: so now we already can plug in new YAQL functions but when implementing a new one we can't do much regarding Mistral functionality11:54
rakhmerovwe basically can only write something simple like math stuff11:55
rakhmerovor something that that access workflow context11:55
rakhmerovthat's it11:55
rakhmerovbut if we want to query DB or something like that then we also need something like API11:55
rakhmerovsomething that we can safely use11:55
rakhmerovit should be stable and well documented etc.11:55
rakhmerovso, for example, using DB API directly (or workflow/utils.py) may not be a good idea because it's an internal API and it may change any time11:56
rakhmerovI think it's important that we work on it now, otherwise people can start using various internal stuff and will be broken by our changes11:57
hparekhagree we need some abstraction11:58
rakhmerovhparekh: ok to take it?11:59
hparekhrakhmerov, yes I am ok11:59
rakhmerovgo ahead please, file a BP (and spec if you think it's needed)12:00
rakhmerovan important part of it will be a good doc12:00
hparekhrakhmerov, sure. :)12:00
rakhmerovhparekh: btw, they filed a bug: https://bugs.launchpad.net/mistral/+bug/160306312:00
openstackLaunchpad bug 1603063 in Mistral "Cannot run Mistral with SSL" [High,New] - Assigned to hardik (hardik-parekh047)12:00
rakhmerovso please look at it12:00
openstackgerritMerged openstack/mistral: Fixing filtering in task controller  https://review.openstack.org/34207612:04
*** Ravikiran_K has joined #openstack-mistral12:04
rakhmerovhparekh: on that SSL bug, they are just expecting same behaviour as for other projects12:09
rakhmerovthat support it12:09
rakhmerovlike Heat12:09
hparekhrakhmerov, yes but we don't support it12:09
rakhmerovso for HEAT we can define options for API like cert_file etc.12:09
rakhmerovyes, right12:10
rakhmerovso it's rather a BP than a bug12:10
hparekhyes right12:10
rakhmerovand many core projects don't use pecan12:10
rakhmerovI don't know exactly how to do it with pecan, need to research12:10
hparekhit can be done using oslo_service instead of wsgi_ref12:11
hparekhIt may work need some experiment12:12
rakhmerovhparekh: ok12:17
rakhmerovso please take it too12:17
rakhmerovbut I think it should be filed as a blueprint12:17
hparekhrakhmerov, yes agree. I have filled blueprint :)12:23
ddejarakhmerov: I double checked what are effects of changing the branch from master to mitaka. And always the direcotry 'mistral/engine/rpc' was gone. But if it is present and there is __init__.py file in it, I got same error as you show me12:24
ddejaso for sure this is the reasone, but I don't know why this is happening, but on newest master and newest mitaka it works well12:26
rakhmerovhparekh: thanks man12:26
rakhmerovddeja: may it depend on git version, for example?12:27
ddejarakhmerov: Oh, I've just reproduce it12:27
rakhmerov:)12:27
ddejasooo12:27
ddejait is due to we git ignore *.pyc files12:27
ddejaso after we run master version12:28
*** toddjohn has joined #openstack-mistral12:28
ddejait creates __init__.pyc in mistral/engine/rpc12:28
*** dprince has joined #openstack-mistral12:28
rakhmerovand then it doesn't get deleted?12:28
ddejaand since .pyc files are not tracked by git, this file is left12:28
ddejayup^^12:28
rakhmerovok12:28
rakhmerovso what's the solution then12:29
ddejaI'll sugest to delete all *.pyc files on startup12:29
rakhmerovddeja: I'm sorry, I have to leave the office now, may be back in a few hours12:29
ddejarakhmerov: no problem12:29
rakhmerovddeja: yeah, sounds good12:29
ddejaI'll talk with Roman if he has some ideas :)12:30
ddeja_gryf: ^^12:30
_gryfso after short discussion we think, that it might be less problematic to rename this package (rpc) to other name, like rpc_backend12:37
_gryfso there wouldn't be name clash after pyc leftover12:37
*** toddjohn has quit IRC12:47
*** d0ugal has quit IRC12:53
*** tonytan4ever has joined #openstack-mistral12:58
openstackgerritRoman Dobosz proposed openstack/mistral: Rename package mistral.engine.rpc to mistral.engine.rpc_backend  https://review.openstack.org/34216613:08
openstackgerritRoman Dobosz proposed openstack/mistral: Rename package mistral.engine.rpc to mistral.engine.rpc_backend  https://review.openstack.org/34216613:16
*** toddjohn has joined #openstack-mistral13:29
*** bigdogstl has joined #openstack-mistral13:31
*** bigdogstl has quit IRC13:31
*** ninag has joined #openstack-mistral13:39
*** ninag has quit IRC13:45
*** vishwanathj has joined #openstack-mistral13:58
*** bobh has joined #openstack-mistral14:01
*** catintheroof has joined #openstack-mistral14:16
*** vishwanathj has quit IRC14:40
*** vishwanathj has joined #openstack-mistral14:40
*** d0ugal has joined #openstack-mistral14:47
*** rrecio has joined #openstack-mistral14:52
*** rrecio_ has joined #openstack-mistral14:53
*** rrecio has quit IRC14:56
*** shardy_ has joined #openstack-mistral15:02
*** shardy has quit IRC15:10
*** d0ugal has quit IRC15:16
*** d0ugal has joined #openstack-mistral15:18
*** toddjohn has quit IRC16:01
*** toddjohn has joined #openstack-mistral16:18
*** bobh has quit IRC16:29
*** d0ugal has quit IRC16:49
*** jpich has quit IRC16:57
shardy_Hi all, any yaql gurus around here (or is there a dedicated channel?)17:05
shardy_http://paste.openstack.org/show/532886/17:05
shardy_I'm looking for the cleanest way to combine two dicts, replacing the values in one based on a lookup of another17:06
*** vishwanathj has quit IRC17:15
*** shardy_ has quit IRC17:25
*** harlowja has joined #openstack-mistral17:28
*** catintheroof has quit IRC17:58
*** toddjohn has quit IRC18:15
*** toddjohn has joined #openstack-mistral18:16
openstackgerritMiles Gould proposed openstack/mistral: Fix some trivial SPAG errors in docs  https://review.openstack.org/34233918:24
*** toddjohn has quit IRC18:50
*** toddjohn has joined #openstack-mistral18:51
*** dprince has quit IRC18:54
*** bobh has joined #openstack-mistral19:06
*** catintheroof has joined #openstack-mistral19:18
*** tonytan4ever has quit IRC19:21
*** tonytan4ever has joined #openstack-mistral19:22
*** gyee has joined #openstack-mistral20:13
*** mgershen has quit IRC20:24
*** mgershen has joined #openstack-mistral20:24
*** mgershen has quit IRC20:25
*** mgershen has joined #openstack-mistral20:26
*** toddjohn has quit IRC20:31
*** tonytan4ever has quit IRC20:31
*** mgershen has quit IRC20:56
*** mgershen has joined #openstack-mistral21:05
*** bobh has quit IRC21:22
*** hparekh has quit IRC21:30
*** tonytan4ever has joined #openstack-mistral21:32
*** toddjohn has joined #openstack-mistral21:32
*** bobh has joined #openstack-mistral21:35
*** toddjohn has quit IRC21:35
*** toddjohn has joined #openstack-mistral21:36
*** tonytan4ever has quit IRC21:37
*** toddjohn has quit IRC21:39
*** toddjohn has joined #openstack-mistral21:39
*** catintheroof has quit IRC21:40
*** hparekh has joined #openstack-mistral21:43
*** toddjohn has quit IRC21:44
*** mgershen has quit IRC21:44
*** bobh has quit IRC22:03
*** openstackgerrit has quit IRC22:03
*** openstackgerrit has joined #openstack-mistral22:03
Ravikiran_KHi22:31
*** tonytan4ever has joined #openstack-mistral22:33
Ravikiran_Ki have installed mistral dashboard and server on two different vms...from dashboard i am trying to access workflows..request is going to the localhost22:34
Ravikiran_Kits not taking mistral endpoint22:34
Ravikiran_Kcan someone help me one this22:35
*** tonytan4ever has quit IRC22:38
*** toddjohn has joined #openstack-mistral22:45
*** toddjohn has quit IRC22:46
*** toddjohn has joined #openstack-mistral22:47
*** toddjohn has quit IRC22:51
*** tonytan4ever has joined #openstack-mistral23:05
*** rrecio_ has quit IRC23:17
*** tonytan4ever has quit IRC23:49

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