| *** bobh has joined #openstack-mistral | 00:06 | |
| *** zhurong has joined #openstack-mistral | 00:30 | |
| *** openstackgerrit has quit IRC | 00:41 | |
| *** bobh has quit IRC | 01:43 | |
| *** mattybrennan has quit IRC | 01:55 | |
| *** mattybrennan has joined #openstack-mistral | 01:56 | |
| *** dougshelley66 has quit IRC | 01:57 | |
| *** bobh has joined #openstack-mistral | 02:13 | |
| *** bobh has quit IRC | 02:25 | |
| *** gkadam-away is now known as gkadam | 02:55 | |
| *** jrist has joined #openstack-mistral | 04:30 | |
| niraj_singh | d0ugal:The issue is solved now. thanks :) | 04:56 |
|---|---|---|
| *** jtomasek has joined #openstack-mistral | 05:29 | |
| d0ugal | niraj_singh: great, what was wrong? | 06:20 |
| rakhmerov | hi, any volunteers for this very cool BP? https://blueprints.launchpad.net/mistral/+spec/mistral-run-workflow-from-execution | 06:39 |
| rakhmerov | toure_biab: may be you? | 06:40 |
| rakhmerov | apetrich: or you? | 06:40 |
| *** jpich has joined #openstack-mistral | 06:58 | |
| d0ugal | rakhmerov: so basically restart an exection? | 07:14 |
| rakhmerov | not exactly | 07:14 |
| d0ugal | or copy an execution maybe is better | 07:14 |
| rakhmerov | copy | 07:15 |
| rakhmerov | yes | 07:15 |
| rakhmerov | the idea is that I don't want to pass all the params again, I don't want to remember them etc. | 07:15 |
| rakhmerov | all is already there in some of the existing executions | 07:15 |
| *** ChanServ sets mode: +o openstack | 07:16 | |
| d0ugal | right, makes sense | 07:16 |
| rakhmerov | d0ugal: I'm just debugging something now and stumbled on it. I know that I'd like to start an execution which has the exact same input, params (and env) as one of the existing | 07:17 |
| d0ugal | yup, I think I would only use it when debugging | 07:17 |
| rakhmerov | but I don't even have the input file because somebody else started this workflow | 07:17 |
| rakhmerov | and I don't know where to find it | 07:17 |
| rakhmerov | d0ugal: mostly yes, debugging | 07:17 |
| rakhmerov | agree | 07:17 |
| rakhmerov | but it's convenience | 07:17 |
| rakhmerov | d0ugal: do you know if we still have a requirement to import modules only? | 07:20 |
| rakhmerov | trying to find it.. | 07:20 |
| d0ugal | rakhmerov: I think it is still a hacking rule, but it was never very good at detecting it | 07:21 |
| rakhmerov | yeah | 07:21 |
| d0ugal | but we don't have to follow it :) | 07:21 |
| rakhmerov | yes, found it: https://docs.openstack.org/hacking/latest/user/hacking.html#imports | 07:21 |
| rakhmerov | ok | 07:22 |
| d0ugal | It is actually one of the checks I never really liked. | 07:22 |
| rakhmerov | me too | 07:22 |
| d0ugal | I would prefer a check that banned importing and aliasing to one or two letters :P | 07:22 |
| *** jtomasek has quit IRC | 07:23 | |
| rakhmerov | :)) | 07:23 |
| *** jtomasek has joined #openstack-mistral | 07:25 | |
| rakhmerov | d0ugal: is the word "superfluous" commonly used in regular English? :) | 07:25 |
| rakhmerov | I'm just reviewing a patch and found it | 07:26 |
| rakhmerov | I thought it's a little fancy ) | 07:26 |
| d0ugal | rakhmerov: I don't think it is that common | 07:26 |
| d0ugal | but everyone should know what it means | 07:26 |
| rakhmerov | ok | 07:26 |
| d0ugal | I have seen it more commonly from non-native speakers, so I wonder if there is a word in a language that translates to it (and is a closer match than unnecessary etc) | 07:27 |
| rakhmerov | redundant? | 07:28 |
| d0ugal | that also works too | 07:28 |
| d0ugal | but that isn't what I mean :) | 07:28 |
| d0ugal | I think I explained it badly. | 07:28 |
| rakhmerov | ooh | 07:28 |
| d0ugal | I don't know how to explain it haha | 07:29 |
| rakhmerov | so "superfluous" is not exactly the same as "redundant" and "unnecessary"? | 07:29 |
| d0ugal | No, they are essentially the same | 07:29 |
| d0ugal | so, yes, my example doesn't really make sense | 07:29 |
| rakhmerov | maybe then superfluous is kind of a slang or something? | 07:29 |
| d0ugal | no, it is a legit word :) | 07:30 |
| d0ugal | I have never thought so much about superfluous before! | 07:30 |
| d0ugal | http://wikidiff.com/unnecessary/superfluous | 07:31 |
| d0ugal | rakhmerov: that explains what I wanted to say quite well | 07:31 |
| rakhmerov | let me see.. | 07:31 |
| d0ugal | "As adjectives the difference between unnecessary and superfluous is that unnecessary is not needed or necessary while superfluous is in excess of what is required or sufficient." | 07:31 |
| rakhmerov | ooooh! | 07:32 |
| rakhmerov | I think I got it | 07:32 |
| rakhmerov | ok | 07:32 |
| rakhmerov | it's probably closer to extra | 07:32 |
| rakhmerov | more than sufficient | 07:32 |
| d0ugal | right | 07:33 |
| rakhmerov | ok | 07:33 |
| rakhmerov | you know, now I think that this word (superfluous) actually better reflects the semantics in the code ) | 07:34 |
| d0ugal | lol | 07:34 |
| rakhmerov | yeah, I'll delete my comment | 07:34 |
| d0ugal | I look forward to seeing it used more | 07:34 |
| d0ugal | but don't use superfluous superfluously ;) | 07:35 |
| *** zhurong has quit IRC | 07:35 | |
| rakhmerov | haha :) | 07:35 |
| rakhmerov | I won't | 07:35 |
| rakhmerov | in fact, I think I've never used it | 07:35 |
| rakhmerov | although knew what it means (more or less) | 07:35 |
| *** gkadam has quit IRC | 07:54 | |
| *** zhurong has joined #openstack-mistral | 07:55 | |
| d0ugal | rakhmerov: do you have a list of values you would expect to be in the execution context? | 08:08 |
| d0ugal | just those that you used in your test patch? | 08:08 |
| rakhmerov | d0ugal: yes, just a second | 08:13 |
| rakhmerov | it's in the cod | 08:13 |
| rakhmerov | code | 08:13 |
| d0ugal | okay, I have all of those | 08:13 |
| d0ugal | I guess we can add more later if we need them | 08:13 |
| rakhmerov | ooh, ok | 08:13 |
| rakhmerov | yes | 08:13 |
| rakhmerov | you're working on it now? | 08:14 |
| d0ugal | rakhmerov: yes | 08:14 |
| rakhmerov | awesome | 08:14 |
| rakhmerov | I'm here, ping me if needed | 08:14 |
| d0ugal | will do thanks, there should be some initial patches up soon | 08:14 |
| rakhmerov | great | 08:15 |
| d0ugal | rakhmerov: btw I am on vacation next week. | 08:28 |
| rakhmerov | ok, good to know | 08:28 |
| rakhmerov | thanks | 08:28 |
| d0ugal | rakhmerov: so if I get this started but don't finish, feel free to update it | 08:28 |
| rakhmerov | sure | 08:28 |
| d0ugal | I'm not sure if you need it urgently | 08:28 |
| d0ugal | I'll also try and find somebody to cover for me. | 08:29 |
| rakhmerov | well, so.. On one hand it's not so urgent. On the other hand, it's a important thing that we didn't account for when designing mistral-lib. So it's better to fix it sooner | 08:30 |
| *** shardy has joined #openstack-mistral | 08:30 | |
| d0ugal | Yup, agree | 08:30 |
| rakhmerov | we found it by trying to create an ad-hoc action on top of std.mistral_http and it didn't work | 08:30 |
| *** openstackgerrit has joined #openstack-mistral | 08:30 | |
| openstackgerrit | Andrea Visnyei proposed openstack/mistral master: Mistral install guide https://review.openstack.org/476499 | 08:30 |
| rakhmerov | "action_context" parameter was missing | 08:30 |
| rakhmerov | and when I looked at it closely I realized that we had this design problem | 08:31 |
| d0ugal | Sure, I just wasn't sure if it was blocking you | 08:32 |
| rakhmerov | not in the next couple of weeks | 08:32 |
| d0ugal | cool | 08:32 |
| rakhmerov | yes | 08:32 |
| d0ugal | rakhmerov: do you know where I can easily get the execution context? (workflow id and name etc) | 08:33 |
| *** csatari_ has joined #openstack-mistral | 08:33 | |
| rakhmerov | you mean at what point? | 08:33 |
| rakhmerov | in the code | 08:33 |
| d0ugal | because I guess I need it around here: https://github.com/openstack/mistral/blob/master/mistral/executors/default_executor.py#L109 | 08:33 |
| rakhmerov | ooh, the only way is to prepare it on the engine side and send over RPC to executor | 08:34 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral-lib master: Add a representation of the ActionContext in mistral-lib https://review.openstack.org/506058 | 08:34 |
| d0ugal | rakhmerov: okay, so I need to add it to the run_action call? | 08:35 |
| rakhmerov | on executor side it's not accessible if not sent from engine | 08:35 |
| rakhmerov | yes, I guess so | 08:35 |
| rakhmerov | in one form or another | 08:35 |
| rakhmerov | I'd look if we can reuse rpc_ctx param | 08:35 |
| rakhmerov | don't remember what it is exactly now | 08:35 |
| rakhmerov | whether it's just an auth context (MistralContext) or something else RPC specific | 08:36 |
| d0ugal | hmm, I can't find where it comes from. | 08:40 |
| d0ugal | brb | 08:41 |
| rakhmerov | d0ugal: yeah, it's a little tricky.. Let me try to find it. | 08:41 |
| rakhmerov | ok | 08:41 |
| *** csatari_ has quit IRC | 08:42 | |
| rakhmerov | d0ugal: as a clue, you can look at https://github.com/openstack/mistral/blob/master/mistral/context.py#L200 and see where this method is called within oslo.messaging | 08:46 |
| rakhmerov | when we initialize RPC we give this RpcContextSerializer to oslo.messaging | 08:46 |
| rakhmerov | https://github.com/openstack/mistral/blob/master/mistral/rpc/oslo/oslo_server.py#L52 | 08:47 |
| rakhmerov | we also call it here: https://github.com/openstack/mistral/blob/master/mistral/services/scheduler.py#L65 | 08:48 |
| rakhmerov | in this case it's just our MistralContext | 08:48 |
| rakhmerov | but not sure what it is in case when o.m calls it | 08:49 |
| rakhmerov | maybe it's just easier to debug it and see | 08:49 |
| *** gkadam has joined #openstack-mistral | 09:01 | |
| d0ugal | right, thanks | 09:03 |
| d0ugal | I'll keep digging | 09:03 |
| d0ugal | I do need to go out shortly for a haircut, but I'll continue when I get back. | 09:03 |
| d0ugal | thansk | 09:03 |
| rakhmerov | haha :) ok | 09:05 |
| *** shardy is now known as shardy_afk | 09:05 | |
| *** zhurong has quit IRC | 09:55 | |
| rakhmerov | d0ugal: have you ever used RetryDecorator from oslo_service.loopingcall ? | 10:03 |
| *** zhurong has joined #openstack-mistral | 10:11 | |
| *** jkilpatr has quit IRC | 10:38 | |
| openstackgerrit | Kupai József proposed openstack/mistral master: Removed NOT IN query from expiration policy. https://review.openstack.org/496721 | 10:46 |
| d0ugal | rakhmerov: no, I hadn't even seen it before. | 10:47 |
| rakhmerov | ok, I believe I found a problem with it.. | 10:47 |
| d0ugal | oh? are we using it? | 10:47 |
| rakhmerov | filing a bug | 10:47 |
| rakhmerov | yes | 10:47 |
| rakhmerov | I wrote @db_utils.retry_on_deadlock decorator based on it | 10:47 |
| d0ugal | ah | 10:47 |
| rakhmerov | to retry transactional methods on DB deadlocks (thanks to mysql) | 10:48 |
| rakhmerov | and it doesn't work as I expected | 10:48 |
| d0ugal | What does it do wrong? | 10:48 |
| rakhmerov | ok, so | 10:48 |
| d0ugal | I can just read the bug once you submit if that is easier :) | 10:49 |
| rakhmerov | yeah, ok ) | 10:49 |
| rakhmerov | d0ugal: https://bugs.launchpad.net/oslo.service/+bug/1718635 | 11:01 |
| openstack | Launchpad bug 1718635 in oslo.service "loopingcall.RetryDecorator doesn't drop _retry_count on successful invocation of a decorated method" [Undecided,New] | 11:01 |
| openstackgerrit | zhangyangyang proposed openstack/mistral-lib master: Cleanup test-requirements https://review.openstack.org/506113 | 11:01 |
| rakhmerov | d0ugal: again, I'm not sure. Maybe it's by design | 11:02 |
| openstackgerrit | zhangyangyang proposed openstack/mistral-extra master: Cleanup test-requirements https://review.openstack.org/506115 | 11:02 |
| rakhmerov | but it's quite surprising to me | 11:02 |
| rakhmerov | or maybe I don't use it properly (???) | 11:03 |
| d0ugal | rakhmerov: ooft, that does sound like a bug to me. | 11:04 |
| rakhmerov | yeah ) | 11:04 |
| rakhmerov | I wonder in how many places it's used now ) | 11:04 |
| rakhmerov | I'm writing an email to ML now so that they check | 11:04 |
| d0ugal | rakhmerov: it looks like nova uses it. | 11:06 |
| rakhmerov | :) | 11:06 |
| d0ugal | Only us and nova I think | 11:06 |
| d0ugal | http://codesearch.openstack.org/?q=RetryDecorator&i=nope&files=&repos= | 11:06 |
| rakhmerov | where? | 11:06 |
| d0ugal | The others are unrelated or it finds their own classes with the same name | 11:06 |
| d0ugal | http://git.openstack.org/cgit/openstack/nova/tree/nova/virt/disk/mount/api.py | 11:06 |
| d0ugal | http://git.openstack.org/cgit/openstack/nova/tree/nova/virt/libvirt/guest.py | 11:06 |
| rakhmerov | yep | 11:07 |
| *** jkilpatr has joined #openstack-mistral | 11:10 | |
| *** gkadam has quit IRC | 11:21 | |
| *** thrash|g0ne is now known as thrash | 11:26 | |
| *** dprince has joined #openstack-mistral | 11:30 | |
| *** fultonj has quit IRC | 11:33 | |
| *** fultonj has joined #openstack-mistral | 11:34 | |
| *** shardy_afk is now known as shardy | 11:35 | |
| *** fultonj has quit IRC | 11:40 | |
| *** fultonj has joined #openstack-mistral | 11:40 | |
| *** zhurong has quit IRC | 11:51 | |
| *** chlong has joined #openstack-mistral | 11:52 | |
| *** gkadam has joined #openstack-mistral | 12:06 | |
| *** bobh has joined #openstack-mistral | 12:26 | |
| d0ugal | rakhmerov: are you still around? | 12:28 |
| *** catintheroof has joined #openstack-mistral | 12:41 | |
| rakhmerov | d0ugal: yes, but not for too long | 12:45 |
| rakhmerov | please write, I'll reply later when I can | 12:46 |
| d0ugal | rakhmerov: I have failed to figure out how to get the execution context values | 12:51 |
| d0ugal | rakhmerov: so I was going to see if you had any more pointers. | 12:51 |
| rakhmerov | ok, let me ping you later | 12:54 |
| *** gkadam has quit IRC | 12:58 | |
| d0ugal | rakhmerov: thanks | 12:58 |
| *** gkadam has joined #openstack-mistral | 12:59 | |
| *** toure_biab is now known as toure | 13:01 | |
| *** gkadam has quit IRC | 13:02 | |
| *** d0ugal has quit IRC | 13:02 | |
| *** gkadam has joined #openstack-mistral | 13:02 | |
| *** d0ugal has joined #openstack-mistral | 13:03 | |
| *** weshay is now known as weshay_bbiab | 13:30 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/mistral master: Updated from global requirements https://review.openstack.org/506168 | 13:30 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral-lib master: Add a representation of the ActionContext in mistral-lib https://review.openstack.org/506058 | 13:44 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: WIP: Testing ad-hoc asynchronous actions https://review.openstack.org/505508 | 13:54 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: [WIP] Pass the new ActionContext to mistral-lib https://review.openstack.org/506185 | 13:54 |
| toure | rakhmerov good morning / evening for you :) so would there be any security risks for that spec | 13:55 |
| toure | https://blueprints.launchpad.net/mistral/+spec/mistral-run-workflow-from-execution | 13:56 |
| d0ugal | toure: what security risk are you thinking of? | 13:56 |
| toure | d0ugal what if there are secrets in the workflow for a tenant? | 13:57 |
| toure | I guess keystone would prevent that violation | 13:58 |
| d0ugal | toure: sure, you should only be able to "copy" a workflow you have permissions for | 13:58 |
| d0ugal | s/workflow/execution/ | 13:58 |
| toure | d0ugal yeah, just trying to think of other use cases other than debugging | 13:58 |
| *** yangyapeng has joined #openstack-mistral | 13:59 | |
| toure | d0ugal but this one should be pretty straight forward to implement | 13:59 |
| d0ugal | aye | 13:59 |
| d0ugal | copy workflow name and inputs from existing execution, start new execution | 14:00 |
| d0ugal | I'm honestly not sure I think it is worth it :0 | 14:00 |
| d0ugal | :) | 14:00 |
| d0ugal | I'd just script it | 14:00 |
| d0ugal | but it is a nice and easy thing to do, so why not. | 14:00 |
| toure | yeah it feels a little forced | 14:00 |
| toure | :) | 14:01 |
| d0ugal | the action doesn't really fit into REST terms | 14:01 |
| d0ugal | because it is essentially a GET and a POST | 14:01 |
| d0ugal | https://stackoverflow.com/questions/18755220/what-is-the-restful-way-to-represent-a-resource-clone-operation-in-the-url | 14:02 |
| toure | this would have to implement a service helper to search and re-run the execution | 14:03 |
| d0ugal | Yeah | 14:04 |
| toure | well it's not like I don't know that side of house by now :) | 14:04 |
| d0ugal | cool, so you can knock this out for us today? :-D | 14:05 |
| toure | almost :) | 14:05 |
| toure | let me see if I can put something together for tomorrow | 14:05 |
| toure | :) | 14:05 |
| toure | but I think I could hack something together for tomorrow | 14:06 |
| d0ugal | haha, mostly | 14:06 |
| d0ugal | haha, mostly joking :) | 14:06 |
| toure | :) hehe | 14:06 |
| toure | it would give me something to else to work on that isn't WEA | 14:06 |
| toure | which is all consuming | 14:07 |
| d0ugal | heh | 14:08 |
| d0ugal | need any help with that? | 14:08 |
| toure | possibly in regards to serializing data in a way pecan can consume it... rakhmerov suggested a polymorphic resource class which is fine, it is just trying to get pecan to understand it | 14:11 |
| toure | which I think I can have sqlalchemy serialize the results as json | 14:12 |
| toure | d0ugal http://blog.mmast.net/sqlalchemy-serialize-json | 14:14 |
| d0ugal | toure: I don't think Pecan has any specific support for polymorphic resources. | 14:15 |
| toure | d0ugal it doesn't | 14:15 |
| toure | :) | 14:15 |
| d0ugal | toure: sure, it is easy to convert a result into JSON, but that kinda skips the rest layer | 14:15 |
| d0ugal | we also support XML, not that I think anyone uses it. | 14:15 |
| toure | hrmm | 14:16 |
| d0ugal | (or we did support xml at one point, I have not looked in some time) | 14:17 |
| d0ugal | toure: I suspect if you want to do it the Pecan way, you will need to create a custom resource class | 14:19 |
| d0ugal | toure: but I guess that is what rakhmerov didn't like | 14:19 |
| d0ugal | toure: maybe you could programatically create it based on the others, rather than duplicating the details. | 14:19 |
| toure | d0ugal that is what I will need a hand with | 14:21 |
| toure | d0ugal brb | 14:21 |
| d0ugal | k | 14:21 |
| *** bobh has quit IRC | 14:25 | |
| *** openstackgerrit has quit IRC | 14:33 | |
| *** openstackgerrit has joined #openstack-mistral | 14:34 | |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: [WIP] Pass the new ActionContext to mistral-lib https://review.openstack.org/506185 | 14:34 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: WIP: Testing ad-hoc asynchronous actions https://review.openstack.org/505508 | 14:34 |
| *** thrash is now known as thrash|biab | 14:41 | |
| *** thrash|biab is now known as thrash | 15:16 | |
| *** bobh has joined #openstack-mistral | 15:26 | |
| *** bobh has quit IRC | 15:30 | |
| *** bobh has joined #openstack-mistral | 15:36 | |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: Add actions for the ironic virtual network interface commands https://review.openstack.org/506264 | 15:40 |
| d0ugal | shardy: ^ we are probably going to need/want this. | 15:40 |
| openstackgerrit | Dougal Matthews proposed openstack/mistral master: Add the Ironic wait_for_provision_state action https://review.openstack.org/506268 | 15:45 |
| *** thrash is now known as thrash|biab | 15:48 | |
| *** weshay_bbiab is now known as weshay | 16:07 | |
| shardy | d0ugal: ack, nice thanks for looking into this :) | 16:33 |
| d0ugal | shardy: np, I left a bunch of comments on the review after each update | 16:35 |
| openstackgerrit | Merged openstack/mistral master: Removed NOT IN query from expiration policy. https://review.openstack.org/496721 | 16:38 |
| *** jtomasek has quit IRC | 16:49 | |
| *** jtomasek has joined #openstack-mistral | 16:50 | |
| *** thrash|biab is now known as thrash | 16:58 | |
| *** jpich has quit IRC | 17:01 | |
| openstackgerrit | Merged openstack/mistral master: Updated from global requirements https://review.openstack.org/506168 | 17:22 |
| *** thrash is now known as thrash|biab | 18:36 | |
| mattybrennan | Will an executor keep doing what it was doing if it gets a TERM signal? | 18:45 |
| *** toure is now known as toure_biab | 18:56 | |
| mattybrennan | I have a kubernetes deployment with a few executors and an engine. I did an update to roll out a new docker image. it seems like the executor image went away without finishing its work | 18:56 |
| *** thrash|biab is now known as thrash | 19:06 | |
| *** weshay is now known as weshay_bbiab | 19:50 | |
| *** jkilpatr has quit IRC | 20:47 | |
| *** jkilpatr has joined #openstack-mistral | 21:04 | |
| *** catintheroof has quit IRC | 21:07 | |
| *** dprince has quit IRC | 21:42 | |
| *** zhenguo has quit IRC | 21:45 | |
| *** zhenguo has joined #openstack-mistral | 22:01 | |
| *** thrash is now known as thrash|g0ne | 22:06 | |
| *** bobh has quit IRC | 22:23 | |
| *** bobh has joined #openstack-mistral | 23:24 | |
| *** bobh has quit IRC | 23:24 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!