| *** threestrands has joined #openstack-mistral | 00:40 | |
| *** threestrands has quit IRC | 00:40 | |
| *** threestrands has joined #openstack-mistral | 00:40 | |
| *** harlowja has quit IRC | 01:13 | |
| *** mfedosin has quit IRC | 02:03 | |
| *** mfedosin has joined #openstack-mistral | 02:27 | |
| *** gkadam has joined #openstack-mistral | 02:55 | |
| *** gkadam has quit IRC | 02:59 | |
| *** hardikjasani has joined #openstack-mistral | 04:30 | |
| *** openstackgerrit has quit IRC | 05:10 | |
| *** lxkong has quit IRC | 05:16 | |
| *** aspiers has quit IRC | 05:16 | |
| *** mwhahaha has quit IRC | 05:16 | |
| *** portdirect has quit IRC | 05:16 | |
| *** lxkong has joined #openstack-mistral | 05:35 | |
| *** aspiers has joined #openstack-mistral | 05:35 | |
| *** mwhahaha has joined #openstack-mistral | 05:35 | |
| *** portdirect has joined #openstack-mistral | 05:35 | |
| *** akovi has joined #openstack-mistral | 05:59 | |
| *** gkadam has joined #openstack-mistral | 07:08 | |
| *** shardy has joined #openstack-mistral | 07:12 | |
| *** openstackgerrit has joined #openstack-mistral | 07:35 | |
| openstackgerrit | Andras Kovi proposed openstack/mistral master: Allow engine commands as task name https://review.openstack.org/582966 | 07:35 |
|---|---|---|
| rakhmerov | d0ugal, akovi, mcdoker181818: hi, I'd appreciate if you review https://review.openstack.org/581671 | 07:44 |
| rakhmerov | mcdoker181818: see my comments about not using this option for executor and notifier | 07:45 |
| *** AlexeyAbashkin has joined #openstack-mistral | 07:45 | |
| rakhmerov | there's no explicit testing for this because I didn't have any good idea how to do this | 07:45 |
| rakhmerov | it's implicitly tested though as I switched the default value to "eventlet" | 07:45 |
| d0ugal | rakhmerov: sure, I'll take a look soon! | 07:48 |
| d0ugal | I just need coffee :) | 07:48 |
| rakhmerov | thanks ) | 07:48 |
| rakhmerov | and good morning! | 07:48 |
| *** jtomasek has joined #openstack-mistral | 07:50 | |
| d0ugal | Morning! | 07:56 |
| d0ugal | btw everyone, I'll be releasing mistral-lib tomorrow if there are any changes | 08:33 |
| d0ugal | This week if the last week for non-client libraries | 08:33 |
| *** threestrands has quit IRC | 08:51 | |
| akovi | d0ugal: What do you think: should we aim to merge these https://review.openstack.org/#/q/topic:service-catalog-issue+(status:open+OR+status:merged) | 09:00 |
| jlejeune | akovi: +1 :) | 09:00 |
| *** nguyenhai_ has joined #openstack-mistral | 09:03 | |
| pgaxatte | jlejeune: haha you were lurking in the shadows :D | 09:05 |
| pgaxatte | hi everyone | 09:06 |
| *** nguyenhai93 has quit IRC | 09:06 | |
| jlejeune | pgaxatte: :) | 09:10 |
| *** xavierhardy has joined #openstack-mistral | 09:11 | |
| xavierhardy | Hi everyone | 09:11 |
| xavierhardy | Is it possible to add more executor thread/process per executor instance in Mistral? | 09:11 |
| xavierhardy | is it executor_thread_pool_size ? | 09:12 |
| xavierhardy | rakhmerov , d0ugal? | 09:14 |
| mcdoker181818 | xavierhardy: Hi, If I understand correctly, no, you can't | 09:16 |
| xavierhardy | what makes you think that? | 09:16 |
| mcdoker181818 | executor_thread_pool_size is number of green threads, which handle a rabbitmq messages | 09:16 |
| xavierhardy | OK, but are those threads not handling the execution as well? Why would you need 64 (default value) threads if there's just one thread handling the execution? | 09:18 |
| mcdoker181818 | For example, if you set 5 that executor can handle only 5 actions at a time. | 09:20 |
| mcdoker181818 | What's problem to launch multiple instances? | 09:21 |
| mcdoker181818 | 64 (default value) threads are green threads inside one real OS thread | 09:27 |
| d0ugal | akovi: It would be good to move them forward. I forget what my comments on the review were :) | 09:28 |
| d0ugal | akovi: we also discussed this at the PTG and I don't remember everything we said either! | 09:29 |
| *** xavierhardy has quit IRC | 09:36 | |
| pgaxatte | is there a reason why the API server is able to parse XML? does mistral accept XML anywhere? apparently it is possible to DoS the api using XML entity expansion | 09:38 |
| d0ugal | pgaxatte: The API just supports both JSON and XML. | 09:39 |
| d0ugal | I do wonder if anyone uses XML :) | 09:39 |
| pgaxatte | ok | 09:39 |
| pgaxatte | yeah i guess noone uses it | 09:40 |
| pgaxatte | thanks d0ugal, i'll report | 09:40 |
| pgaxatte | bacl | 09:40 |
| pgaxatte | back | 09:40 |
| pgaxatte | our security team is checking mistral to see what could be exploited | 09:41 |
| d0ugal | Interesting. It will be good to hear of any other findings :) | 09:41 |
| pgaxatte | yeah absolutely | 09:42 |
| pgaxatte | I hope this helps everyone | 09:42 |
| pgaxatte | d0ugal: is there a way to completely disable xml support? | 09:54 |
| *** xavierhardy has joined #openstack-mistral | 09:55 | |
| pgaxatte | apparently at least nova and cinder don't support it anymore (or we have set something in conf that I'm not aware of) | 09:55 |
| d0ugal | pgaxatte: Not that I know of, but I have never tried. I'll take a look. | 09:56 |
| *** xavierhardy has quit IRC | 09:56 | |
| *** xavierhardy1 has joined #openstack-mistral | 09:56 | |
| pgaxatte | ok thanks d0ugal, i'll see on my side if I find anything | 09:58 |
| d0ugal | pgaxatte: I can't see anything obvious. | 09:58 |
| d0ugal | I guess the support comes with pecan by default? | 09:58 |
| xavierhardy1 | mcdoker181818 => OK that was my understanding | 09:59 |
| xavierhardy1 | so by the default, it can handle 64 actios at a time | 09:59 |
| d0ugal | pgaxatte: maybe it is a wsme feature actually... | 10:00 |
| pgaxatte | d0ugal: do you know if pecan is used in other openstack products? | 10:01 |
| d0ugal | pgaxatte: yes, is many of them | 10:01 |
| mcdoker181818 | xavierhardy1: if I understand correctly :) | 10:01 |
| pgaxatte | d0ugal: i'll try to compare with cinder to see where mistral differs because cinder does not support xml | 10:02 |
| xavierhardy1 | thanks | 10:03 |
| d0ugal | pgaxatte: it doesn't look like cinder uses pecan | 10:04 |
| jlejeune | akovi: thanks to your patch, I can today run actions on different openstack regions using ENV variables | 11:02 |
| jlejeune | but is there a way to define ENV variable directly in workflows ? | 11:03 |
| jlejeune | I 've tried to play with mistral environment but it doesn't work as expected | 11:03 |
| jlejeune | can I source ENV authentification variables somewhere else ? | 11:04 |
| jlejeune | to be accessible in workflows ? | 11:04 |
| akovi | jlejeune: no, currently, it is possible to run workflows against multiple regions but the auth parameters have to be passed over the API | 11:06 |
| akovi | and an execution can be targeted on a single region | 11:07 |
| akovi | :( | 11:07 |
| jlejeune | I see | 11:07 |
| akovi | rakhmerov: d0ugal: I'm trying to fix a concurrency issue in the tempest tests and have a question http://logs.openstack.org/66/582966/6/check/mistral-devstack/dd7e2ac/testr_results.html.gz | 11:11 |
| akovi | Now that the tempest tests are isolated from Mistral core, would it be possible to use the python-mistralclient instead of reimplementing it in tempest? | 11:12 |
| akovi | There is lots of tricking with urls and json parsing and... just feels bad | 11:13 |
| d0ugal | rakhmerov: I think we should have a release note for https://review.openstack.org/#/c/581671/ | 11:16 |
| d0ugal | rakhmerov: since there is a new config option | 11:16 |
| d0ugal | rakhmerov: but I think we can add that in a follow up patch. | 11:17 |
| openstackgerrit | Andras Kovi proposed openstack/mistral-tempest-plugin master: Always use id of workflow instead of name https://review.openstack.org/583550 | 11:22 |
| d0ugal | akovi: I think tempest encourages projects to not use their own clients, but I forget why | 11:24 |
| d0ugal | akovi: I would be fine with using mistralclient, but I guess we want some simple rest client tests still | 11:24 |
| akovi | d0ugal: that's my train of thoughts here as well haha | 11:27 |
| d0ugal | akovi: Our tempest tests are horrible. I don't really know how to improve them because I think it is a huge job. | 11:28 |
| d0ugal | We also need to move the tempest tests from mistralclient | 11:28 |
| pgaxatte | d0ugal: neutron seems to use pecan we'll try against it | 12:00 |
| d0ugal | Good idea | 12:03 |
| openstackgerrit | Brad P. Crochet proposed openstack/mistral master: Install mistral-lib as a sibling for tox jobs https://review.openstack.org/583232 | 12:17 |
| pgaxatte | d0ugal: neutron does not respond in XML ever | 12:20 |
| pgaxatte | whatever Accept header I put it always answers in JSON | 12:21 |
| d0ugal | Interesting, so we need to see how they do it :) | 12:21 |
| openstackgerrit | Brad P. Crochet proposed openstack/mistral master: Add user creds trusts to db https://review.openstack.org/462056 | 12:27 |
| pgaxatte | d0ugal: there does not seem to be any support for xml in pecan.... I dont understand *what* answers in XML | 12:50 |
| pgaxatte | d0ugal: ok as you suggested it's apparently wsme | 12:55 |
| pgaxatte | the resources are inherited from it and it seems to be able to serialize to XML | 12:55 |
| *** hardikjasani has quit IRC | 13:10 | |
| *** toure|gone is now known as toure | 13:14 | |
| openstackgerrit | Merged openstack/mistral master: Fix rev-id parameter in mistral-db-manage https://review.openstack.org/583214 | 13:15 |
| pgaxatte | well this is scary: https://docs.python.org/3/library/xml.html#xml-vulnerabilities | 13:22 |
| d0ugal | :) | 13:23 |
| d0ugal | pgaxatte: https://github.com/openstack/neutron/blob/master/neutron/api/v2/resource.py#L41-L43 | 13:24 |
| d0ugal | I suspect that is what controls it | 13:24 |
| pgaxatte | d0ugal: yes it is forced to answer in json | 13:25 |
| d0ugal | Looks a bit like a hack :) | 13:25 |
| d0ugal | pgaxatte: https://github.com/openstack/neutron/commit/1b4ad95758c6cff7bc9a461f445c4b4e67964434 | 13:25 |
| d0ugal | It was removed 4 years ago :) | 13:25 |
| pgaxatte | with wsme you could tweak it from here: https://github.com/openstack/wsme/blob/master/wsme/rest/protocol.py#L18 | 13:25 |
| d0ugal | s/It/XML/ | 13:25 |
| pgaxatte | yeah they did well | 13:26 |
| pgaxatte | d0ugal: I wonder if xml support could be removed from here: https://github.com/openstack/mistral/blob/master/mistral/api/controllers/v2/service.py | 13:28 |
| d0ugal | pgaxatte: possibly | 13:29 |
| pgaxatte | i'll try that | 13:30 |
| d0ugal | We need to make it optional and deprecate it first | 13:30 |
| *** _Adary has quit IRC | 13:30 | |
| *** bobh has joined #openstack-mistral | 13:35 | |
| openstackgerrit | Merged openstack/mistral master: Add the config option for Oslo Messaging executor type https://review.openstack.org/581671 | 13:39 |
| *** bobh has quit IRC | 13:40 | |
| *** bobh has joined #openstack-mistral | 13:57 | |
| *** bobh has quit IRC | 14:01 | |
| pgaxatte | d0ugal: yes that's what I had in mind | 14:10 |
| d0ugal | Great | 14:10 |
| *** akovi has quit IRC | 14:22 | |
| openstackgerrit | Brad P. Crochet proposed openstack/mistral master: Add user creds trusts to db https://review.openstack.org/462056 | 14:22 |
| *** gkadam is now known as gkadam-afk | 15:59 | |
| *** gkadam-afk is now known as gkadam | 16:23 | |
| *** AlexeyAbashkin has quit IRC | 16:36 | |
| *** AlexeyAbashkin has joined #openstack-mistral | 16:39 | |
| *** AlexeyAbashkin has quit IRC | 16:43 | |
| *** xavierhardy1 has quit IRC | 16:49 | |
| *** shardy has quit IRC | 17:15 | |
| *** gkadam has quit IRC | 17:38 | |
| *** jtomasek has quit IRC | 17:50 | |
| *** xavierhardy has joined #openstack-mistral | 18:51 | |
| *** harlowja has joined #openstack-mistral | 18:54 | |
| *** xavierhardy has quit IRC | 18:56 | |
| *** gkadam has joined #openstack-mistral | 20:45 | |
| *** gkadam has quit IRC | 21:04 | |
| *** threestrands has joined #openstack-mistral | 23:50 | |
| *** threestrands has quit IRC | 23:50 | |
| *** threestrands has joined #openstack-mistral | 23:50 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!