| *** yangyapeng has quit IRC | 00:02 | |
| *** bobh has joined #openstack-mistral | 00:30 | |
| *** bobh has quit IRC | 01:10 | |
| *** yangyapeng has joined #openstack-mistral | 01:11 | |
| *** daidv has joined #openstack-mistral | 01:15 | |
| *** yangyapeng has quit IRC | 01:29 | |
| *** yangyapeng has joined #openstack-mistral | 01:29 | |
| *** yangyape_ has joined #openstack-mistral | 01:40 | |
| *** yangyapeng has quit IRC | 01:41 | |
| *** daidv has quit IRC | 01:50 | |
| *** daidv has joined #openstack-mistral | 01:51 | |
| *** bobh has joined #openstack-mistral | 02:06 | |
| *** gkadam has quit IRC | 02:06 | |
| *** daidv has quit IRC | 02:23 | |
| *** daidv has joined #openstack-mistral | 02:24 | |
| *** bobh has quit IRC | 03:08 | |
| *** daidv has quit IRC | 03:18 | |
| *** daidv has joined #openstack-mistral | 03:19 | |
| *** gkadam has joined #openstack-mistral | 03:32 | |
| *** gkadam has quit IRC | 03:32 | |
| *** gkadam has joined #openstack-mistral | 03:32 | |
| *** daidv has quit IRC | 03:40 | |
| *** daidv has joined #openstack-mistral | 03:41 | |
| *** rbrady is now known as rbrady-afk | 03:41 | |
| *** daidv has quit IRC | 04:13 | |
| *** daidv has joined #openstack-mistral | 04:14 | |
| *** Dinesh_Bhor has joined #openstack-mistral | 04:14 | |
| openstackgerrit | Renat Akhmerov proposed openstack/mistral master: 'all' parameter breaks task context https://review.openstack.org/510384 | 04:21 |
|---|---|---|
| rakhmerov | d0ugal: good morning, can you fix https://review.openstack.org/#/c/515749 so that we can merge it? | 04:31 |
| *** jaosorior has joined #openstack-mistral | 04:54 | |
| *** katkapilatova has quit IRC | 05:15 | |
| *** katkapilatova has joined #openstack-mistral | 05:15 | |
| *** yangyape_ has quit IRC | 05:24 | |
| *** yangyapeng has joined #openstack-mistral | 05:25 | |
| *** zhurong has joined #openstack-mistral | 05:48 | |
| *** Kevin_Zheng has quit IRC | 06:01 | |
| *** jtomasek has joined #openstack-mistral | 06:26 | |
| *** ijiqiupgcp has joined #openstack-mistral | 07:20 | |
| *** ijiqiupgcp has quit IRC | 07:20 | |
| *** AlexeyAbashkin has joined #openstack-mistral | 07:23 | |
| d0ugal | rakhmerov: I can try, but gerrit isn't loading for me now | 07:52 |
| rakhmerov | щл | 07:58 |
| rakhmerov | ok | 07:58 |
| rakhmerov | d0ugal: I won't be able to attend the meeting today | 07:59 |
| rakhmerov | again | 07:59 |
| apetrich | d0ugal, I thought it was just me | 07:59 |
| rakhmerov | hopefully next time I'll be already fine | 07:59 |
| d0ugal | rakhmerov: okay, that is a shame. Somebody from tripleo was hoping to come and talk about eventlet :) | 07:59 |
| apetrich | 35 mins to post a comment | 07:59 |
| d0ugal | rakhmerov: no worries, I'll see if they can come in here and ask about it. | 07:59 |
| rakhmerov | d0ugal: eventlet? :) | 07:59 |
| rakhmerov | what's the context? I didn't get it | 07:59 |
| d0ugal | rakhmerov: yeah and something about our monkeypatching. | 08:00 |
| rakhmerov | ooh | 08:00 |
| d0ugal | I don't remember it fully, but it is causing a problem with an aciton they are trying to write | 08:00 |
| rakhmerov | d0ugal: I'm really sorry, it's quite a critical situation for us now | 08:00 |
| rakhmerov | ooh, really? | 08:00 |
| rakhmerov | hm... | 08:00 |
| d0ugal | rakhmerov: sure, that is fine. I'll see if they can come in here earlier and then maybe you'll be around | 08:00 |
| rakhmerov | yes | 08:00 |
| rakhmerov | I can spend some time discussing that, yes | 08:01 |
| d0ugal | thanks | 08:01 |
| rakhmerov | d0ugal: do you remember who it was? | 08:01 |
| d0ugal | rakhmerov: social in #tripleo | 08:01 |
| rakhmerov | ok | 08:01 |
| d0ugal | I forget where they live, but I think somewhere in the US | 08:01 |
| rakhmerov | but not Bob, right? | 08:02 |
| d0ugal | apetrich: did you manage to post the comment? It just fails really slowly for me | 08:02 |
| rakhmerov | ooh, gosh... | 08:02 |
| d0ugal | rakhmerov: not Bob | 08:02 |
| rakhmerov | (I need rest) | 08:02 |
| d0ugal | :) | 08:02 |
| rakhmerov | Bob is from Nokia! :) | 08:02 |
| d0ugal | I thought Bob was from Nokia haha | 08:02 |
| rakhmerov | he is, yes! | 08:02 |
| apetrich | d0ugal, finally did | 08:02 |
| d0ugal | rakhmerov: can you remember what needs fixing in that review? I have the code locally (the mistral-lib patch, right?) but I can't see gerrit still. | 08:03 |
| apetrich | d0ugal, that was my comment | 08:03 |
| rakhmerov | d0ugal: one test is failing | 08:04 |
| rakhmerov | to check deprecations | 08:04 |
| apetrich | d0ugal, action context is still using target_id | 08:04 |
| d0ugal | oh really | 08:04 |
| apetrich | the review didn't go in for the task_execution_id | 08:04 |
| d0ugal | They pass locally! | 08:04 |
| rakhmerov | d0ugal: unexpected keyword argument "task_execution_id" in __init__ | 08:04 |
| apetrich | s/target_/task_/ | 08:04 |
| d0ugal | Unless I have a different copy than is on gerrit, but I don't know how that would happen | 08:05 |
| d0ugal | I guess I'll wait. I probably can't submit a change anyway | 08:05 |
| rakhmerov | yes | 08:05 |
| * d0ugal is about to reboot | 08:32 | |
| d0ugal | \quit | 08:34 |
| d0ugal | umm | 08:34 |
| *** d0ugal has quit IRC | 08:34 | |
| *** d0ugal has joined #openstack-mistral | 08:35 | |
| *** d0ugal has quit IRC | 08:35 | |
| *** d0ugal has joined #openstack-mistral | 08:35 | |
| *** d0ugal has quit IRC | 08:35 | |
| openstackgerrit | Dougal Matthews proposed openstack/mistral-lib master: Add deprecated context parameters to the new security context https://review.openstack.org/515749 | 08:35 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral-lib master: Rename task_id to task_execution_id https://review.openstack.org/514675 | 08:35 |
| *** d0ugal has joined #openstack-mistral | 08:40 | |
| *** d0ugal has joined #openstack-mistral | 08:40 | |
| *** d0ugal has quit IRC | 08:40 | |
| apetrich | I miss d0ugal | 08:44 |
| apetrich | and gerrit | 08:44 |
| *** oidgar has joined #openstack-mistral | 08:45 | |
| *** d0ugal has joined #openstack-mistral | 08:48 | |
| *** social has joined #openstack-mistral | 09:16 | |
| social | mornin | 09:16 |
| d0ugal | Hey | 09:17 |
| d0ugal | rakhmerov: social joined here :) | 09:17 |
| rakhmerov | hi | 09:17 |
| rakhmerov | yes | 09:17 |
| social | just gimme sec, booting up, coffee does not work yet | 09:18 |
| rakhmerov | ok | 09:18 |
| d0ugal | :) | 09:18 |
| social | TLDR: https://review.openstack.org/#/c/511254/ -> ansible uses multiprocessing which fails because of eventlet monkeypatching os and threads -> https://github.com/eventlet/eventlet/issues/425 | 09:19 |
| social | so question a) can we not monkeypatch os and threads (I guess there will be more =[) b) should I find workaround where I add patch to mistral to save original modules and provide it to actions so I can load _multiprocessing and override the impot? | 09:20 |
| social | b is long enough to know it's not really good way | 09:20 |
| rakhmerov | social: I believe "a" is hardly feasible | 09:21 |
| rakhmerov | at least for threads | 09:21 |
| rakhmerov | but b) actually makes more sense to me | 09:22 |
| rakhmerov | since generally speaking actions are supposed to run on a separate component, executor | 09:22 |
| social | the issue there is that it should work in theory but I don't have enough insides of python to be able to say where exactly I should override it | 09:22 |
| rakhmerov | not engine | 09:22 |
| social | hmm | 09:22 |
| social | can we monkeypatch per component? | 09:23 |
| social | I mean no monekypatching on executor? | 09:23 |
| rakhmerov | social: that's what I'm thinking, yes | 09:23 |
| rakhmerov | yes | 09:23 |
| d0ugal | I think eventually we want actions to run outside the Mistral process - this would solve problems like this | 09:23 |
| rakhmerov | yeah, the only thing now is that actions can now run on local/embedded executor | 09:25 |
| rakhmerov | essentially on the engine side | 09:25 |
| rakhmerov | in this case it won't probably work | 09:25 |
| social | rakhmerov: in that case should I just fail to run the action or still work on b) | 09:27 |
| rakhmerov | thinking.. | 09:27 |
| *** d0ugal has quit IRC | 09:28 | |
| rakhmerov | well, so, as an option (although I don't really like it much) we could do monkeypatching per component (skip it for remote executors) but using local executors would be just limited in this way | 09:28 |
| rakhmerov | and documented and so on.. | 09:28 |
| social | rakhmerov: how much work you think it is? I think we need this quite asap as running ansible without any feedback for user is kinda wrong | 09:31 |
| social | rakhmerov: so other option except adding ansible python is to just stream the output of running ansible back | 09:32 |
| rakhmerov | social: I don't think it's too much work but we'll need someone's help, indeed | 09:32 |
| rakhmerov | need to find someone who could take care of that | 09:32 |
| rakhmerov | social: I also need to read about the problem itself, I wonder what's wrong with it | 09:33 |
| rakhmerov | maybe it should be just configured somehow to monkeypatch multiprocessing correctly? | 09:33 |
| rakhmerov | we found a similar issue with MySQLdb driver and found that it was already fixed in some of the latest eventlet versions | 09:34 |
| rakhmerov | they just have a flag for that now | 09:34 |
| *** shardy has joined #openstack-mistral | 09:34 | |
| *** matbu has joined #openstack-mistral | 09:36 | |
| social | rakhmerov: so you mean closing https://github.com/eventlet/eventlet/issues/425 ? :) | 09:37 |
| *** d0ugal has joined #openstack-mistral | 09:38 | |
| rakhmerov | sorry, I just didn't read carefully yet | 09:38 |
| d0ugal | social, rakhmerov - sorry, my connection dropped. not sure if I missed anything | 09:39 |
| rakhmerov | social: I don't think it's too much work but we'll need someone's help, indeed | 09:40 |
| rakhmerov | rakhmerov | 09:40 |
| rakhmerov | need to find someone who could take care of that | 09:40 |
| rakhmerov | rakhmerov | 09:40 |
| rakhmerov | social: I also need to read about the problem itself, I wonder what's wrong with it | 09:40 |
| rakhmerov | rakhmerov | 09:41 |
| rakhmerov | maybe it should be just configured somehow to monkeypatch multiprocessing correctly? | 09:41 |
| rakhmerov | rakhmerov | 09:41 |
| rakhmerov | we found a similar issue with MySQLdb driver and found that it was already fixed in some of the latest eventlet versions | 09:41 |
| rakhmerov | rakhmerov | 09:41 |
| rakhmerov | they just have a flag for that now | 09:41 |
| rakhmerov | (last few lines is what I said before) | 09:41 |
| d0ugal | aha, thanks. | 09:41 |
| rakhmerov | np ) | 09:41 |
| rakhmerov | d0ugal, social: so I think b) is not that difficult to do | 09:41 |
| rakhmerov | hope so | 09:42 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: Pass the new ActionContext to mistral-lib https://review.openstack.org/506185 | 09:42 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: Use the new action context in MistralHTTPAction https://review.openstack.org/505508 | 09:42 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: Remove the __init__ method from the test action https://review.openstack.org/513061 | 09:42 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: Rename task_id to task_execution_id https://review.openstack.org/514688 | 09:42 |
| rakhmerov | social: can you file a bug at Launchpad? | 09:42 |
| rakhmerov | please | 09:42 |
| rakhmerov | so that we don't lose the info | 09:42 |
| social | on it | 09:42 |
| rakhmerov | thanks | 09:43 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: WIP Test to demonstrate that task() doesn't work in continue-on https://review.openstack.org/516227 | 09:50 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral-lib master: Rename task_id to task_execution_id https://review.openstack.org/514675 | 09:50 |
| social | rakhmerov: https://bugs.launchpad.net/mistral/+bug/1728552 I hope it makes some sense | 09:53 |
| openstack | Launchpad bug 1728552 in Mistral "mistral should not monkeypatch executor" [Undecided,New] | 09:53 |
| d0ugal | rakhmerov: are you aware of this limitation? https://review.openstack.org/#/c/516227/ (see the comments) | 09:53 |
| d0ugal | social: thanks! | 09:53 |
| social | and about the ansible via python api, should we create some bug for that too and target it for qeens? | 09:56 |
| social | we kinda need it in pike but I understand it's too late | 09:57 |
| d0ugal | queens seems more realistic | 09:57 |
| d0ugal | when we have a fix I guess we can evaluate how hard it would be to backport | 09:57 |
| *** zhurong has quit IRC | 10:05 | |
| *** yangyapeng has quit IRC | 10:08 | |
| *** yangyapeng has joined #openstack-mistral | 10:09 | |
| *** yangyapeng has quit IRC | 10:13 | |
| rakhmerov | d0ugal: nope | 10:14 |
| rakhmerov | interesting.. | 10:14 |
| d0ugal | rakhmerov: I am trying to figure out how to fix it. I managed to break tripleo CI by removing the task name in a continue-on... :( | 10:15 |
| *** gkadam_ has joined #openstack-mistral | 10:20 | |
| *** gkadam_ has quit IRC | 10:20 | |
| *** zhurong has joined #openstack-mistral | 10:20 | |
| *** thrash|bbl is now known as thrash | 11:00 | |
| *** jtomasek has quit IRC | 11:21 | |
| *** jtomasek has joined #openstack-mistral | 11:51 | |
| *** dprince has joined #openstack-mistral | 12:16 | |
| *** yangyapeng has joined #openstack-mistral | 12:20 | |
| mfedosin | rakhmerov: d0ugal hey folks! do you have 5 minutes to reach an agreement on https://review.openstack.org/#/c/514055/4 ? | 12:21 |
| mfedosin | first of all, I agree with rakhmerov's comment and I will update the patch | 12:21 |
| mfedosin | but my main concern is the api impact | 12:22 |
| mfedosin | because after this patch we can't guarantee that the listing results will be the same all the time | 12:23 |
| mfedosin | (name and created_at are not unique fields) | 12:23 |
| mfedosin | so, we have two options: | 12:24 |
| mfedosin | if we are okay with the small api impact we can proceed as is | 12:24 |
| mfedosin | or I can append the lists of default sorting parameters with 'id' | 12:25 |
| mfedosin | for example, here it will be sort_keys='created_at,id' https://github.com/openstack/mistral/blob/master/mistral/api/controllers/v2/workflow.py#L198 | 12:26 |
| mfedosin | I suppose the second option is better | 12:26 |
| mfedosin | tell me what you think | 12:26 |
| *** gkadam has quit IRC | 12:37 | |
| *** shardy is now known as shardy_lunch | 12:42 | |
| *** zhurong has quit IRC | 12:50 | |
| rakhmerov | mfedosin: my opinion, I'd leave it as it's in your latest patchset | 12:58 |
| rakhmerov | d0ugal: what do you think? | 12:58 |
| *** apetrich has quit IRC | 13:30 | |
| *** jtomasek_ has joined #openstack-mistral | 13:30 | |
| *** apetrich has joined #openstack-mistral | 13:30 | |
| d0ugal | rakhmerov: just back from lunch. Reading. | 13:31 |
| *** jtomasek has quit IRC | 13:32 | |
| d0ugal | mfedosin, rakhmerov - I am fine with how you done it in the latest patchset too | 13:33 |
| mfedosin | rakhmerov: d0ugal thanks! I'll update the patch in a minute | 13:33 |
| *** shardy_lunch is now known as shardy | 13:46 | |
| *** vishwanathj has joined #openstack-mistral | 14:34 | |
| openstackgerrit | Brad P. Crochet proposed openstack/mistral master: DNM: TEST Zuul v3 jobs https://review.openstack.org/509428 | 14:46 |
| openstackgerrit | Lukas Bezdicka proposed openstack/mistral master: Don't monkeypatch the executor https://review.openstack.org/516345 | 15:13 |
| *** AlexeyAbashkin has quit IRC | 15:30 | |
| *** AlexeyAbashkin has joined #openstack-mistral | 15:31 | |
| *** AlexeyAbashkin has quit IRC | 15:35 | |
| *** thrash is now known as thrash|biab | 15:42 | |
| *** harlowja has joined #openstack-mistral | 17:19 | |
| *** AlexeyAbashkin has joined #openstack-mistral | 17:30 | |
| *** thrash|biab is now known as thrash | 17:33 | |
| *** AlexeyAbashkin has quit IRC | 17:35 | |
| *** oidgar has quit IRC | 17:41 | |
| *** dprince has quit IRC | 17:50 | |
| *** social has quit IRC | 18:29 | |
| *** dprince has joined #openstack-mistral | 18:37 | |
| *** dprince has quit IRC | 19:01 | |
| *** dprince has joined #openstack-mistral | 19:01 | |
| *** AlexeyAbashkin has joined #openstack-mistral | 19:10 | |
| openstackgerrit | Toure Dunnon proposed openstack/mistral master: [WIP] Running new workflow based on an existing execution. https://review.openstack.org/506652 | 19:12 |
| *** shardy has quit IRC | 19:14 | |
| openstackgerrit | Mike Fedosin proposed openstack/mistral master: Normalize sorting https://review.openstack.org/514055 | 19:17 |
| openstackgerrit | Toure Dunnon proposed openstack/python-mistralclient master: [WIP] Running new workflow based on existing execution. https://review.openstack.org/506653 | 19:52 |
| *** toure is now known as toure_biab | 20:01 | |
| *** bobh has joined #openstack-mistral | 20:06 | |
| *** bobh has quit IRC | 20:18 | |
| *** bobh has joined #openstack-mistral | 20:22 | |
| *** AlexeyAbashkin has quit IRC | 20:37 | |
| *** bobh has quit IRC | 20:39 | |
| *** jtomasek_ has quit IRC | 20:58 | |
| *** thrash is now known as thrash|g0ne | 21:30 | |
| *** dprince has quit IRC | 22:02 | |
| *** AlexeyAbashkin has joined #openstack-mistral | 22:29 | |
| *** AlexeyAbashkin has quit IRC | 22:33 | |
| *** AlexeyAbashkin has joined #openstack-mistral | 23:29 | |
| *** AlexeyAbashkin has quit IRC | 23:33 | |
| *** bobh has joined #openstack-mistral | 23:47 | |
| *** histrio has quit IRC | 23:51 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!