Tuesday, 2017-04-25

*** zhurong has joined #openstack-mistral01:50
*** thrash is now known as thrash|g0ne02:35
*** bobh has quit IRC03:02
*** gongysh has joined #openstack-mistral03:17
*** zhurong has quit IRC03:21
*** bobh has joined #openstack-mistral04:03
*** bobh has quit IRC04:07
*** warface has joined #openstack-mistral04:35
*** tung_doan has joined #openstack-mistral04:46
*** gkadam has joined #openstack-mistral05:00
*** gongysh has quit IRC05:11
*** warface has quit IRC05:44
*** warface has joined #openstack-mistral05:45
*** zhurong has joined #openstack-mistral05:52
*** zhurong has quit IRC06:01
openstackgerritMerged openstack/python-mistralclient master: fix release note formatting  https://review.openstack.org/45934106:07
openstackgerritMerged openstack/python-mistralclient master: Explicitly set 'builders' option  https://review.openstack.org/45794006:07
*** jaosorior_away is now known as jaosorior06:16
*** gongysh has joined #openstack-mistral06:17
gongyshhttps://bugs.launchpad.net/mistral/+bug/168558306:18
gongyshdo you guys have any idea on this bug?06:18
openstackLaunchpad bug 1685583 in Mistral "action does not be deleted when workflow execution is delete" [Undecided,New]06:18
apetrichd0ugal, trying to figure out what is missing for mistral-lib release in rdo. I'm looking at the built packages and don't see it there07:00
apetrichd0ugal, but today is holiday in Italy and my go to person is off07:01
d0ugaldang07:01
d0ugalapetrich: I'll try and take a look shortly, just catching up :)07:01
d0ugalapetrich: where do you see the built packages?07:01
apetrichd0ugal, no worries.07:01
apetrichd0ugal, https://buildlogs.centos.org/centos/7/cloud/x86_64/rdo-trunk-master/07:02
d0ugalhttps://www.rdoproject.org/documentation/rdo-packaging/#how-to-add-a-new-package-to-rdo-trunk07:03
d0ugalapetrich: ^ we are on step 5 there now07:03
apetrichd0ugal, oh I see07:04
apetrichcool07:04
apetrichI didn't see that07:05
*** jpich has joined #openstack-mistral07:33
*** fultonj_ has joined #openstack-mistral07:41
*** fultonj has quit IRC07:41
gongyshkong, d0ugal ,hi08:32
*** gongysh has quit IRC08:58
*** gongysh has joined #openstack-mistral09:16
*** sharatss has joined #openstack-mistral09:27
*** tuan_ has joined #openstack-mistral09:31
openstackgerritSharat Sharma proposed openstack/mistral master: Fix doc generation for python 3  https://review.openstack.org/45720409:38
*** shardy has joined #openstack-mistral09:41
d0ugalgongysh: hi09:49
gongyshd0ugal, https://bugs.launchpad.net/mistral/+bug/168558309:49
openstackLaunchpad bug 1685583 in Mistral "action does not be deleted when workflow execution is delete" [Undecided,New]09:49
gongyshis there a way to stop the left over running action if we delete the workflow execution?09:50
d0ugalgongysh: Good question. I'm not sure, I don't have any actions that run forever09:51
d0ugalIMO users shouldn't be allowed to delete running workflows - they should be required to stop them first.09:51
*** gongysh has quit IRC10:08
*** thrash|g0ne is now known as thrash10:27
thrashd0ugal: welcome back10:27
d0ugalthrash: thanks :)10:27
*** mgershen has joined #openstack-mistral10:30
*** mgershen1 has quit IRC10:33
thrashd0ugal: so this is fun... The nova python api has a 'get_keys()' method on the Flavor object that it returns... Which has additional info.10:42
thrashd0ugal: https://docs.openstack.org/developer/python-novaclient/ref/v2/flavors.html10:43
thrashd0ugal: so... we don't necessarily get back all info from the auto-generated actions. *sigh*10:43
*** jkilpatr has quit IRC10:49
d0ugalthrash: damg10:54
d0ugaldamn10:54
d0ugalI guess this is the downside of action generation10:55
thrashd0ugal: It's still useful... Just means that you need to use a custom action for anything other than base cases...10:58
thrashd0ugal: perhaps a proposal to put client generation in the base action? A client manager like OSC?10:59
tuan_jaosorior: Hi, it seems that the tripleo gate is not fixed yet10:59
tuan_:(10:59
tuan_do you have some updated information about it10:59
thrashd0ugal: we have all of the 'get_x_client()' in the TripleoBase...10:59
thrashd0ugal: seems like that might be something useful11:00
d0ugalthrash: yeah, we will have that in mistral-extra when we move the openstack actions there11:00
thrashd0ugal: cool11:00
d0ugalthrash: there was some related discussion in "[openstack-dev] [mistral] Mistral Custom Actions API Design" a while ago11:01
thrashd0ugal: ack. I'll go look at that thread.11:02
d0ugalthrash: I'm not sure how we should expose client constructor args to users.11:04
d0ugalthrash: for example, this is quite tripleo specific: https://github.com/openstack/tripleo-common/blob/master/tripleo_common/actions/base.py#L65-L7111:04
jaosoriorthrash, d0ugal: Hey, if oyu have time can you look at the issue that tuan_ was encountering? I think it's gonna bite us at some point.11:05
d0ugaltuan_: what's the issue?11:05
thrashd0ugal: that's a good question...11:07
d0ugalthrash: rbrady was looking at this problem too - I forget what his use-case was tho'11:07
tuan_d0ugal: the tripleo gate fails with the error of app11:08
tuan_ERROR(app)11:08
*** bobh has joined #openstack-mistral11:08
tuan_i will send you the log11:08
d0ugaltuan_: which patch?11:09
tuan_d0ugal: this patch: https://review.openstack.org/#/c/455407/11:09
d0ugaltuan_: okay, I'll look into the logs - give me a few mins.11:09
tuan_even before this patch, another guy had also the same problem11:09
tuan_with the latest master at that time11:09
tuan_jaosorior: could you correct me if i am saying something wrong11:10
tuan_:)11:10
jaosoriortuan_: will try. I'm in the middle of trying to debug another issue though :/ sorry if I take some time to respond11:10
tuan_jaosorior: oh, i am sorry too to disturb you11:10
tuan_so focus on your stuff11:10
tuan_i will contact to dougal for this problem11:11
*** jkilpatr has joined #openstack-mistral11:11
openstackgerritluong tuan proposed openstack/mistral master: Refactor mistral context using oslo_context  https://review.openstack.org/45540711:11
*** bobh has quit IRC11:13
d0ugaltuan_: I can't find where it fails in the logs :(11:14
tuan_d0ugal: yep, we had the same result too11:14
tuan_:)11:14
jaosoriorok, back for a bit11:15
tuan_remember that before this patch11:15
jaosoriord0ugal: so there are two issues here.11:15
tuan_jaosorior: so you turn11:15
jaosoriord0ugal: One is that there is some error that is not getting logged in mistral. (might be some middleware, who knows)11:15
jaosoriord0ugal: the second issue is that the mistralclient is hiding the errors it gets from the server11:15
jaosoriorso, for instance, if we see an error 500. It will just say ERROR (app)11:15
jaosoriord0ugal: we had a similar issue IIRC in tripleoclient, that's why I pinged you about the second error.11:16
d0ugalright11:17
d0ugalyikes11:17
tuan_jaosorior: i also consider that why 500 error happened11:17
tuan_but no error in mistral-api server11:17
jaosoriortuan_: most likely is some middleware issue. those tend not to get logged for some reason.11:17
tuan_well, even client logs out error11:18
tuan_but with 500 as Internal Server Error11:18
tuan_it should show out some error from server side11:18
tuan_and in mistral-api server, we had DEBUG mode as log level11:18
d0ugaljaosorior, tuan_ - there is almost nothing in the mistral logs, maybe it isn't starting correctly?11:19
tuan_d0ugal: with the vanilla mistral11:19
tuan_everything works well11:19
tuan_i have no idea with mistral with triplo11:20
tuan_tripleo11:20
d0ugaljaosorior: the undercloud install failed in https://review.openstack.org/#/c/455407/11:20
d0ugalon PS511:20
tuan_yep11:20
d0ugaloh11:21
d0ugalI see the ERROR app11:21
jaosoriord0ugal: it's the error tuan_ is talking about.11:21
d0ugalokay, sorry, it seems I am just catching up11:21
jaosoriorno logs in the mistral server11:21
d0ugalsorry :P11:21
jaosoriorbut some strange ERROR (app)   as the sole output11:21
d0ugalYou need to come to me with easier problems after I have been off!11:21
jaosoriord0ugal: no biggie, we should probably do better explaining the context of this11:21
d0ugalnp11:22
d0ugalOkay, doing some experimenting now I see where this happens11:22
d0ugaljaosorior: I think our cut removes the error11:25
d0ugal"cut -f 2 -d ' '"11:25
d0ugalit should look more like http://paste.openstack.org/show/607851/11:26
jaosoriord0ugal: by the way, I'm replacing those https://review.openstack.org/#/c/458808/11:30
d0ugaljaosorior: oh, interesting - that will be less error prone11:30
d0ugaljaosorior: I am going to base a patch on yours, and add a depends on for tuan_11:31
d0ugal's patch11:31
jaosoriord0ugal: awesome, thanks for looking into this11:31
*** tuan_ has quit IRC11:47
*** catintheroof has joined #openstack-mistral11:48
*** tuan_ has joined #openstack-mistral11:50
*** weshay_PTO is now known as weshay11:59
*** dprince has joined #openstack-mistral12:02
*** bobh has joined #openstack-mistral12:02
*** catintheroof has quit IRC12:08
*** tuan_ has quit IRC12:12
openstackgerritDougal Matthews proposed openstack/python-mistralclient master: Add a missing space to the help message for execution-create  https://review.openstack.org/45366612:12
*** tuan_ has joined #openstack-mistral12:14
*** chlong has quit IRC12:20
shardytuan_: Hi!  You had some questions related to trusts?12:22
tuan_shardy: Hi, here you are12:23
tuan_yep12:23
tuan_actually, partially related to12:23
tuan_:D12:23
tuan_since now i am finding a place to store trustor_id and trustee_id12:23
tuan_i saw that in Heat we store them in User_Creds table12:23
shardyYeah we have a table for that12:24
tuan_what is the purpose of that table12:24
tuan_just store information of user12:24
tuan_?12:24
shardytuan_: yeah, previously (or when deferred_auth_method=password is selected) we stored the user ID of the user that created the stack, and encrypted the password in the context12:24
shardytuan_: but now, with the default configuration, we store only the trust ID and the trustor/trustee details12:25
shardye.g no passwords or other sensitive data12:25
shardyWhen I say "now", I mean like since 2013 ;)12:25
shardyI just used that table because we already had it, I guess you can store it somewhere in the mistral DB?12:26
tuan_shardy: Perfect12:26
tuan_in Mistral i do not see any suitable table to store it12:26
tuan_:D12:26
tuan_may be we can create a new table to store them12:27
tuan_we could not store trustee/trustor_id within workbook table12:27
tuan_:D12:27
shardytuan_: Yeah, I guess you need to associate the trust with a workflow execution?12:28
shardyIn heat we create the user_creds record then reference it from the stack table12:28
tuan_and in some actions related to stack (create, update, etc.), the trust is associcated?12:29
shardytuan_: on create of the stack we create and store the trust12:29
tuan_shardy: and other actions just get the trust from db12:30
tuan_?12:30
shardytuan_: yes exactly12:30
tuan_shardy: thanks Steve12:30
tuan_so i think in mistral, we need somewhere to store the trust12:31
shardyI guess it's simpler for mistral as you only have one lifecycle operation, e.g you create the execution (and the trust) then it runs (potentially consuming the trust to reauthenticate)12:31
shardy?12:31
tuan_and the trust is created when we create a workbook, workflow12:31
shardytuan_: but creating a workbook or workflow is independent of it's execution?12:31
shardyI would expect the trust to be created at the point of running the workflow, not storing its definition12:32
tuan_shardy: i agree12:32
tuan_let's create it only when running12:32
shardyOk, sounds good then :)12:32
tuan_what about a table call trust_creds in order to store the trust information (trust_id and trustor_id)12:33
tuan_?12:33
shardytuan_: I don't have any strong opinion on the naming - heat uses user_creds because it supports more than one credential method, not just trusts12:34
tuan_shardy: yeah, so i think let create something general and then may be later, it can be used in another purpose12:35
tuan_so user_creds seems OK for me12:35
shardyI'm not a mistral expert but +1 FWIW :)12:35
tuan_i think i will go forward with it and let's wait for the comments from others12:35
tuan_shardy: thanks :D12:36
tuan_and thank you again for your time12:36
shardytuan_: sounds good, please add me to the review when you post it and I'll try to help with review feedback12:36
tuan_shardy: yep, definitely i will12:36
apetrichd0ugal, so it merged12:46
d0ugalapetrich: yup! I am waiting for it to show up here: http://trunk.rdoproject.org/centos7-master/report.html12:46
d0ugalapetrich: I guess that is the next step?12:46
d0ugalapetrich: looking at the commit and build datetimes, it seems it might take about an hour?12:47
d0ugalor longer12:47
apetrichd0ugal, I assume so. I'm going to look at the promote ci. it might be there12:47
d0ugalapetrich: good idea, where do I see that?12:49
apetrichd0ugal, it is one of those I think https://ci.centos.org/search/?q=promote12:49
apetrichd0ugal, ignore that. promote will pass from master into master-passed-ci12:50
d0ugalah12:50
apetrichnot relevant12:50
d0ugalokay :)12:50
*** sharatss has quit IRC12:53
*** sharat has joined #openstack-mistral12:53
*** catintheroof has joined #openstack-mistral13:01
apetrichd0ugal, we might have to have a merging commit for it to trigger13:02
d0ugalapetrich: ah, makes sense.13:03
d0ugalapetrich: commit in mistral-lib?13:03
apetrichd0ugal, aye13:03
d0ugalhrm, I'm nnot sure what I could do lol13:04
apetrichyeah. that's the thing13:04
apetrichI'm talking to people lets see if I can figure this out13:05
apetrichnever mind. we probably better ask on rdo13:10
apetrichd0ugal, so next commit to mistral-lib or to the dist git13:17
d0ugalapetrich: :( I can't think of anything that is needed right now13:19
d0ugalapetrich: if you want to make a commit to mistral-lib that just tweaks a docstring or something I'll +2 it :P13:20
d0ugalapetrich: the README could be updated, it contains lies :)13:21
*** toure|gone is now known as toure13:33
apetrichd0ugal, will look at it. s/lies<enter><esc>dd:wq13:35
*** bobh has quit IRC13:38
*** chlong has joined #openstack-mistral13:40
*** sharat has quit IRC13:41
*** livelace has joined #openstack-mistral13:46
openstackgerritAdriano Petrich proposed openstack/mistral-lib master: Fix documentation for custom actions  https://review.openstack.org/45969513:47
apetrichd0ugal, ^^13:47
openstackgerritDougal Matthews proposed openstack/mistral-lib master: Remove the unused ApplicationContextNotFoundException  https://review.openstack.org/45969813:48
d0ugalapetrich: technically Result can be used for successful results too btw13:48
apetrichd0ugal, I don't think I saw an example of that on tripleo-common13:50
apetrichd0ugal, mostly was either **mistral_result that I assume can be successful or errors13:51
d0ugalapetrich: https://github.com/openstack/mistral-lib/blob/e8814640bb6c9325eb9c9bb3dbd295f65bcfb1df/mistral_lib/actions/types.py#L2413:52
d0ugalapetrich: depending what arg you use determines if it was an error or just data being returned13:52
apetrichd0ugal, let me fix that doc13:53
openstackgerritAdriano Petrich proposed openstack/mistral-lib master: Fix documentation for custom actions  https://review.openstack.org/45969514:00
apetrichd0ugal, I think it is better ^14:00
*** warface has quit IRC14:08
d0ugalapetrich: there was a build btw: https://trunk.rdoproject.org/centos7/report.html15:08
apetrichd0ugal, woot15:09
openstackgerritBrad P. Crochet proposed openstack/mistral master: [WIP] - DO NOT MERGE - Implement the sensitive data blueprint.  https://review.openstack.org/45974715:09
thrashd0ugal: ^^^ extreme beginning. :D15:10
*** gkadam has quit IRC15:10
d0ugalthrash: neat!15:10
d0ugalLGMT +215:10
d0ugal:P15:10
thrashd0ugal: hehe... It adds the 'hidden' field to the dsl, and.. That's about it.15:10
thrashBut it is documented! :P15:10
d0ugalthrash: I spotted and appreciated that!15:11
openstackgerritBrad P. Crochet proposed openstack/mistral master: [WIP] - DO NOT MERGE - Implement the sensitive data blueprint.  https://review.openstack.org/45974715:13
thrashd0ugal: and... a release note. lol15:13
d0ugalthrash: lol, I almost joked about that.15:13
thrashhehe15:13
*** tuan_ has quit IRC15:19
*** shardy is now known as shardy_afk15:25
*** shardy_afk is now known as shardy15:49
*** jpich has quit IRC16:06
*** chlong has quit IRC16:10
*** chlong has joined #openstack-mistral16:24
openstackgerritToure Dunnon proposed openstack/mistral master: [WIP] Workflow Error Analysis (do not merge!)  https://review.openstack.org/45544716:55
*** clenimar has joined #openstack-mistral16:58
openstackgerritToure Dunnon proposed openstack/mistral master: [WIP] Workflow Error Analysis (do not merge!)  https://review.openstack.org/45544717:18
*** jaosorior is now known as jaosorior_away17:19
*** chlong has quit IRC17:37
*** livelace has quit IRC17:40
*** chlong has joined #openstack-mistral17:53
*** dprince has quit IRC18:03
*** fultonj_ is now known as fultonj18:09
*** shardy has quit IRC18:36
*** dprince has joined #openstack-mistral18:39
tung_doand0ugal: thrash: Hi, could you guys please let me know if Mistral have some ways to trigger a workflow using a custom event/notification? Seems just OpenStack notifications are introduced in this spec: https://specs.openstack.org/openstack/mistral-specs/specs/newton/approved/event-notification-trigger.html18:46
*** Kevin_Zheng has quit IRC18:47
thrashtung_doan: not sure I follow about an "OpenStack" notification...18:48
d0ugaltung_doan: you can trigger a workflow with a zaqar queue19:10
d0ugaltung_doan: https://www.openstack.org/videos/video/building-self-healing-applications-with-aodh-zaqar-and-mistral19:10
*** catinthe_ has joined #openstack-mistral19:28
*** catintheroof has quit IRC19:30
*** chlong has quit IRC19:46
*** chlong has joined #openstack-mistral19:57
*** bobh has joined #openstack-mistral20:00
*** jkilpatr has quit IRC20:10
*** catinthe_ has quit IRC20:42
*** jkilpatr has joined #openstack-mistral20:48
*** dprince has quit IRC21:08
*** chlong has quit IRC21:28
*** harlowja has quit IRC22:34
*** bobh has quit IRC22:57
*** openstack has joined #openstack-mistral23:28
*** harlowja has joined #openstack-mistral23:29
kongtung_doan: hi, yeah, mistral does support run workflow based on notifications defined. We have the feature supported in API, but lack of documentation for now.23:35
kongIn order to use that feature, mistral needs to be deployed in OpenStacn environment because it needs to listen to MQ23:36
kongbut for event/notifications out of openstack, d0ugal is right, you need several openstack services combination (at least zaqar and mistral)23:39
*** bobh has joined #openstack-mistral23:50

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