*** AlexeyAbashkin has joined #openstack-mistral | 00:10 | |
*** AlexeyAbashkin has quit IRC | 00:14 | |
*** bobh has joined #openstack-mistral | 00:22 | |
*** gkadam has joined #openstack-mistral | 00:23 | |
*** gkadam_ has quit IRC | 00:25 | |
*** AlexeyAbashkin has joined #openstack-mistral | 00:46 | |
*** AlexeyAbashkin has quit IRC | 00:51 | |
*** gkadam has quit IRC | 01:07 | |
*** Kevin_Zheng has joined #openstack-mistral | 01:21 | |
*** bobh has quit IRC | 02:11 | |
*** AlexeyAbashkin has joined #openstack-mistral | 02:47 | |
*** AlexeyAbashkin has quit IRC | 02:52 | |
*** AlexeyAbashkin has joined #openstack-mistral | 03:08 | |
*** AlexeyAbashkin has quit IRC | 03:12 | |
*** yangyape_ has joined #openstack-mistral | 04:09 | |
*** yangyapeng has quit IRC | 04:11 | |
*** gkadam has joined #openstack-mistral | 04:19 | |
*** jaosorior has joined #openstack-mistral | 05:05 | |
*** mcdoker181818 has joined #openstack-mistral | 05:10 | |
*** mcdoker181818 has quit IRC | 07:17 | |
rakhmerov | d0ugal: can you please review https://review.openstack.org/#/c/512520/ again? | 07:27 |
---|---|---|
d0ugal | rakhmerov: sure | 07:28 |
d0ugal | rakhmerov: I have a bunch of reviews open if you get time :) | 07:28 |
rakhmerov | yes, ok | 07:29 |
rakhmerov | :) | 07:29 |
rakhmerov | trying to survive here but I'll find time :) | 07:29 |
d0ugal | haha, they can wait until next week | 07:29 |
d0ugal | rakhmerov: maybe just this one so I can start the release process. https://review.openstack.org/#/c/514675/ | 07:29 |
rakhmerov | yep | 07:30 |
rakhmerov | no problem | 07:30 |
rakhmerov | d0ugal: seems like 'task_execution_id' from the __init__ parameters is not used anywhere | 07:32 |
d0ugal | rakhmerov: whoops, line 87 is wrong. One moment | 07:35 |
d0ugal | 82 I mean | 07:36 |
rakhmerov | yes, right | 07:36 |
openstackgerrit | Dougal Matthews proposed openstack/mistral-lib master: Rename task_id to task_execution_id https://review.openstack.org/514675 | 07:37 |
d0ugal | rakhmerov: ^ | 07:37 |
rakhmerov | yep | 07:37 |
*** oidgar has joined #openstack-mistral | 07:38 | |
rakhmerov | ok, looks good now | 07:39 |
d0ugal | Thanks | 07:39 |
rakhmerov | so task_id now has higher priority | 07:39 |
d0ugal | Yeah | 07:39 |
d0ugal | well, sort-of | 07:40 |
d0ugal | but if you pass both that is weird :) | 07:40 |
rakhmerov | yep | 07:40 |
rakhmerov | that's ok | 07:40 |
d0ugal | I don't mind changing the order | 07:40 |
apetrich | d0ugal, it might be me as a non native speaker but I had to reread "It will be removed in the future" twice. I'd rather read "Task_id will be removed in the future" | 07:42 |
apetrich | d0ugal, but that is a very minor nit. LGTM | 07:42 |
* d0ugal re-reads | 07:43 | |
d0ugal | I think that is okay, but happy to take suggestions. | 07:43 |
apetrich | d0ugal, it might just be morning. I can read it normaly now | 07:43 |
d0ugal | lol | 07:43 |
d0ugal | apetrich: It is difficult to think with that bright thing outside | 07:44 |
*** AlexeyAbashkin has joined #openstack-mistral | 07:56 | |
*** oidgar has quit IRC | 08:08 | |
*** mcdoker181818 has joined #openstack-mistral | 08:16 | |
mcdoker181818 | rakhmerov: What do you think about pluginaze a TaskPolicy through the stevedore? It would allow users to develop own a task policy. | 09:03 |
mcdoker181818 | Good morning all :) | 09:03 |
rakhmerov | hi | 09:03 |
rakhmerov | mcdoker181818: the idea sounds good but the implementation of policies is pretty hairy | 09:04 |
rakhmerov | and involves a lot of interactions with internal engine mechanisms | 09:04 |
rakhmerov | I thought about this idea initially when I was developing some of the policies | 09:05 |
rakhmerov | but we first need to understand how to make the engine more flexible etc. | 09:05 |
rakhmerov | so believe me, it's really not easy | 09:05 |
mcdoker181818 | than what do you think about 'fail-on' policy that would allow to fail a task based on true\false expression after execution? | 09:14 |
mcdoker181818 | after task execution | 09:14 |
openstackgerrit | Merged openstack/mistral master: Add a config option to disable cron triggers https://review.openstack.org/512520 | 09:51 |
*** daidv has quit IRC | 10:12 | |
*** gkadam has quit IRC | 10:34 | |
rakhmerov | mcdoker181818: task or action? | 10:40 |
d0ugal | it should fail the task IMO | 10:43 |
mcdoker181818 | task | 10:46 |
*** gkadam has joined #openstack-mistral | 10:47 | |
d0ugal | mcdoker181818: Can you write up an example of what fail-on would look like? | 10:52 |
d0ugal | a full task or workflow | 10:52 |
d0ugal | I don't think I really understand what you could do with fail-on that you couldn't do with the fail engine command. | 10:52 |
d0ugal | mcdoker181818: maybe what you really want is a new engine command to fail the task but not the workflow? | 10:56 |
d0ugal | but I still don't know why you would want that :) | 10:56 |
d0ugal | why is it so important that the task fails? | 10:56 |
rakhmerov | yeah, I'm trying to see where it would be useful | 11:01 |
rakhmerov | and we also need to keep in mind that "on-error" and "on-success" refer to action error or success | 11:02 |
rakhmerov | task is just an action call essentially | 11:02 |
d0ugal | good point | 11:06 |
openstackgerrit | Dougal Matthews proposed openstack/mistral master: Pass the new ActionContext to mistral-lib https://review.openstack.org/506185 | 11:07 |
openstackgerrit | Dougal Matthews proposed openstack/mistral master: Use the new action context in MistralHTTPAction https://review.openstack.org/505508 | 11:07 |
openstackgerrit | Dougal Matthews proposed openstack/mistral master: Remove the __init__ method from the test action https://review.openstack.org/513061 | 11:07 |
openstackgerrit | Dougal Matthews proposed openstack/mistral master: Rename task_id to task_execution_id https://review.openstack.org/514688 | 11:07 |
mcdoker181818 | For example, std.http returns some json. I check value. If value equals 'RUNNING' I retried task (retry -> on-continue). But if json equals 'ERROR'. I want that task has ERROR state. And for example, tasks will be triggered in 'on-error'. | 11:09 |
d0ugal | mcdoker181818: why not just do it in an on-complete/on-success? | 11:09 |
rakhmerov | do what? | 11:10 |
d0ugal | on-complete: - error_task: <% task().result = "ERROR" %> | 11:10 |
rakhmerov | yep | 11:10 |
d0ugal | I understand a it, you want the task to be an error because that is conceptually how you see it | 11:11 |
d0ugal | but really the action didn't fail | 11:11 |
rakhmerov | probably my main concern here is: what will having the task in ERROR state give us? | 11:11 |
* d0ugal will bbiab | 11:12 | |
rakhmerov | in order to handle the task's ERROR state we'll need something else | 11:12 |
rakhmerov | because, like I said, on-complete/on-error/on-success is all about actions | 11:12 |
rakhmerov | what to do if the action(s) failed, succeeded or just finished (no matter what the result is) | 11:13 |
rakhmerov | now task may get to ERROR state if its action(s) succeeded only if some post action run things failed | 11:14 |
rakhmerov | like "publish", "on-xxx" clause etc. | 11:14 |
rakhmerov | e.g. an error during YAQL/Jinja expression evaluation | 11:14 |
mcdoker181818 | The administrator see tasks but not actions. Now this task will has SUCCESS state. Administrutor must retried this task. But he cann't, right? | 11:23 |
mcdoker181818 | I will do example of implementation with examples to Monday :) | 11:37 |
d0ugal | The administrator should also pay attention to failing workflows, so you can show it that way | 11:43 |
d0ugal | the workflow itself should do the retrying if that is needed anyway tho' | 11:44 |
mcdoker181818 | All tasks has success state, how do i show users the root cause? | 11:52 |
mcdoker181818 | rakhmerov: > only if some post action run things failed | 12:02 |
mcdoker181818 | Not. For example is timeout policy. My action has success state, but task is failed | 12:02 |
*** dprince has joined #openstack-mistral | 12:05 | |
*** gkadam has quit IRC | 12:37 | |
*** toure_biab is now known as toure | 12:38 | |
d0ugal | mcdoker181818: I would improve the workflow output if the cause isn't clear. | 12:55 |
mcdoker181818 | Maybe the task output too. But users oriented by color (SUCCESS-green, ERROR-red) :D | 13:05 |
mcdoker181818 | I tried to show you examples on Monday. Maybe then you'll change your mind. | 13:05 |
d0ugal | Sounds good :) | 13:06 |
d0ugal | I'm not totally against it, we just need to be careful what we add. | 13:06 |
*** shardy has quit IRC | 13:07 | |
mcdoker181818 | I agree with you. | 13:07 |
rakhmerov | mcdoker181818: language is always a very sensitive part and changes may have serious long-term impact on end users | 13:17 |
rakhmerov | that's why we discuss them during long time and with lots of people | 13:18 |
mcdoker181818 | Yep, I understand. This is why I offer opportunity to pluginaze policy too :) | 13:21 |
mcdoker181818 | I have dev build with policy with steveodore, I show on Monday too | 13:22 |
mcdoker181818 | rakmerov: While you're here can it will be merged? https://review.openstack.org/#/c/510384/ | 13:24 |
mcdoker181818 | rakhmerov: | 13:25 |
rakhmerov | mcdoker181818: I'll review, yes | 13:28 |
openstackgerrit | Renat Akhmerov proposed openstack/mistral master: 'all' parameter breaks task context https://review.openstack.org/510384 | 13:32 |
mcdoker181818 | Thank you | 13:36 |
openstackgerrit | Merged openstack/mistral master: Fix launcher tests https://review.openstack.org/512574 | 13:46 |
*** Dinesh_Bhor has quit IRC | 14:04 | |
*** d0ugal_ has joined #openstack-mistral | 14:04 | |
*** d0ugal has quit IRC | 14:05 | |
*** bobh has joined #openstack-mistral | 14:55 | |
*** toure is now known as toure_biab | 14:57 | |
*** bobh has quit IRC | 14:58 | |
*** toure_biab is now known as toure | 15:00 | |
*** thrash is now known as thrash|biab | 15:06 | |
*** AlexeyAbashkin has quit IRC | 15:13 | |
openstackgerrit | Dougal Matthews proposed openstack/mistral-lib master: Rename task_id to task_execution_id https://review.openstack.org/514675 | 15:43 |
openstackgerrit | Dougal Matthews proposed openstack/mistral-lib master: Add deprecated context parameters to the new security context https://review.openstack.org/515749 | 15:43 |
*** thrash|biab is now known as thrash | 15:44 | |
*** jaosorior has quit IRC | 15:48 | |
*** AlexeyAbashkin has joined #openstack-mistral | 16:53 | |
*** AlexeyAbashkin has quit IRC | 16:57 | |
*** thrash is now known as thrash|bbl | 17:21 | |
*** mcdoker181818 has quit IRC | 17:48 | |
*** harlowja has quit IRC | 18:05 | |
*** harlowja has joined #openstack-mistral | 18:05 | |
*** dprince has quit IRC | 18:48 | |
*** dprince has joined #openstack-mistral | 19:13 | |
*** bobh has joined #openstack-mistral | 20:02 | |
*** bobh has quit IRC | 20:04 | |
*** bobh has joined #openstack-mistral | 20:05 | |
*** bobh has quit IRC | 20:09 | |
*** dprince has quit IRC | 20:27 | |
*** jtomasek has quit IRC | 22:14 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!