| *** Qiming has quit IRC | 01:12 | |
| *** flwang has quit IRC | 01:12 | |
| *** dims has quit IRC | 01:12 | |
| *** yanyanhu has joined #openstack-zun | 01:12 | |
| *** flwang has joined #openstack-zun | 01:13 | |
| *** Qiming has joined #openstack-zun | 01:14 | |
| *** dims has joined #openstack-zun | 01:14 | |
| openstackgerrit | Wenzhi Yu proposed openstack/zun: Enable etcd DB backend testing pipeline https://review.openstack.org/412756 | 01:23 |
|---|---|---|
| *** Wenzhi has joined #openstack-zun | 01:28 | |
| *** Wenzhi has quit IRC | 01:44 | |
| *** Wenzhi has joined #openstack-zun | 01:44 | |
| *** jianyi has joined #openstack-zun | 02:02 | |
| *** zhurong has joined #openstack-zun | 02:06 | |
| *** zhurong has quit IRC | 02:15 | |
| *** zhurong has joined #openstack-zun | 02:16 | |
| *** Wenzhi has quit IRC | 02:18 | |
| *** Wenzhi has joined #openstack-zun | 02:18 | |
| *** prameswar has joined #openstack-zun | 02:18 | |
| openstackgerrit | Kevin Zhao proposed openstack/python-zunclient: Add support for attach detach and resize https://review.openstack.org/417681 | 02:27 |
| openstackgerrit | Kevin Zhao proposed openstack/python-zunclient: Add support for attach detach and resize https://review.openstack.org/417681 | 02:41 |
| openstackgerrit | Kevin Zhao proposed openstack/python-zunclient: Add support for attach detach and resize https://review.openstack.org/417681 | 02:59 |
| openstackgerrit | huhaoran proposed openstack/zun: Code fix for bug#1652868 https://review.openstack.org/417341 | 02:59 |
| openstackgerrit | Kevin Zhao proposed openstack/python-zunclient: Add support for attach detach and resize https://review.openstack.org/417681 | 03:07 |
| *** kevinz has joined #openstack-zun | 03:09 | |
| *** yanyanhu has quit IRC | 03:13 | |
| *** hongbin has joined #openstack-zun | 03:15 | |
| *** yanyanhu has joined #openstack-zun | 03:18 | |
| *** zhurong has quit IRC | 03:43 | |
| *** janki has joined #openstack-zun | 03:43 | |
| jianyi | hi hongbin zun-api run in control node, zun-compute run in compute node, is that right? This is just my understanding. | 04:04 |
| hongbin | jianyi: yes, it is right | 04:04 |
| hongbin | jianyi: thanks for adding zun to kolla. appreciate your efforts :) | 04:05 |
| jianyi | hongbin,Thank you for your review,this is very helpful to me :) | 04:10 |
| hongbin | jianyi: good to hear | 04:10 |
| *** sheel has joined #openstack-zun | 04:11 | |
| hongbin | jianyi: btw, do you know what is the easiest way to setup a kolla-ansible dev environment (the quickstart guide looks very complicated) | 04:14 |
| openstackgerrit | Sudipta Biswas proposed openstack/zun: add the cpuset to container object https://review.openstack.org/417743 | 04:47 |
| *** hongbin has quit IRC | 04:50 | |
| openstackgerrit | Sudipta Biswas proposed openstack/zun: add the cpuset to container object https://review.openstack.org/417743 | 04:52 |
| *** Wenzhi has quit IRC | 05:00 | |
| *** Wenzhi has joined #openstack-zun | 05:00 | |
| *** Wenzhi has quit IRC | 05:05 | |
| *** zhurong has joined #openstack-zun | 05:27 | |
| *** diga has joined #openstack-zun | 05:28 | |
| openstackgerrit | Digambar proposed openstack/zun: Implement Cinder integration with zun https://review.openstack.org/417747 | 05:31 |
| jianyi | hongbin,i also refer to the quickstart guide , Don't know the orher way | 05:38 |
| *** Wenzhi has joined #openstack-zun | 05:44 | |
| openstackgerrit | Sudipta Biswas proposed openstack/zun: add cpuset to the docker driver https://review.openstack.org/417751 | 06:09 |
| openstackgerrit | Sudipta Biswas proposed openstack/python-zunclient: add cpusets during container creation https://review.openstack.org/417753 | 06:24 |
| openstackgerrit | Joe Cropper proposed openstack/zun: Fix exception handling https://review.openstack.org/417566 | 06:32 |
| *** zhangjl has joined #openstack-zun | 06:44 | |
| openstackgerrit | Maohaijun proposed openstack/zun: Missing import '_' in file zun/container/driver.py https://review.openstack.org/417764 | 07:15 |
| *** openstackgerrit has quit IRC | 08:18 | |
| *** pksingh has joined #openstack-zun | 08:29 | |
| *** pksingh has quit IRC | 08:31 | |
| *** Wenzhi has quit IRC | 08:45 | |
| *** jwcroppe has quit IRC | 08:48 | |
| *** sudipto has joined #openstack-zun | 08:51 | |
| *** sudipto_ has joined #openstack-zun | 08:51 | |
| *** vikasc has quit IRC | 08:53 | |
| *** flaper87 has joined #openstack-zun | 08:55 | |
| *** vikasc has joined #openstack-zun | 08:55 | |
| *** prameswar has quit IRC | 09:01 | |
| *** jianyi has quit IRC | 09:04 | |
| sudipto_ | mkrai, hi, are you there? | 09:38 |
| *** zhurong has quit IRC | 09:40 | |
| mkrai | hi sudipto_ | 09:42 |
| sudipto_ | mkrai, if i want to convert a value that's given by the user into another type - what's the best place to do it? I am referring to the user entering a string representing a "cpuset", but while storing that "cpuset", I want to store it as a listofinterger | 09:43 |
| sudipto_ | the validation of the string being correct or not also can be determined by whether it's convertible into a list or not. | 09:44 |
| sudipto_ | The current validation framework we have, does not seem to allow me to do that... | 09:44 |
| mkrai | It is good to do at the API layer itself | 09:45 |
| mkrai | Because we create the db object there only | 09:45 |
| sudipto_ | mkrai, yeah, in the API Layer, i thought the validator could have a way to check that... | 09:46 |
| sudipto_ | otherwise, it can be base.py method...but not sure if that's a good thing. | 09:46 |
| mkrai | the jsonschema will validate it. Isn't it? | 09:46 |
| sudipto_ | the jsonschema seems to validate things on 'type' | 09:47 |
| sudipto_ | i am trying to convert one 'type' to another 'type' and check if the first 'type' is correct or not | 09:47 |
| sudipto_ | but yeah, i think it should be possible in the jsonschema | 09:47 |
| mkrai | ok so I think the jsonschema can validate the acceptable type and then you can have another method to parse/change it to the desirable typr | 09:48 |
| mkrai | But should be done at API layer itself | 09:48 |
| sudipto_ | ok | 09:48 |
| mkrai | Yeah I haven't had look at jsonschema but I guess it should have a way | 09:49 |
| sudipto_ | looks like i can write a validator to validate the input, wondering i can store the converted input somehow or not.. | 09:53 |
| *** openstackgerrit has joined #openstack-zun | 09:55 | |
| openstackgerrit | Merged openstack/zun: Fix exception handling https://review.openstack.org/417566 | 09:55 |
| *** pksingh has joined #openstack-zun | 09:58 | |
| *** yanyanhu has quit IRC | 10:07 | |
| *** sudipto has quit IRC | 10:11 | |
| *** sudipto_ has quit IRC | 10:11 | |
| *** sudipto has joined #openstack-zun | 10:12 | |
| *** sudipto_ has joined #openstack-zun | 10:12 | |
| mkrai | Do you have patch uploaded for this sudipto_ ? | 10:12 |
| *** pksingh has quit IRC | 10:15 | |
| *** kevinz has quit IRC | 10:16 | |
| *** zhangjl has quit IRC | 10:16 | |
| *** prameswar has joined #openstack-zun | 10:21 | |
| sudipto_ | mkrai, writing one... | 10:24 |
| mkrai | ok I can have a look then and hopefully suggest something | 10:29 |
| openstackgerrit | Merged openstack/zun: Fix incorrect filtering at ZunService dbapi https://review.openstack.org/417539 | 10:35 |
| openstackgerrit | Merged openstack/zun: Missing import '_' in file zun/container/driver.py https://review.openstack.org/417764 | 10:35 |
| *** diga has quit IRC | 11:14 | |
| *** sudipto has quit IRC | 11:21 | |
| *** sudipto_ has quit IRC | 11:21 | |
| *** kevinz has joined #openstack-zun | 12:16 | |
| openstackgerrit | Wenzhi Yu proposed openstack/zun: Enable etcd DB backend testing pipeline https://review.openstack.org/412756 | 12:27 |
| *** janki has quit IRC | 12:38 | |
| *** catintheroof has joined #openstack-zun | 12:42 | |
| *** sudipto_ has joined #openstack-zun | 13:02 | |
| *** sudipto has joined #openstack-zun | 13:02 | |
| *** sudipto_ has quit IRC | 13:04 | |
| *** sudipto has quit IRC | 13:04 | |
| *** sudipto has joined #openstack-zun | 13:06 | |
| *** sudipto_ has joined #openstack-zun | 13:06 | |
| *** sudipto has quit IRC | 13:07 | |
| *** sudipto_ has quit IRC | 13:07 | |
| *** sudipto_ has joined #openstack-zun | 13:08 | |
| *** sudipto has joined #openstack-zun | 13:08 | |
| *** zhurong has joined #openstack-zun | 13:23 | |
| *** diga has joined #openstack-zun | 13:35 | |
| *** kevinz has quit IRC | 13:49 | |
| *** kevinz has joined #openstack-zun | 13:50 | |
| *** hongbin has joined #openstack-zun | 14:12 | |
| *** kevinz has quit IRC | 14:15 | |
| *** pksingh has joined #openstack-zun | 14:25 | |
| *** zhurong has quit IRC | 14:29 | |
| *** jwcroppe has joined #openstack-zun | 14:46 | |
| sudipto_ | hongbin, good morning. | 15:06 |
| hongbin | sudipto_: hey | 15:07 |
| openstackgerrit | Pradeep Kumar Singh proposed openstack/zun: Make start/stop/pause/unpause/restart/kill action async https://review.openstack.org/416904 | 15:07 |
| sudipto_ | I registered a BP today: https://blueprints.launchpad.net/zun/+spec/cpuset-container - this is just for adding the cpuset as an option to zun create - the scheduler/capability part will be done subsequently. what do you think? | 15:07 |
| sudipto_ | I wrote some code too for that, but i have some conflicts w.r.t resolving one issue, that i can discuss here...with pksingh being around. | 15:08 |
| hongbin | sudipto_: approve the bp | 15:08 |
| pksingh | hello sudipto, hongbin | 15:09 |
| hongbin | sudipto_: sure, go ahead, (i am not sure what conflict yet) | 15:09 |
| sudipto_ | hongbin, thanks. Now to the implementation problem i have been having is w.r.t the way we would want to provide the user this option for cpuset. I assume we would want the user to enter a string which would be something like "1,2,3-6" | 15:10 |
| sudipto_ | But in order for us to build on exclusive cpusets later, we would want this string value to be stored as a "Set" in the DB - that way we can compare the sets for exclusivity etc... | 15:10 |
| sudipto_ | The validation logic we have right now, can do the check for a proper string in the API layer, but in the API layer, i would also like to covert this above mentioned string to something like: [1, 2, 3, 4, 5, 6] and store it against the container object. | 15:11 |
| sudipto_ | What should be the best possible way to do this? | 15:11 |
| hongbin | i think i need to review the patch before answering the question :) | 15:12 |
| sudipto_ | The patchset is a bit incomplete coz of this. | 15:13 |
| pksingh | sudipto_: you can write your own validators, and validate the input | 15:13 |
| hongbin | i haven't looked into the details yet, but it seems here is an option: | 15:14 |
| hongbin | in the cli, support "1,2,3-6", in the rest api support [1,2,3,4,5,6] | 15:15 |
| sudipto_ | pksingh, yeah, i thought of writing one like that "maximum" "minimum" thing you have in there for now, but i was thinking if i could write a function like this: https://github.com/openstack/nova/blob/32a864fd17d4d08df899d1f4c2a143febed5da92/nova/virt/hardware.py#L57 and then have that validate the input | 15:15 |
| sudipto_ | hongbin, i think then you are suggesting that the conversion from string to list, can be done in the zunclient itself right? | 15:15 |
| sudipto_ | (i think that's a good suggestion) | 15:15 |
| hongbin | sudipto_: yes | 15:16 |
| sudipto_ | Alright sounds good. pksingh - what's the way to validate a list? | 15:16 |
| sudipto_ | parameter_type per say... | 15:16 |
| pksingh | sudipto_: https://spacetelescope.github.io/understanding-json-schema/reference/array.html | 15:17 |
| sudipto_ | that helps :)_ | 15:17 |
| sudipto_ | thank you!! BTW, if you really want a spec i could write one...but I feel that a spec is worthy for a feature when we write the exclusivity and scheduler part, what do you think? | 15:18 |
| pksingh | sudipto_: no sir :), i just wanted to understand the feature more, may be i need to look more in documentation for understanding the functionality first | 15:19 |
| sudipto_ | pksingh, https://agileek.github.io/docker/2014/08/06/docker-cpuset/ something like this can be done post this is in | 15:20 |
| pksingh | sudipto_: i thought you are binding a container to a cpu core, i dont know whether i am correct or not | 15:20 |
| hongbin | sudipto_: i think you could write down the document in the bp (instead of writting a spec) | 15:20 |
| sudipto_ | hongbin, sounds fair, i will write more details there. | 15:21 |
| sudipto_ | pksingh, yeah binding the core to container, but not giving any exclusivity yet. | 15:21 |
| pksingh | hongbin: sudipto, please review my patch https://review.openstack.org/#/c/416904/, i need to rebase it again and again, since it changes the manager :) | 15:22 |
| sudipto_ | pksingh, aye aye sir! | 15:22 |
| hongbin | it conflicts with 5 patches :) | 15:23 |
| hongbin | ok, let's merge it as long as the gate is clear | 15:23 |
| pksingh | hongbin: :) | 15:23 |
| sudipto_ | done | 15:24 |
| pksingh | sudipto_: thannks for the link, will look into it. Man you english is too good, sometimes i need to look internet | 15:24 |
| pksingh | sudipto_: :) | 15:24 |
| sudipto_ | pksingh, i would tread that compliment carefully :) | 15:25 |
| * hongbin is going to drive to office and will be offline in about half an hour | 15:26 | |
| *** hongbin has quit IRC | 15:26 | |
| pksingh | sudipto_: can user select the CPU cores in public cloud, where they don't know about the compute hosts, just random thought | 15:27 |
| *** pksingh has quit IRC | 15:29 | |
| *** pksingh has joined #openstack-zun | 15:30 | |
| pksingh | sudipto_: can user select the CPU cores in public cloud, where they don't know about the compute hosts, just random thought | 15:34 |
| *** pksingh has quit IRC | 15:46 | |
| openstackgerrit | Merged openstack/zun: Fix the incorrect initialization of context roles https://review.openstack.org/417648 | 15:46 |
| sudipto_ | pksingh, good question, ideally not. This would be more of an API capability that docker presently supports. This would be just an optional field that could be set for specific cases.. | 16:00 |
| sudipto_ | However if you have some thoughts around that, i am more than happy to hear :) | 16:00 |
| openstackgerrit | Merged openstack/zun: Make start/stop/pause/unpause/restart/kill action async https://review.openstack.org/416904 | 16:06 |
| openstackgerrit | Digambar proposed openstack/zun: Implement Cinder integration with zun https://review.openstack.org/417747 | 16:08 |
| *** jayaura has quit IRC | 16:19 | |
| *** mkrai has quit IRC | 16:19 | |
| sudipto_ | from the cloud use case per say - it should be something like enforce_exclusive_cpusets (with a value of how many crusts)- and then zun would be making sure that happens via this cpuset support... | 16:22 |
| *** hongbin has joined #openstack-zun | 16:23 | |
| *** Qiming has quit IRC | 16:34 | |
| openstackgerrit | Namrata proposed openstack/python-zunclient: Add OSC Plugin for openstack appcontainer exec https://review.openstack.org/417964 | 16:39 |
| *** Qiming has joined #openstack-zun | 16:40 | |
| openstackgerrit | Sudipta Biswas proposed openstack/zun: add cpuset to the docker driver https://review.openstack.org/417751 | 16:46 |
| *** Guest10938 has joined #openstack-zun | 16:52 | |
| *** sudipto has quit IRC | 17:32 | |
| *** sudipto_ has quit IRC | 17:33 | |
| *** mkrai has joined #openstack-zun | 17:33 | |
| *** sudipto_ has joined #openstack-zun | 17:39 | |
| *** sudipto has joined #openstack-zun | 17:39 | |
| *** sudipto has quit IRC | 17:39 | |
| *** sudipto_ has quit IRC | 17:39 | |
| openstackgerrit | Namrata proposed openstack/python-zunclient: Add OSC Plugin for openstack appcontainer logs https://review.openstack.org/418007 | 18:18 |
| *** harlowja_ has joined #openstack-zun | 18:31 | |
| *** harlowja has quit IRC | 18:31 | |
| *** prameswar has quit IRC | 18:32 | |
| openstackgerrit | Namrata proposed openstack/python-zunclient: Add OSC Plugin for openstack appcontainer kill https://review.openstack.org/418060 | 19:18 |
| openstackgerrit | Namrata proposed openstack/python-zunclient: Add OSC Plugin for openstack appcontainer stop https://review.openstack.org/418086 | 20:03 |
| openstackgerrit | Namrata proposed openstack/python-zunclient: Add OSC Plugin for openstack appcontainer run https://review.openstack.org/418103 | 20:24 |
| *** catinthe_ has joined #openstack-zun | 20:59 | |
| *** catintheroof has quit IRC | 21:00 | |
| *** catinthe_ has quit IRC | 22:24 | |
| *** harlowja_ has quit IRC | 23:40 | |
| *** harlowja has joined #openstack-zun | 23:41 | |
| *** diga has quit IRC | 23:51 | |
| *** diga has joined #openstack-zun | 23:52 | |
| diga | hongbin: Hi | 23:52 |
| hongbin | diga: sup | 23:52 |
| diga | hongbin: as you commented in the spec, Fuxi is not supporting multi tenancy | 23:53 |
| hongbin | diga: yes | 23:54 |
| hongbin | diga: it doesn't support multi-tenancy | 23:54 |
| diga | hongbin: If we need tenancy support in fuxi, we can add it in future ? that is doable correct ? | 23:54 |
| hongbin | diga: i am not sure exactly. right now, i couldn't think of a feasible approach to support multi-tenancy in fux | 23:56 |
| diga | hongbin: okay | 23:56 |
| diga | hongbin: we can start with Fuxi now, but in this case, we want to implement volume support such a way that it is extendable in future to support multiple volume dirver | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!