| *** CerealBox has joined #openstack-ironic | 00:03 | |
| CerealBox | I've updated the "troubleshooting" section of https://wiki.openstack.org/wiki/Baremetal with some additional information (thanks NobodyCam!) | 00:04 |
|---|---|---|
| NobodyCam | CerealBox: Awesome ... my I suggest one thing | 00:05 |
| CerealBox | yes please. I wanted to find out how to fix the formatting, BTW | 00:05 |
| NobodyCam | can you remove any info that is releated to you ... ie your ips and user names | 00:06 |
| NobodyCam | that more so that people dont have you info! | 00:07 |
| NobodyCam | s/you/your/ | 00:07 |
| CerealBox | I already did. Do you still see any? The IPs still showing up is for illustration only | 00:07 |
| NobodyCam | ahh I just saw this: ipmitool -I lanplus -H 7.7.7.4 -U ADMIN -f /tmp/tmp1mDmnB power status | 00:08 |
| NobodyCam | :-p | 00:08 |
| CerealBox | Yes, maybe I should have used 192.168.x.y instead | 00:09 |
| CerealBox | The log files entries are all messed up due to the wiki formatting (all CRLF removed). How do I fix this? I tried the editing options that the wiki page provides, but they didn't work for me | 00:09 |
| clarkb | the wiki allows for literal text | 00:09 |
| NobodyCam | CerealBox: formatting like the one I just updated | 00:10 |
| NobodyCam | :-p | 00:10 |
| NobodyCam | CerealBox: take a look at the agenda https://wiki.openstack.org/wiki/Meetings/Ironic | 00:11 |
| NobodyCam | do a edit on it.. it has a far part of the formatting going on | 00:12 |
| NobodyCam | s/part/bit/ | 00:12 |
| CerealBox | thanks, looking at it now | 00:13 |
| * NobodyCam notes that is now after 5 and all of his answers from this point will include the --beer --type: apple options | 00:14 | |
| Haomeng | CerealBox: one question, can we use devstack to depoly our nova baremetal node, I found devstack support it, and add devstack steps into our document | 00:17 |
| NobodyCam | Haomeng: devstack will support nova baremetal but ironic is fully intrgrated with it yet | 00:18 |
| CerealBox | Haomeng: Yes, devstack should work too, but honestly I don't know for sure because I didn't try | 00:18 |
| NobodyCam | ironic is NOT fully... | 00:18 |
| Haomeng | I have try, encounter some br99 bridge issue,willtry | 00:24 |
| *** matsuhashi has joined #openstack-ironic | 00:31 | |
| NobodyCam | br99 sounds like a network bridge | 00:32 |
| Haomeng | yes | 00:32 |
| Haomeng | clarkb:ping | 00:32 |
| Haomeng | looks like dev-quickstart.rst#n29 line29 is for Fedora sys, but my sys is Ubuntu | 00:33 |
| Haomeng | this is for ubuntu - "sudo apt-get install python-dev swig libssl-dev python-pip libmysqlclient-dev libxml2-dev libxslt-dev libpq-dev" | 00:33 |
| Haomeng | not sure if we can add 'hacking' here | 00:33 |
| Haomeng | or I just add new lineto install 'hacking' | 00:34 |
| CerealBox | thanks clarkb, NobodyCam. I've updated the wiki to use 192.168.* addresses and the formatting looks much better now. | 00:40 |
| NobodyCam | :) awesome Ty CerealBox | 00:40 |
| CerealBox | Thank YOU for the awesome help over the past few days. I'm planning to update to Havana next, and try this again. | 00:41 |
| NobodyCam | :) let us know how it goes! | 00:42 |
| CerealBox | will do! | 00:42 |
| *** nosnos has joined #openstack-ironic | 00:43 | |
| *** WillOchanda has quit IRC | 00:54 | |
| * NobodyCam wanders away | 00:56 | |
| clarkb | Haomeng: you don't need nose | 00:57 |
| clarkb | even on fedora you shouldn't need it | 00:57 |
| *** CerealBox is now known as CerealBox_ | 01:06 | |
| *** CerealBox_ is now known as CerealBox | 01:07 | |
| *** CerealBox is now known as CerealBox_ | 01:07 | |
| *** CerealBox_ is now known as CerealBox_Away | 01:09 | |
| *** WillOchanda has joined #openstack-ironic | 01:19 | |
| *** CerealBox_Away has left #openstack-ironic | 01:20 | |
| *** sjing has joined #openstack-ironic | 01:24 | |
| openstackgerrit | linggao proposed a change to openstack/python-ironicclient: Modifies CLI to show nodes by instance uuid https://review.openstack.org/53485 | 01:25 |
| *** anniec has quit IRC | 01:25 | |
| *** linggao has quit IRC | 01:30 | |
| *** anniec has joined #openstack-ironic | 01:31 | |
| *** WillOchanda has quit IRC | 01:32 | |
| *** anteaya has quit IRC | 01:37 | |
| *** anniec has quit IRC | 01:41 | |
| *** anniec has joined #openstack-ironic | 01:47 | |
| Haomeng | clarkb: yes, but I am running on Ubuntu | 01:58 |
| clarkb | Haomeng: that shouldn't matter | 01:59 |
| clarkb | ubuntu or fedora, you don't need nose. You can use nose to run the tests but testr is the preferred and tested way | 01:59 |
| Haomeng | but I am confusing, why I follow the step, but the H101 is missed by local flake8 | 01:59 |
| clarkb | because H101 comes from hacking | 02:00 |
| clarkb | you need to make sure hacking is installed. The best way to do that is to run the tests with tox. `tox` | 02:00 |
| Haomeng | I checked, the 'hacking' is defined as dependency in our test-requirements, I think once we run 'tox -evenv' or 'testr init', the required lib should be installed, but it is not true case in my ubuntu dev env | 02:02 |
| Haomeng | let me reinstall another ubuntu to have a look, thank you clarkt | 02:03 |
| clarkb | testr init won't install the dependencies | 02:04 |
| clarkb | and if you use tox -evenv, you have to source the venv after wards | 02:05 |
| clarkb | the easiest way to run the tests is literally just `tox` | 02:05 |
| Haomeng | yes, I follow the steps | 02:05 |
| Haomeng | Setting up a local environment for development can be done with tox: | 02:06 |
| Haomeng | 02:06 | |
| Haomeng | # create virtualenv | 02:06 |
| Haomeng | tox -evenv -- echo 'done' | 02:06 |
| Haomeng | 02:06 | |
| Haomeng | # activate the virtualenv | 02:06 |
| Haomeng | source .tox/venv/bin/activate | 02:06 |
| Haomeng | 02:06 | |
| Haomeng | # run testr init | 02:06 |
| Haomeng | testr init | 02:06 |
| clarkb | Haomeng: maybe you ran that before hacking became a dependency? | 02:07 |
| Haomeng | I want to fix this document to make sure other guys will not encounter such same with hacking issue with me | 02:12 |
| clarkb | Haomeng: the way to do that is to just use `tox` | 02:12 |
| Haomeng | now your work-around works, just run 'pip install hacking' can fix my issue, but not sure how to change the dev guide, so need your comments, thank you | 02:12 |
| Haomeng | yes | 02:13 |
| *** rloo has quit IRC | 02:45 | |
| openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: Support uniform MAC address with colons https://review.openstack.org/52761 | 03:21 |
| *** anniec has quit IRC | 03:23 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Conductors maintan driver list in the DB https://review.openstack.org/51743 | 03:41 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Fix a few missing i18n https://review.openstack.org/53491 | 03:41 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Remove several old TODO messages https://review.openstack.org/53492 | 03:42 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Remove redundant test stubs from conductor/manager https://review.openstack.org/53493 | 03:42 |
| *** prekarat has quit IRC | 04:23 | |
| *** rwsu has quit IRC | 05:42 | |
| *** matsuhashi has quit IRC | 06:04 | |
| openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: Use assertItemsEqual to compare the list https://review.openstack.org/53522 | 06:08 |
| *** matsuhashi has joined #openstack-ironic | 06:09 | |
| *** rwsu has joined #openstack-ironic | 06:21 | |
| *** rwsu has quit IRC | 06:26 | |
| *** prekarat has joined #openstack-ironic | 06:34 | |
| *** jianingy1afk has joined #openstack-ironic | 06:45 | |
| *** jianingy_afk has quit IRC | 06:45 | |
| *** jianingy_afk has joined #openstack-ironic | 06:45 | |
| *** jianingy1afk has quit IRC | 06:49 | |
| *** matsuhashi has quit IRC | 06:54 | |
| *** matsuhashi has joined #openstack-ironic | 06:55 | |
| *** matsuhashi has quit IRC | 06:59 | |
| *** matsuhashi has joined #openstack-ironic | 07:01 | |
| GheRivero | morning all | 07:39 |
| *** romcheg has joined #openstack-ironic | 07:39 | |
| *** jistr has joined #openstack-ironic | 07:50 | |
| *** martyntaylor has joined #openstack-ironic | 07:55 | |
| *** lucasagomes has joined #openstack-ironic | 08:20 | |
| *** derekh has joined #openstack-ironic | 08:22 | |
| Haomeng | lucasagomes: ping | 08:33 |
| Haomeng | GheRivero: morning:) | 08:34 |
| lucasagomes | Haomeng, morning buddy, how r u? | 08:35 |
| GheRivero | morning all | 08:35 |
| Haomeng | Lucus, need your help, do you have idea about this defect https://bugs.launchpad.net/ironic/+bug/1244072, not sure why port object works with default value, but node does not work, strange:) | 08:35 |
| lucasagomes | Haomeng, oh, off the top of my head I don't know, gotta dig in the code a bit to see what's going on | 08:38 |
| Haomeng | no problem, ok will go on to check code:) | 08:38 |
| Haomeng | and another findings, some port properties can not be listed by command line, is there some display configurations to control the output, any comments? | 08:38 |
| romcheg | Morning all | 08:39 |
| romcheg | Sorry folks, I got sick so was unavailable for the last two days | 08:39 |
| Haomeng | morning, romcheg | 08:40 |
| romcheg | Morning Haomeng | 08:40 |
| romcheg | I heard you have a question for me | 08:41 |
| Haomeng | Just find helping, if you can help me thankyou | 08:41 |
| Haomeng | two defects, no idea how to fix now, any comments? https://bugs.launchpad.net/ironic/+bug/1244072 https://bugs.launchpad.net/ironic/+bug/1244107 | 08:42 |
| Haomeng | these are tested out by me today:) | 08:42 |
| Haomeng | want to fix by myself:) | 08:42 |
| Haomeng | welcome all of your comments:_ | 08:43 |
| romcheg | Ok, let's take a look | 08:43 |
| Haomeng | thank you romcheg:) | 08:43 |
| Haomeng | no rush, if you have time:) | 08:44 |
| lucasagomes | Haomeng, oh I see, weird, port-show should show all the port attributes (https://github.com/openstack/ironic/blob/master/ironic/api/controllers/v1/port.py#L48-L54) | 08:48 |
| lucasagomes | romcheg, welcome back :) | 08:48 |
| lucasagomes | romcheg, feeling better now buddy? | 08:48 |
| Haomeng | ok, let me debug it to find the bugs:) | 08:49 |
| romcheg | Hi lucasagomes. Better now but have to attend my doctor for a few days more | 08:49 |
| romcheg | How're you? | 08:49 |
| *** sjing has quit IRC | 08:50 | |
| Haomeng | maybe it is filted by IronicClient code | 08:50 |
| Haomeng | let me try with api call | 08:50 |
| romcheg | Haomeng: you're always welcome to fix bugs :) | 08:50 |
| lucasagomes | romcheg, I see :) | 08:50 |
| lucasagomes | all is well here | 08:50 |
| Haomeng | yes, I am focusing on Ironic these days, and our boss let me work with Ironic team full time to do more Contribution | 08:50 |
| Haomeng | but I am new guy, need all of you supporting:) | 08:51 |
| romcheg | Haomeng, lucasagomes: I think the bug might be in the client so we need to add that project to it, agree? | 08:51 |
| Haomeng | thanks for all supporting:) | 08:51 |
| Haomeng | yes | 08:51 |
| Haomeng | let me do a API call first to compare the result first | 08:51 |
| romcheg | Haomeng: that's a nice idea | 08:52 |
| Haomeng | :) | 08:52 |
| lucasagomes | romcheg, the port show? yes sir! | 08:52 |
| Haomeng | I tested with API call, it can return these properteis: | 08:55 |
| Haomeng | { | 08:55 |
| Haomeng | "node_id": "be83a857-aa23-493b-b13a-f01ebfcef68c", | 08:55 |
| Haomeng | "extra": {}, | 08:55 |
| Haomeng | "uuid": "0ad5d609-3072-4499-97fc-ab2d666313b8", | 08:55 |
| Haomeng | "links": [ | 08:55 |
| Haomeng | { | 08:55 |
| Haomeng | "href": "http://9.123.136.76:6385/v1/ports/0ad5d609-3072-4499-97fc-ab2d666313b8", | 08:55 |
| Haomeng | "rel": "self" | 08:55 |
| Haomeng | }, | 08:55 |
| Haomeng | { | 08:55 |
| Haomeng | "href": "http://9.123.136.76:6385/ports/0ad5d609-3072-4499-97fc-ab2d666313b8", | 08:55 |
| Haomeng | "rel": "bookmark" | 08:55 |
| Haomeng | } | 08:55 |
| romcheg | Haomeng: please use http://paste.openstack.org for big snippets | 08:56 |
| Haomeng | thanks | 08:56 |
| lucasagomes | does nodes return the internal attributes as well (created_at, updated_at) ? | 08:57 |
| Haomeng | sorry, I dont know such way to paste a lot of information | 08:57 |
| romcheg | lemme check that | 08:58 |
| Haomeng | port api call does not return created_at, updated_a | 08:58 |
| romcheg | I suspect the problem might be in the object model | 08:58 |
| Haomeng | but our node-show can return created_at, updated_at , but with empty value | 08:58 |
| Haomeng | let me check | 08:58 |
| Haomeng | thank you romcheg | 08:58 |
| romcheg | the object model seems to be fine | 09:04 |
| Haomeng | romcheg and lucasagomes, I will take care my Daughter, will be back after 1+ hours... | 09:07 |
| Haomeng | another finding | 09:07 |
| *** matsuhashi has quit IRC | 09:08 | |
| Haomeng | our node api call can not return created_at and updated_at | 09:08 |
| *** max_lobur has joined #openstack-ironic | 09:08 | |
| *** matsuhashi has joined #openstack-ironic | 09:09 | |
| Haomeng | but "ironic node-show" can return created_at and updated_at with empty value, strange | 09:09 |
| Haomeng | will be back... | 09:09 |
| lucasagomes | Haomeng, I think these are internal values that's why they are not shown | 09:09 |
| *** matsuhashi has quit IRC | 09:09 | |
| Haomeng | yes, agree with you Lucas | 09:09 |
| *** matsuhashi has joined #openstack-ironic | 09:09 | |
| Haomeng | but "ironic node-show" can return created_at and updated_at two fields, with empty value, strange | 09:09 |
| lucasagomes | Haomeng, https://github.com/openstack/python-ironicclient/blob/master/ironicclient/v1/node_shell.py#L24-L27 | 09:10 |
| Haomeng | yes, Lucas, this is client code, let me check | 09:11 |
| lucasagomes | the CLI wants to show the create_at and update_at, but as they r not present on the body message it shows an empty value | 09:11 |
| Haomeng | here is gap | 09:11 |
| Haomeng | api can not return create_at and update_at, but cli want to populate them | 09:11 |
| lucasagomes | yes | 09:11 |
| Haomeng | if create_at and update_at are required for us to show, we have to enable from api level | 09:12 |
| lucasagomes | exactly | 09:12 |
| Haomeng | I will try to debug API code and let you know the result | 09:12 |
| Haomeng | thank you Lucas | 09:12 |
| Haomeng | and check all our APIs | 09:12 |
| lucasagomes | now, should we show it in the API? as they r internal values we might need to discuss if there's any value in showing them or not | 09:12 |
| lucasagomes | :) | 09:13 |
| Haomeng | yes | 09:13 |
| Haomeng | if we should not show it by api, we have to remove the define from cli display list | 09:13 |
| lucasagomes | yes | 09:13 |
| romcheg | Is't that a topic for the meeting? | 09:13 |
| lucasagomes | same for the "reservation" it's a internal value used by ResourceManage to control the locks of the nodes | 09:14 |
| lucasagomes | romcheg, I don't think so really | 09:14 |
| lucasagomes | I think the CLI is wrong and just should be removed from there, I don't see any value in showing such attributes to be honest | 09:14 |
| Haomeng | and from my view, 'created_at', 'updated_at' are internal fields, but it is meaningful for us to check the object create time and latest updat time for investigations | 09:14 |
| Haomeng | ok | 09:14 |
| lucasagomes | yea maybe for some debug propose | 09:14 |
| Haomeng | yes | 09:15 |
| lucasagomes | well we can talk about it :) | 09:15 |
| Haomeng | :) | 09:15 |
| Haomeng | thank you Lucas and RomCheg | 09:15 |
| romcheg | maybe it's reasonable to show created_at and updated_at if the —debug option is specified& | 09:15 |
| romcheg | ? | 09:15 |
| Haomeng | I think created_at and updated_at are usefuy by my personal view:) | 09:16 |
| Haomeng | will be back, see you:) | 09:16 |
| lucasagomes | romcheg, yea... now thinking a bit more, yea such values might be kinda important, knowing when it was created is a good information hmmm | 09:20 |
| *** liteIRC has joined #openstack-ironic | 09:22 | |
| romcheg | lucasagomes: is there an easy way to show some fields only in debug mode? | 09:23 |
| romcheg | I'm not familiar with the client unfortunately | 09:24 |
| lucasagomes | oh the problem is that the API is not returning those fields so needs to be changed in the API first | 09:24 |
| lucasagomes | if you specify the --debug it will show the json from the request on the CLI but that won't contain the created_at, updated_at fields | 09:25 |
| max_lobur | Hi romcheg , lucasagomes , Haomeng | 09:29 |
| romcheg | Morning max_lobur | 09:29 |
| lucasagomes | max_lobur, morning :) | 09:29 |
| max_lobur | romcheg, to implement what you want we probably would need to send debug flag to the api in json | 09:30 |
| max_lobur | because currently --debug is purely client option | 09:30 |
| max_lobur | so the api will need to know whether to send created_at deleted_at back or not | 09:30 |
| romcheg | max_lobur: the API server accepts the debug option as well | 09:35 |
| romcheg | However, that should be done on start | 09:36 |
| *** liteIRC has quit IRC | 09:36 | |
| max_lobur | romcheg, true | 09:36 |
| lucasagomes | should those options appear only by debug propose? created_at and updated_at sounds like a relevant information even for not debug propose | 09:37 |
| romcheg | lucasagomes, max_lobur: Passing the date and time through the API is only the part of the task | 09:38 |
| romcheg | the second part is how and we represent the date to the user | 09:38 |
| romcheg | Both date and time a sensitive to the time zone | 09:39 |
| romcheg | So the problems start appearing when the client and the server are located in different time zones | 09:39 |
| max_lobur | romcheg, good point | 09:40 |
| romcheg | While it's an internal value, we should just use UTC | 09:40 |
| *** liteIRC has joined #openstack-ironic | 09:40 | |
| romcheg | If we want to make this an public information, however, we need to care about time zones | 09:40 |
| max_lobur | FYI nova shows created, update fields in show command, and it shows it in UTC | 09:41 |
| max_lobur | created | 2013-10-24T09:37:39Z | 09:41 |
| lucasagomes | romcheg, very good point yea | 09:41 |
| romcheg | We need to be better than nova :) | 09:42 |
| max_lobur | argee=) | 09:42 |
| lucasagomes | hehe | 09:42 |
| max_lobur | IMO possible way to show time in current TZ is to patch values from response according to the current time zone on the client | 09:45 |
| romcheg | Indeed | 09:45 |
| romcheg | That does not require changing the API | 09:46 |
| max_lobur | yep | 09:46 |
| *** liteIRC has quit IRC | 09:55 | |
| *** prekarat has quit IRC | 10:10 | |
| *** prekarat has joined #openstack-ironic | 10:22 | |
| Haomeng | I am back, ok, I will check API side to return these created_at and update_dt | 10:30 |
| Haomeng | then try to enhance client to display time value againsting local TZ | 10:31 |
| Haomeng | not sure if we have such fields definations in API side, like the client code _print_node_show method define the dislplay fields list | 10:40 |
| *** matsuhashi has quit IRC | 10:46 | |
| *** nosnos has quit IRC | 11:12 | |
| *** nosnos has joined #openstack-ironic | 11:12 | |
| *** nosnos has quit IRC | 11:17 | |
| romcheg | I think it will be also useful if NobodyCam and devananda read this discussion ^^ | 11:53 |
| Haomeng | :) | 11:54 |
| Haomeng | OK, will wait NobodyCam and devananda comments then to do fix | 11:55 |
| romcheg | Haomeng: I think you can do a patch. If the guys don't like it, they can always reject it | 11:58 |
| *** prekarat has quit IRC | 11:58 | |
| romcheg | Haomeng: in this case there's no reason to be afraid of rejections :) | 11:58 |
| *** prekarat has joined #openstack-ironic | 11:58 | |
| *** anteaya has joined #openstack-ironic | 12:07 | |
| Haomeng | romcheg, ok | 12:11 |
| Haomeng | :) | 12:11 |
| Haomeng | I think port fields returned by API are controled by Port class fields defination | 12:12 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/python-ironicclient: Add node-set-power-state to cli https://review.openstack.org/53356 | 12:20 |
| *** jdob has joined #openstack-ironic | 12:26 | |
| *** linggao has joined #openstack-ironic | 12:39 | |
| openstackgerrit | Jenkins proposed a change to openstack/ironic: Updated from global requirements https://review.openstack.org/53371 | 13:16 |
| *** jdob has quit IRC | 13:16 | |
| *** jdob has joined #openstack-ironic | 13:16 | |
| *** jbjohnso has joined #openstack-ironic | 13:22 | |
| *** rloo has joined #openstack-ironic | 13:32 | |
| *** jdob has quit IRC | 13:34 | |
| *** jdob has joined #openstack-ironic | 13:35 | |
| openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: node-show display created_at,updated_at empty value https://review.openstack.org/53627 | 13:37 |
| *** yuriyz has joined #openstack-ironic | 13:42 | |
| openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Fix exceptions handling in controllers https://review.openstack.org/53632 | 13:44 |
| romcheg | Haomeng: I just would like to remind you that this kind of change should be also covered by unittests | 13:45 |
| openstackgerrit | Yuriy Zveryanskyy proposed a change to openstack/ironic: Fix status codes in node controller https://review.openstack.org/52920 | 13:48 |
| Haomeng | I have run cli function tested already in my local env | 13:49 |
| Haomeng | and just commit first patch to review if can be accepted for our soution | 13:49 |
| Haomeng | will add unittest codes in next patchs | 13:49 |
| romcheg | Haomeng: no problem in that case :) | 13:51 |
| Haomeng | I will go to sleep, will add testcase code tomorrow morning,thanks for your reminder nomchg | 13:51 |
| Haomeng | :) | 13:51 |
| yuriyz | Good night Haomeng :) | 13:52 |
| romcheg | Good night | 13:52 |
| Haomeng | nice day, yuriyz, romcheg | 13:52 |
| Haomeng | :) | 13:52 |
| *** rloo_ has joined #openstack-ironic | 14:03 | |
| *** rloo__ has joined #openstack-ironic | 14:04 | |
| *** rloo___ has joined #openstack-ironic | 14:07 | |
| *** rloo has quit IRC | 14:07 | |
| *** rloo_ has quit IRC | 14:08 | |
| *** rloo__ has quit IRC | 14:09 | |
| *** rloo___ has quit IRC | 14:11 | |
| *** rwsu has joined #openstack-ironic | 14:20 | |
| *** prekarat1 has joined #openstack-ironic | 14:28 | |
| NobodyCam | Good morning Ironic | 14:31 |
| *** prekarat has quit IRC | 14:31 | |
| max_lobur | Morning NobodyCam =) | 14:32 |
| NobodyCam | hey hey max_lobur how is your day going. well I hope:) | 14:32 |
| max_lobur | so far so good =) | 14:33 |
| NobodyCam | :) | 14:34 |
| yuriyz | Morning Ironic | 14:34 |
| NobodyCam | mornig yuriyz | 14:35 |
| NobodyCam | morning even | 14:35 |
| linggao | Morning all. | 14:36 |
| NobodyCam | morning linggao | 14:36 |
| linggao | Hi Haomeng, I saw your comment on my code regarding sync with doc. Is that something I should change in code? | 14:36 |
| linggao | Morning NobodyCam | 14:37 |
| linggao | NobodyCam, devananda, do we have a common API doc that I can update for the changes I have made /nodes/?instance_uud=<something> ? | 14:41 |
| NobodyCam | linggao: this ? http://docs.openstack.org/developer/ironic/ ? | 14:43 |
| linggao | NobodyCam, I am not sure this is the one that Haomeng talking about. | 14:47 |
| NobodyCam | linggao: what review? | 14:47 |
| linggao | https://review.openstack.org/#/c/53262/ | 14:47 |
| NobodyCam | linggao: http://docs.openstack.org/developer/ironic/ | 14:53 |
| NobodyCam | bad paste | 14:53 |
| NobodyCam | linggao: http://docs.openstack.org/developer/ironic/dev/api-spec-v1.html | 14:53 |
| linggao | NobodyCam, that looks like the right one. | 14:54 |
| NobodyCam | :-p | 14:54 |
| NobodyCam | :0 | 14:55 |
| linggao | Should I update it after the patch is approved? | 14:55 |
| linggao | Or we have someone dedicated to update it? | 14:55 |
| NobodyCam | we do not have a dedicated person | 14:55 |
| linggao | hmm, wonder how I can update it. There is not even a button for me to login. :) Maybe only core people can do it. | 14:59 |
| *** rloo has joined #openstack-ironic | 15:01 | |
| NobodyCam | linggao: here maybe https://github.com/openstack/ironic/blob/master/doc/source/dev/api-spec-v1.rst | 15:02 |
| NobodyCam | add any edits to that file along with the review | 15:03 |
| linggao | NobodyCam, thanks a lot. | 15:03 |
| NobodyCam | :) | 15:04 |
| *** prekarat1 has quit IRC | 15:10 | |
| * NobodyCam makes mo coffee... brb | 15:16 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Required fields on nodes https://review.openstack.org/53664 | 15:24 |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Expose /drivers on the API https://review.openstack.org/53157 | 15:24 |
| NobodyCam | morning lucasagomes | 15:24 |
| lucasagomes | NobodyCam, good morning :) how r u? | 15:25 |
| NobodyCam | doing good.... slowly getting use this thing called "The Sun"... | 15:25 |
| lucasagomes | nice! here it's getting cold now | 15:26 |
| lucasagomes | and no sun, just cloudy | 15:26 |
| * lucasagomes I actually likes it | 15:27 | |
| linggao | NobodyCam, I do not even know how to start change that doc https://github.com/openstack/ironic/blob/master/doc/source/dev/api-spec-v1.rst | 15:28 |
| linggao | it is not up-to-date at all. | 15:28 |
| NobodyCam | :) lol only going to be 90F here today :) | 15:28 |
| NobodyCam | linggao: :( | 15:29 |
| lucasagomes | linggao, :( yea it's not | 15:29 |
| lucasagomes | wow 90F is like 30C, that's pretty hot | 15:29 |
| NobodyCam | hehehe :) ahh palm springs | 15:29 |
| lucasagomes | here it's 14C right now (57F) | 15:29 |
| *** ben_duyujie has joined #openstack-ironic | 15:29 | |
| linggao | I'll just ignore it for now with my patch, can I? | 15:29 |
| NobodyCam | linggao: how far off is that doc | 15:30 |
| NobodyCam | lucasagomes: Mr J. didn't like the expose driver patch | 15:31 |
| NobodyCam | lucasagomes: also dose that one need a doc update? | 15:31 |
| linggao | NobodyCam, take Node for example http://docs.openstack.org/developer/ironic/dev/api-spec-v1.html#node | 15:31 |
| linggao | A lot of attributes mentioned there are not in ironic. | 15:32 |
| lucasagomes | NobodyCam, it because it depends on the patch that add the list of drivers to the db | 15:32 |
| lucasagomes | and that's failing on py26 cause it uses one assert that is only supported by py27+ | 15:32 |
| linggao | arch, cpu, mem, ram, state etc. | 15:32 |
| linggao | I am afraid to change it. | 15:33 |
| lucasagomes | linggao, yea pretty much the whole thing should be updated | 15:34 |
| NobodyCam | linggao: ack .. please comment on the review that the doc still need updating | 15:34 |
| linggao | NobodyCam, ok. will do. thanks | 15:34 |
| NobodyCam | prob need a bug for api-spec-doc way out of date | 15:34 |
| NobodyCam | :) | 15:34 |
| NobodyCam | brb | 15:34 |
| lucasagomes | we have one I think | 15:35 |
| rloo | hi, quick question. As a general rule, should we LOG something when an exception is raised? | 15:35 |
| lucasagomes | if not I'm going to create | 15:35 |
| linggao | thanks lucasagomes. | 15:35 |
| rloo | or leave it to something that catches the exception, to maybe log it. | 15:35 |
| lucasagomes | https://bugs.launchpad.net/ironic/+bug/1202748 | 15:35 |
| linggao | ok, I'll put that link in the review. | 15:36 |
| NobodyCam | lucasagomes: :) | 15:37 |
| yuriyz | Hello lucasgomes rloo | 15:42 |
| rloo | hi yuriyz | 15:42 |
| openstackgerrit | linggao proposed a change to openstack/ironic: Supports get node by instance uuid in API https://review.openstack.org/53262 | 15:44 |
| yuriyz | Where is Martyn Taylor? He stop work with https://review.openstack.org/49582 | 15:44 |
| NobodyCam | martyntaylor: are you around? | 15:46 |
| martyntaylor | martyntaylor: I am around | 15:46 |
| martyntaylor | yuriyz: it's been inactive for week, going to try to get to is asap | 15:46 |
| NobodyCam | hey hey martyntaylor :) good mornign | 15:46 |
| NobodyCam | morning even | 15:47 |
| martyntaylor | NobodyCam: hey morning fella | 15:47 |
| NobodyCam | :) | 15:47 |
| martyntaylor | NobodyCam: well nearly 5pm here :D | 15:47 |
| NobodyCam | lol almost 9 here :) | 15:47 |
| yuriyz | martyntaylor, I like your idea for validating, only some code cleanup needed | 15:47 |
| martyntaylor | haha | 15:47 |
| martyntaylor | yuriyz: yeah mate, I haven't forgotten about it I assure you :P | 15:47 |
| martyntaylor | yuriyz: we just have a few bits of prep for openstack summit so I had to change priorities, but will get to it asap | 15:48 |
| martyntaylor | yuriyz: I will get to it before the end of the week ;) | 15:49 |
| NobodyCam | martyntaylor: awesome :) | 15:49 |
| * NobodyCam notes that is tomorrow is friday | 15:49 | |
| yuriyz | martyntaylog, good travel to HK :) | 15:49 |
| NobodyCam | heheheh | 15:49 |
| NobodyCam | yuriyz: are you comming? | 15:50 |
| martyntaylor | yuriyz: haha thanks :D | 15:50 |
| yuriyz | No, only romcheg | 15:50 |
| martyntaylor | doh :( | 15:50 |
| NobodyCam | :-P | 15:50 |
| romcheg | Morning everyone | 15:50 |
| romcheg | Good news everyone! | 15:50 |
| NobodyCam | morning romcheg :) | 15:51 |
| romcheg | Devstack now installs ironic client | 15:51 |
| NobodyCam | woot :) | 15:51 |
| romcheg | the authentication patch was merged as well | 15:51 |
| * NobodyCam high fives romcheg :) | 15:51 | |
| NobodyCam | bbt... brb | 16:01 |
| *** yuriyz has quit IRC | 16:06 | |
| *** yuriyz has joined #openstack-ironic | 16:11 | |
| *** jistr has quit IRC | 16:16 | |
| *** WillOchanda has joined #openstack-ironic | 16:26 | |
| openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/python-ironicclient: Add driver-list https://review.openstack.org/53683 | 16:27 |
| lucasagomes | romcheg, w00t! | 16:28 |
| devananda | g'morning, all | 16:33 |
| NobodyCam | good morning devananda | 16:33 |
| romcheg | Morning devananda | 16:34 |
| devananda | i seem to have a lot of reviews to catch up on ... :) | 16:34 |
| devananda | Haomeng: hi! sorry i missed your comments yesterday -- was on a plane | 16:34 |
| yuriyz | Morning devananda, I posted some comments on my keystone-related patchs | 16:35 |
| NobodyCam | devananda: http://eavesdrop.openstack.org/irclogs/%23openstack-ironic/ | 16:35 |
| NobodyCam | :) | 16:35 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Fix a few missing i18n https://review.openstack.org/53491 | 16:35 |
| *** martyntaylor has quit IRC | 16:37 | |
| *** ben_duyujie has quit IRC | 16:38 | |
| *** yuriyz has quit IRC | 16:39 | |
| NobodyCam | loging errors in test seems odd to me. | 16:39 |
| devananda | ok, still some catching up to do, but on the created_at, updated_at discussion | 16:41 |
| devananda | yes, our API should expose those for every resource | 16:41 |
| devananda | and I think client should list them by default | 16:41 |
| devananda | rloo: on exception logging, there are cases where we expect an exception to be caught at a higher layer of code and dealt with there | 16:45 |
| devananda | rloo: generally, we don't need to LOG when we raise, but we probably want to LOG when we catch the exception | 16:46 |
| devananda | lucasagomes: comments on API ^ | 16:46 |
| *** romcheg has left #openstack-ironic | 16:46 | |
| lucasagomes | devananda, I endorse ur statement :) as comments on https://review.openstack.org/#/c/53627/ | 16:48 |
| devananda | :) | 16:48 |
| lucasagomes | I think that those information should be exposed in the api as well | 16:48 |
| lucasagomes | one note... the CLI also is trying to show the "reserved" field | 16:49 |
| lucasagomes | on nodes | 16:49 |
| lucasagomes | I don't think that should be exposed, that's an internal flag to control the node locks | 16:49 |
| lucasagomes | the user shouldn't care about that | 16:49 |
| NobodyCam | lucasagomes: does 53356 imply a change to 52902? | 16:52 |
| devananda | lucasagomes: possibly, but it could be useful for troubleshooting | 16:53 |
| lucasagomes | NobodyCam, it does, I the chassis-list one also | 16:53 |
| devananda | lucasagomes: eg, knowing which conductor locked a node and, oh hey, that conductor is offline .... | 16:53 |
| lucasagomes | NobodyCam, I have to update that man page in general | 16:53 |
| devananda | lucasagomes: so taht will tie into the break-resource-locks blueprint | 16:53 |
| *** harlowja has joined #openstack-ironic | 16:54 | |
| NobodyCam | hehe :) | 16:54 |
| lucasagomes | devananda, yea but should it be exposed by default? as u said it's good for developers to troubleshoot things | 16:54 |
| lucasagomes | and as a dev I'm happy to go to the db and check it | 16:54 |
| lucasagomes | the normal user shouldn't really care IMO | 16:54 |
| lucasagomes | NobodyCam, ^ thoughts? | 16:57 |
| NobodyCam | lucasagomes: would it be worth showing "reserved" if --debug is included? so admins could get that info easily? | 16:57 |
| devananda | lucasagomes: "normal user" == cloud admin | 16:57 |
| *** albionandrew has joined #openstack-ironic | 16:58 | |
| devananda | IMO, --debug shouldn't be used to get information about the resource. It should be used to debug the call to the API | 16:58 |
| devananda | ironic node-show --all-fields | 16:58 |
| lucasagomes | devananda, the reserved will have the hostname of the conductor? or it's just a flag like 1 or 0? | 16:58 |
| *** hemnafk is now known as hemna | 16:58 | |
| devananda | would be better IF we need to hide some fields by default, but I dont think we do | 16:59 |
| devananda | lucasagomes: hsotname | 16:59 |
| lucasagomes | I see | 16:59 |
| lucasagomes | hmm | 16:59 |
| *** derekh has quit IRC | 16:59 | |
| NobodyCam | we dont have a --verbose do we | 17:00 |
| NobodyCam | devananda: I do tent to aggree with you. our target audince is admins, | 17:01 |
| NobodyCam | *tend | 17:02 |
| lucasagomes | NobodyCam, good point I think it's on the CLI but doesnt seems to work | 17:02 |
| lucasagomes | I mean it returns the same with --verbose or without it hehe | 17:02 |
| * lucasagomes think that we need a bug for it | 17:02 | |
| lucasagomes | we also have the --debug which dumps the json request and http codes etc | 17:03 |
| NobodyCam | lucasagomes: there are commnets on the man page review to clearify the --debug | 17:04 |
| lucasagomes | I see, yea I will try to submit a new patch for that tomorrow (came to the office today, and it's already 18:04 here) | 17:04 |
| devananda | --debug dumping http and json is good | 17:05 |
| NobodyCam | lol enjoy the drive home | 17:05 |
| lucasagomes | yea --debug is pretty useful | 17:05 |
| lucasagomes | now idk what we should do with the --verbose | 17:05 |
| NobodyCam | :) | 17:06 |
| devananda | mordred: any opinion on unittest2 for py26 compat? | 17:07 |
| clarkb | devananda: please don't :) | 17:08 |
| clarkb | devananda: testtools should cover many things that unittest2 offers and keeps the number of dependencies down | 17:08 |
| clarkb | devananda: is there a particular feature you need? | 17:08 |
| devananda | clarkb: assertItemsEqual | 17:08 |
| devananda | also, unittest2 is still in global-requirements.txt | 17:09 |
| clarkb | devananda: yes some projects still use it, but we are slowly killing it | 17:09 |
| devananda | k | 17:09 |
| devananda | clarkb: is there an equivalent in testtools? | 17:09 |
| devananda | ah. looks like nova has collected a little library of matchers | 17:10 |
| clarkb | http://testtools.readthedocs.org/en/latest/for-test-authors.html#contains may need to be combined with another matcher to get the exact behavior | 17:11 |
| devananda | thanks. i'll poke at it | 17:12 |
| NobodyCam | quick walkies...brb | 17:13 |
| max_lobur | Hi devananda | 17:14 |
| lucasagomes | devananda, NobodyCam I'm done for today, have a g'night | 17:14 |
| max_lobur | Sorry, I didn't think of 2.6 support when proposed assertItemsEqual | 17:15 |
| *** lucasagomes has quit IRC | 17:16 | |
| devananda | max_lobur: isn't it fun trying to support 2.6, 2.7, and 3.3? :) | 17:16 |
| NobodyCam | devananda: :-p | 17:16 |
| max_lobur | to avoid unittest2 we could add our implementation of assertItemsEqual to tests/base.py TestCase class | 17:16 |
| max_lobur | like asserEquals(sorted(seq_1), sorted(seq_2)) | 17:17 |
| devananda | max_lobur: or use testtools | 17:17 |
| devananda | your idea is certainly simpler. but over time we'll probably want to inherit some of the other matches from nova | 17:18 |
| max_lobur | IMO contains will be less obvious | 17:18 |
| devananda | eg, DictEqual, DictContains, DictHasSameKeys, etc | 17:18 |
| max_lobur | devananda, true | 17:18 |
| devananda | see noav/tests/matchers.py | 17:18 |
| *** jianingy_afk has quit IRC | 17:19 | |
| max_lobur | for 2.6 and other versions support - yea=) just need to have a big poster above the monitor - DON'T FORGET ABOUT 2.6 AND 3.3 =) | 17:20 |
| *** hemna has quit IRC | 17:22 | |
| *** rloo has quit IRC | 17:27 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Conductors maintan driver list in the DB https://review.openstack.org/51743 | 17:32 |
| openstackgerrit | A change was merged to openstack/ironic: Fixes RootController to allow URL without version tag https://review.openstack.org/52516 | 17:36 |
| *** martyntaylor has joined #openstack-ironic | 17:38 | |
| *** martyntaylor has left #openstack-ironic | 17:38 | |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Regenerate the sample config file. https://review.openstack.org/53702 | 17:41 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Group and unify ipmi configurations https://review.openstack.org/52615 | 17:41 |
| * mordred votes for testtools matches | 17:52 | |
| mordred | matchers | 17:52 |
| mordred | we should also consider upstreaming the matchers in nova | 17:52 |
| *** datajerk has quit IRC | 17:52 | |
| openstackgerrit | linggao proposed a change to openstack/python-ironicclient: Modifies CLI to show nodes by instance uuid https://review.openstack.org/53485 | 17:55 |
| devananda | NobodyCam: I'm +A'ing some things to get the queue down | 18:02 |
| NobodyCam | linggao: ^^^ can you check i18n on ^^^ | 18:03 |
| NobodyCam | devananda: :) | 18:03 |
| linggao | NobodyCam, ? | 18:04 |
| *** derekh has joined #openstack-ironic | 18:04 | |
| NobodyCam | https://review.openstack.org/#/c/53485/2/ironicclient/v1/node_shell.py around lines 42-48 | 18:04 |
| NobodyCam | _("I18n text") | 18:05 |
| openstackgerrit | A change was merged to openstack/ironic: Updated from global requirements https://review.openstack.org/53371 | 18:05 |
| NobodyCam | it for translation | 18:05 |
| NobodyCam | it's* | 18:05 |
| openstackgerrit | A change was merged to openstack/ironic: Support uniform MAC address with colons https://review.openstack.org/52761 | 18:06 |
| *** datajerk has joined #openstack-ironic | 18:07 | |
| linggao | NobodyCam, what do you mean by "check i18n on"? :) | 18:07 |
| devananda | linggao: i18n == internationalization | 18:08 |
| devananda | _ is syntactic magic for translating strings to locale-specific langauges, when present | 18:08 |
| NobodyCam | lol there are several i18n updates needed i node_shell | 18:08 |
| NobodyCam | but look at line 153 | 18:09 |
| NobodyCam | that raise is i18n | 18:09 |
| * NobodyCam wouldn't -1 if linggao fixed some of the others too. | 18:10 | |
| linggao | :) | 18:10 |
| NobodyCam | like #143, 119 , ect | 18:10 |
| linggao | NobodyCam, do you mean change | 18:11 |
| NobodyCam | devananda: in the client do we need to i18n the prints... ie line 120 in node_shell.py | 18:11 |
| linggao | raise exc.CommandError('Node not found: %s' % n) | 18:11 |
| linggao | to | 18:11 |
| linggao | raise exc.CommandError(_('Node not found: %s') % n) | 18:12 |
| linggao | ? | 18:12 |
| NobodyCam | linggao: :) yeppers | 18:12 |
| linggao | ah, that's something new for me. thanks for the help. | 18:13 |
| NobodyCam | :) | 18:13 |
| devananda | Haomeng: i've commented on the sort() bug and review | 18:13 |
| NobodyCam | linggao: there is a open bug for fixing all ironic's i18n issues | 18:14 |
| linggao | NobodyCam, I'll take a look at it. | 18:15 |
| NobodyCam | linggao: fixing the others I pointed out IS out side the scope of your patch... | 18:16 |
| *** romcheg has joined #openstack-ironic | 18:16 | |
| linggao | yes :) phew! | 18:16 |
| NobodyCam | :) | 18:17 |
| NobodyCam | jsut saying I wouldn't -1 it fix they happen to get fixed :) | 18:17 |
| NobodyCam | hehehe | 18:17 |
| NobodyCam | s/it/if/ | 18:17 |
| linggao | lol | 18:17 |
| NobodyCam | mordred: i18n question... if you have a second? | 18:19 |
| NobodyCam | gah who put that diff all side by side button there | 18:20 |
| NobodyCam | :-p | 18:20 |
| *** derekh has quit IRC | 18:22 | |
| *** anniec has joined #openstack-ironic | 18:22 | |
| devananda | GheRivero: ping | 18:22 |
| *** romcheg has left #openstack-ironic | 18:23 | |
| NobodyCam | lol seems many of the other clients have nit done i18n updates yet | 18:30 |
| linggao | So NobodyCam, for i18n, every string should be use gettext alas _(), right? | 18:31 |
| linggao | like | 18:31 |
| NobodyCam | not every | 18:31 |
| NobodyCam | def the raise's | 18:31 |
| linggao | @utils.arg('node', metavar='<id>', | 18:31 |
| linggao | help=_("ID, UUID or instance UUID of node")) | 18:31 |
| mordred | NobodyCam: on the phone | 18:31 |
| linggao | how about this? | 18:32 |
| linggao | @utils.arg('node', metavar='<id>', | 18:32 |
| linggao | help="ID, UUID or instance UUID of node") | 18:32 |
| linggao | wrong paste | 18:32 |
| NobodyCam | mordred: :) all good.. just have a quick i18n question when your free | 18:32 |
| linggao | fields = ['uuid', 'instance_uuid', 'power_state', 'target_power_state', | 18:32 |
| linggao | 'provision_state', 'target_provision_state', 'driver', | 18:32 |
| linggao | 'driver_info', 'properties', 'extra', | 18:32 |
| linggao | 'created_at', 'updated_at', 'reservation'] | 18:32 |
| *** rloo has joined #openstack-ironic | 18:33 | |
| linggao | Should each filed be gettext-ed? | 18:33 |
| NobodyCam | linggao: nope those are field names translating them would break things | 18:33 |
| linggao | ah, you are right. | 18:34 |
| NobodyCam | linggao: think of it as text that gets displayed to uesrs | 18:34 |
| NobodyCam | mainly I see it raises | 18:35 |
| NobodyCam | it in* | 18:35 |
| NobodyCam | my question for mordred is things like https://github.com/openstack/python-ironicclient/blob/master/ironicclient/v1/node_shell.py#L100 | 18:35 |
| NobodyCam | that one I'm not sure about | 18:35 |
| linggao | https://github.com/openstack/python-ironicclient/blob/master/ironicclient/v1/node_shell.py#L54 | 18:36 |
| linggao | that should, right? | 18:36 |
| NobodyCam | linggao: good question... lets wait for mordred he may point us in the right direction. I have not seen other projects do that, | 18:38 |
| linggao | ok | 18:38 |
| NobodyCam | but I can see values in it | 18:38 |
| NobodyCam | value | 18:38 |
| NobodyCam | :-p | 18:38 |
| linggao | yes, we better do it as we changing the file. | 18:39 |
| *** rloo has quit IRC | 18:46 | |
| *** rloo has joined #openstack-ironic | 18:46 | |
| devananda | NobodyCam: long discussion here you may be interested in: https://bugs.launchpad.net/ironic/+bug/1237688 | 18:47 |
| devananda | lifeless: also you may be interested in, too ^ | 18:48 |
| NobodyCam | devananda: rloo and I chatted about that yesterday | 18:49 |
| devananda | :) | 18:50 |
| lifeless | devananda: btw assertItemsEqual is a matcher in testtools. | 18:51 |
| NobodyCam | I like rloo Idea #1 or #2 and #3 | 18:51 |
| NobodyCam | reading now | 18:51 |
| devananda | lifeless: ack | 18:51 |
| lifeless | devananda: there isn't an exact match, but you can curry one up quickly or point me at the actual tests that are ugly and I'll see what makes sense to me | 18:56 |
| lifeless | devananda: bug commented on | 18:56 |
| NobodyCam | morning lifeless :) | 18:57 |
| devananda | lifeless: assertItemsEqual came up in teh context of unsorted lists. in this case, I think it's sufficient to just continue to use assertEqual(sorted(a), sorted(b)) | 18:59 |
| devananda | we sidetracked into the "why is assertItemsEqual failing in py26" area a bit .... :) | 18:59 |
| *** jianingy_afk has joined #openstack-ironic | 19:03 | |
| clarkb | devananda: I wonder why you are using a ordered type if order doesn't matter? | 19:04 |
| clarkb | it doesn't really hurt except for in cases like this where you are checking equality | 19:04 |
| *** anniec has quit IRC | 19:06 | |
| NobodyCam | clarkb: are you familiar with the i18n stuff? | 19:07 |
| clarkb | NobodyCam: sort of | 19:07 |
| clarkb | NobodyCam: I am familiar with the automation around it but less so code internals | 19:07 |
| NobodyCam | https://github.com/openstack/python-ironicclient/blob/master/ironicclient/v1/node_shell.py#L100 | 19:08 |
| NobodyCam | should something like that be print _("balh") | 19:08 |
| *** anniec has joined #openstack-ironic | 19:12 | |
| devananda | lifeless: seen this before? http://www.opencompute.org/assets/download/OpenComputeProjectOpenMachinesManagement.pdf | 19:14 |
| NobodyCam | brb | 19:14 |
| NobodyCam | back | 19:16 |
| *** WillOchanda has quit IRC | 19:17 | |
| devananda | i thought i saw an OCP-based ramdisk for hw mgmt, but can't find the link any more | 19:17 |
| devananda | it's mentioned as a goal in taht 2-yr-old paper | 19:17 |
| linggao | NobodyCam, for i18n, I got "global name '_' is not defined" error for exception | 19:18 |
| lifeless | devananda: I hadn't actually opened it before :) | 19:18 |
| linggao | something mut be missing. | 19:18 |
| *** WillOchanda has joined #openstack-ironic | 19:19 | |
| NobodyCam | linggao: from ironic.openstack.common.gettextutils import _ | 19:22 |
| NobodyCam | where are you geting that error? | 19:22 |
| linggao | https://github.com/openstack/python-ironicclient/blob/master/ironicclient/v1/node_shell.py#L38 | 19:24 |
| linggao | If I add _() to the string and run "ironic node-show 5" where 5 is an invalide number | 19:24 |
| linggao | without _(), the command returns "Node not found: ..." | 19:25 |
| NobodyCam | humm | 19:28 |
| NobodyCam | one sec | 19:28 |
| linggao | Adding the following works: | 19:28 |
| linggao | from gettext import gettext as _ | 19:29 |
| NobodyCam | look like something with the loacalization: | 19:32 |
| NobodyCam | https://github.com/openstack/python-ironicclient/blob/master/ironicclient/openstack/common/gettextutils.py#L26-L34 | 19:32 |
| NobodyCam | not a whole lot going on there | 19:33 |
| linggao | I did "from ironicclient.openstack.common import gettextutils" in the code, still not working | 19:35 |
| clarkb | linggao: you wuold need to do gettextutils._ with that import I think | 19:35 |
| NobodyCam | lol guess we should add a locale dir and ironicclient.pot file to the client huh | 19:36 |
| NobodyCam | clarkb: I think the client env is not correct | 19:36 |
| NobodyCam | no locale or pot file | 19:37 |
| clarkb | NobodyCam: it probably isn't. Getting those bits correct has been problematic in general | 19:37 |
| NobodyCam | linggao: one sec going to do a test here | 19:37 |
| WillOchanda | NobodyCam & rloo: thanks for your help yesterday, we got our driver up and running! | 19:38 |
| linggao | clarkb, from ironicclient.openstack.common.gettextutils import _ works | 19:38 |
| NobodyCam | WillOchanda: AWESOME!!!! | 19:38 |
| rloo | WillOchanda: yay! | 19:38 |
| NobodyCam | your going to submit it ????!!!????? | 19:38 |
| WillOchanda | not yet, its very hacky now while we finalize APIs on our system, we'll clean it up and submit it in decemberish | 19:38 |
| WillOchanda | right now its just good for some show and tell in hong kong | 19:39 |
| NobodyCam | oh WillOchanda : keep an eye on https://review.openstack.org/#/c/53356/ | 19:39 |
| NobodyCam | :) awesome look forward to seeing it in action there | 19:40 |
| WillOchanda | NobodyCam: that'll be useful, looking forward | 19:40 |
| WillOchanda | on to the next challenge - playing with /vendor_passthru/ :) | 19:40 |
| WillOchanda | are there any examples of this being implemented? | 19:40 |
| NobodyCam | WillOchanda: that may not be as well documented (yet) as was the power state stuff :-/ | 19:41 |
| WillOchanda | that may be ok.. my understanding is if i call /vendor_passthrough/myfunction it'll map to myfunction in the driver, right? | 19:42 |
| neo_ | hi NobodyCam, how are you? | 19:42 |
| NobodyCam | hey neo_ ... doing good. | 19:43 |
| devananda | WillOchanda: that's the idea, yes | 19:43 |
| NobodyCam | that image deploy? | 19:43 |
| devananda | WillOchanda: there should be an example in the PXE driver | 19:43 |
| neo_ | cool. just a quick upate... i tried with the partitioned image but i see the same issue. am stuck with this issue | 19:43 |
| WillOchanda | great, let me take a look | 19:43 |
| NobodyCam | neo_: still the blkid error | 19:43 |
| devananda | NobodyCam: I'm updating your proposal for communicating with the nodes & merging with one of romcheg's | 19:43 |
| neo_ | yes | 19:43 |
| NobodyCam | devananda: :) thank you | 19:44 |
| NobodyCam | neo_: ok lets take another look at the big picture. | 19:44 |
| NobodyCam | you CAN deploy the ubuntu cloud image | 19:45 |
| neo_ | yes | 19:45 |
| neo_ | i can. its working just fine. i get the ubuntu cloud image from disk-image-builder | 19:45 |
| neo_ | and that goes through fine. | 19:45 |
| NobodyCam | but you are NOT able to deploy your image | 19:45 |
| NobodyCam | ok... | 19:45 |
| neo_ | yes. correct. not able to deploy my image. i changed my image to be the "same" as the ubuntu cloud image. | 19:46 |
| neo_ | i got the ubuntu cloud image, qcow2, changed to vmdk, brought up a vm, then added some packages in the vm, then got the vmdk and converted back to qcow2 and this qcow2 has the same issue. | 19:46 |
| NobodyCam | yes | 19:46 |
| NobodyCam | you can not convert vmdk directly to qcow2 | 19:47 |
| neo_ | i converted to raw in between ;) | 19:47 |
| neo_ | yes i did | 19:47 |
| NobodyCam | I have had good luck going to raw | 19:47 |
| NobodyCam | ok and same issue | 19:48 |
| neo_ | yes | 19:48 |
| NobodyCam | and CB is working on a DIB element for your project? | 19:48 |
| devananda | NobodyCam: I've copied your proposal and my edited summary to https://etherpad.openstack.org/p/IcehouseIronicNodeCommunication | 19:49 |
| neo_ | i am sorry, what is that? | 19:49 |
| NobodyCam | diskimage-builder is part of the TripleO (OpenstackOnOpenstack) story | 19:50 |
| *** albionandrew has quit IRC | 19:50 | |
| NobodyCam | it is the tool I use to create my images | 19:50 |
| NobodyCam | https://github.com/openstack/diskimage-builder | 19:51 |
| NobodyCam | when I last spoke with CBox he was going to work in a element for you | 19:51 |
| neo_ | NobodyCam: yes you are right. i used DIB, yes i used it and could successfully deploy ubuntu cloud image on the baremetal server hardware. | 19:51 |
| NobodyCam | (I think) :-p | 19:51 |
| linggao | NobodyCam, I think what's mising now for i18n | 19:52 |
| linggao | We need to add the following to ironicclient/__init_.py | 19:52 |
| linggao | from ironicclient.openstack.common import gettextutils | 19:52 |
| linggao | gettextutils.install('ironicclient') | 19:52 |
| NobodyCam | ok.. if I understand what you are doing. quck recap: install you custom app and deploy.. | 19:53 |
| linggao | then we can use _('blah") everywhere in the ironicclient code. | 19:53 |
| linggao | no more import in each file. | 19:53 |
| NobodyCam | linggao: yes that sounds correct | 19:53 |
| rloo | devananda: do you have a few minutes to discuss https://bugs.launchpad.net/ironic/+bug/1237688 | 19:54 |
| *** albionandrew has joined #openstack-ironic | 19:54 | |
| devananda | rloo: yes | 19:54 |
| devananda | and then i need to get lunch :) | 19:54 |
| NobodyCam | neo_: see pm | 19:54 |
| linggao | NobodyCam, should I include it in my current patch or in a seperate patch? | 19:54 |
| rloo | devananda: one thing I may be missing. When someone does a change-power-request and it fails, won't some error be returned to them? | 19:55 |
| NobodyCam | should really be a seperate patch and then your current should depend on it | 19:55 |
| devananda | rloo: no | 19:57 |
| devananda | rloo: see conductor/rpcapi change_node_power_state | 19:57 |
| neo_ | i am sorry i dint get you NobodyCam, | 19:57 |
| devananda | rloo: it is a "cast" -- which means there is no return possible from the manager | 19:57 |
| WillOchanda | is the 'disk' property implemented for node objects yet? ie - should I be able to set it? | 19:58 |
| linggao | NobodyCam, sure. | 19:58 |
| NobodyCam | neo_: I sent you a private message | 19:58 |
| devananda | WillOchanda: what "disk" property? if you're reading the API spec, it's probably misleading you | 19:58 |
| rloo | devananda: ohhh. that explains things. Hmm. Now i have to think about it again. thx. | 19:58 |
| devananda | rloo: np. yea, that's why all this is an issue | 19:59 |
| WillOchanda | devananda: yup reading the API spec, each node has fields for disk, ram, arch, etc | 19:59 |
| *** albionandrew has quit IRC | 19:59 | |
| devananda | WillOchanda: ah. so, all that has been subsumed by the 'properties' nested resource | 19:59 |
| devananda | APIdoc is out of date | 19:59 |
| WillOchanda | devananda: i suspected that, is 'properties' a free-form dictionary now, or is there a prescribed list of keys i can use? | 20:00 |
| devananda | WillOchanda: so you can set what ever you want in there -- it's a list of key: value pairs | 20:00 |
| devananda | the goal, which i haven't communicated clearly i nthe docs, is for properties to represent what is exposed to nova | 20:00 |
| rloo | devananda: in that case, I think having the 'error' be avail is the easiest soln (well, if we assume that the user wants some way of getting it outside of mining the logs) | 20:00 |
| devananda | eg, # cpus, cpu arch, ram, and any other extra specs that the scheduler needs to know | 20:00 |
| devananda | rloo: i think that's a fair assumption :) | 20:01 |
| *** albionandrew has joined #openstack-ironic | 20:01 | |
| devananda | rloo: users -- even cloud admins -- may not have physical access to the log files | 20:01 |
| devananda | WillOchanda: whereas driver-required info, such as "how big a root volume to create" should be set on the driver_info dict | 20:01 |
| devananda | WillOchanda: see drivers.modules/pxe.py: 94 - 100 | 20:02 |
| devananda | imaeg_source, root_gb, deploy_kernel, etc | 20:02 |
| devananda | these will come from Nova and be used by the PXE driver when deploying the image | 20:02 |
| WillOchanda | ah ok, got it | 20:03 |
| WillOchanda | now for how to poke it, is it something like this: | 20:03 |
| WillOchanda | curl -X PATCH -d "disk=20000" http://localhost:6385/v1/nodes/db811f96-980f-4acb-ba7c-95a391413d9b/properties | 20:03 |
| WillOchanda | (because that doesn't work :/ | 20:03 |
| devananda | WillOchanda: you acn use the CLI to do it | 20:03 |
| devananda | PATCH syntax is a bit more complex | 20:04 |
| devananda | if you want to write your own client, you need to follow the RFC spec .... /me googles for it | 20:05 |
| WillOchanda | alright, thx, i'll keep playing | 20:05 |
| devananda | eg, see "ironic node-update <uuid> replace driver_info/root_gb=100 | 20:07 |
| devananda | s/see/try/ | 20:07 |
| openstackgerrit | linggao proposed a change to openstack/python-ironicclient: Modifies CLI to show nodes by instance uuid https://review.openstack.org/53485 | 20:07 |
| WillOchanda | great, got it working | 20:08 |
| WillOchanda | and if i want to modify this property from within the driver can i do node['properties'] = {"disk" : 20000} ? | 20:09 |
| devananda | that would wipe out the current properties | 20:10 |
| devananda | i think you want to do this: node['driver_info']['disk'] = 20000 | 20:11 |
| WillOchanda | ah, right, so node['properties']['disk'] = ok yup | 20:11 |
| devananda | yea | 20:11 |
| WillOchanda | thanks for fielding newbie questions, i'm not quite a python pro | 20:11 |
| WillOchanda | (obviously) | 20:12 |
| devananda | np | 20:12 |
| devananda | bbiab | 20:14 |
| *** lnxnut has joined #openstack-ironic | 20:20 | |
| openstackgerrit | linggao proposed a change to openstack/python-ironicclient: Enable gettext alias for all ironicclient modules https://review.openstack.org/53730 | 20:20 |
| *** lnxnut has quit IRC | 20:22 | |
| *** lnxnut has joined #openstack-ironic | 20:23 | |
| linggao | anyone, how do I add dependency for a patch that is under review? It had no depencies before. | 20:25 |
| *** lnxnut has quit IRC | 20:27 | |
| NobodyCam | linggao: git review -d ##### | 20:30 |
| NobodyCam | git checkout YourPatchesBranch | 20:30 |
| NobodyCam | git rebase | 20:30 |
| NobodyCam | git review | 20:30 |
| linggao | NobodyCam, let me try... | 20:31 |
| NobodyCam | answer yes to git question | 20:32 |
| *** jdob has quit IRC | 20:33 | |
| openstackgerrit | linggao proposed a change to openstack/python-ironicclient: Modifies CLI to show nodes by instance uuid https://review.openstack.org/53485 | 20:33 |
| NobodyCam | :) | 20:33 |
| linggao | yeah! | 20:35 |
| linggao | thanks | 20:35 |
| NobodyCam | :) great job linggao :) | 20:36 |
| linggao | thanks a lot NobodyCam for your patience and help. | 20:37 |
| WillOchanda | devananda: did the path for vendor_passthru change from this - http://localhost:6385/v1/nodes/add06ead-2de2-4012-8b79-60575c17d918/vendor_passthru ? | 20:38 |
| WillOchanda | oops, guess he stepped out, NobodyCam do you know? | 20:39 |
| NobodyCam | WillOchanda: he's getting a bite to eat | 20:43 |
| NobodyCam | looks right | 20:44 |
| NobodyCam | humm do_vendor_action | 20:45 |
| NobodyCam | running out for quick walk | 20:46 |
| WillOchanda | can't make that work either, alright | 20:46 |
| NobodyCam | brb about 10 / 20 min | 20:46 |
| *** linggao has quit IRC | 20:57 | |
| *** anniec has quit IRC | 20:58 | |
| NobodyCam | back | 20:59 |
| *** anniec has joined #openstack-ironic | 21:04 | |
| *** CerealBox has joined #openstack-ironic | 21:06 | |
| *** jbjohnso has quit IRC | 21:10 | |
| *** WillOchanda has quit IRC | 21:21 | |
| devananda | back | 21:41 |
| NobodyCam | wb | 21:43 |
| *** WillOchanda has joined #openstack-ironic | 21:47 | |
| WillOchanda | devananda: back from lunch? | 21:53 |
| devananda | ya | 21:53 |
| WillOchanda | any ideas on /vendor_passthru/? | 21:54 |
| WillOchanda | is it implemented? | 21:54 |
| *** anniec has quit IRC | 22:04 | |
| *** anniec has joined #openstack-ironic | 22:22 | |
| *** holoirc has joined #openstack-ironic | 22:29 | |
| * NobodyCam start a new build and goes for quick walies | 22:35 | |
| NobodyCam | *starts | 22:35 |
| devananda | WillOchanda: it should be | 22:41 |
| devananda | WillOchanda: did you look at drivers/modules/pxe.py: 582 - 591? | 22:41 |
| devananda | WillOchanda: ooh. it may not be in the API yet | 22:41 |
| devananda | WillOchanda: haha! yea .... http://git.openstack.org/cgit/openstack/ironic/tree/ironic/api/controllers/v1/node.py#n296 | 22:43 |
| devananda | so all the bits are on the back end (RPC, manager, driver) but the API isn't plugged in yet | 22:43 |
| WillOchanda | ahh, bummer! | 22:46 |
| devananda | WillOchanda: i'm implementing it now | 22:47 |
| devananda | small oversight :) | 22:47 |
| WillOchanda | :), is it a big effort? | 22:47 |
| devananda | it's done already (i think). just writing unit tests now | 22:48 |
| devananda | was ~4 lines of code | 22:48 |
| WillOchanda | sweet | 22:49 |
| *** liteIRC has joined #openstack-ironic | 22:49 | |
| *** pky has joined #openstack-ironic | 22:51 | |
| *** michchap has joined #openstack-ironic | 22:56 | |
| * NobodyCam brb | 23:00 | |
| *** anniec has quit IRC | 23:00 | |
| *** anniec has joined #openstack-ironic | 23:03 | |
| devananda | as usual, unit tests are trickier than the actual code :) | 23:04 |
| WillOchanda | :) | 23:05 |
| *** holoirc has quit IRC | 23:05 | |
| *** liteIRC has quit IRC | 23:06 | |
| WillOchanda | i'm also trying to get this PATCH stuff right - so i'm doing PATCH to /v1/UUID content-type=application/json, and my body is a json-encoded list of lines that look like this | 23:06 |
| WillOchanda | {"path": "/driver_info/address", "value": "10.216.142.87", "op": "add"} | 23:06 |
| WillOchanda | anything obviously wrong with that? | 23:07 |
| devananda | hrm. this is why we really need up-to-date API docs :) | 23:10 |
| devananda | i am not familiar with the PATCH syntax to off-hand spot anything wrong there | 23:10 |
| *** albionandrew has quit IRC | 23:10 | |
| devananda | and would point you to lucasgomes.... | 23:10 |
| devananda | but he's probably asleep | 23:11 |
| devananda | using our python library is not an option, eh? | 23:11 |
| WillOchanda | i'm open to that, didn't know one existed | 23:12 |
| WillOchanda | is it part of the main codebase? | 23:12 |
| devananda | openstack/python-ironicclient | 23:12 |
| devananda | has a python API lib and a CLI | 23:12 |
| NobodyCam | neo_: get all that going? | 23:17 |
| devananda | hrm, why is _default not working? | 23:18 |
| *** michchap has quit IRC | 23:19 | |
| *** michchap has joined #openstack-ironic | 23:22 | |
| *** anniec_ has joined #openstack-ironic | 23:24 | |
| *** anniec has quit IRC | 23:25 | |
| *** anniec_ is now known as anniec | 23:25 | |
| Haomeng | devananda: ping | 23:43 |
| devananda | Haomeng: pong | 23:43 |
| Haomeng | good afternoon:) | 23:43 |
| Haomeng | I will add the common created_at updated_at fields into our base class APIBase, is ok? | 23:44 |
| devananda | thanks! | 23:44 |
| Haomeng | then all our API will have these properties, and I found our IronicBase Inherit oslo common class for these fields TimestampMixin, so our Ironic is same with other projects | 23:48 |
| devananda | yep | 23:50 |
| Haomeng | devananda: one more question, how do you think if we support client to display time in local TZ format? | 23:50 |
| Haomeng | by default, we store/display in UTC | 23:50 |
| Haomeng | I think this isnot required | 23:52 |
| NobodyCam | linggao here | 23:52 |
| NobodyCam | ? | 23:52 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Implement /nodes/UUID/vendor_passthru in the API https://review.openstack.org/53761 | 23:53 |
| devananda | WillOchanda: ^^^ | 23:53 |
| devananda | woops, pep8 .... | 23:53 |
| openstackgerrit | Devananda van der Veen proposed a change to openstack/ironic: Implement /nodes/UUID/vendor_passthru in the API https://review.openstack.org/53761 | 23:53 |
| NobodyCam | 53730 is going to conflict with shell.py in 53356 | 23:54 |
| devananda | Haomeng: UTC is required | 23:54 |
| devananda | Haomeng: all times should be stored in the database in UTC | 23:54 |
| Haomeng | yes, it is our current code default behavior | 23:54 |
| Haomeng | I want toknow show we support client to display time in local TZ format? | 23:55 |
| devananda | Haomeng: as far as the client, I think silently converting to local TZ in the CLI is ... problematic in some cases. | 23:55 |
| devananda | imagine if you are the admin | 23:55 |
| Haomeng | :) | 23:55 |
| devananda | you SELECT some records from the DB, then you "ironic node-show" the same records | 23:55 |
| devananda | and the times are totally different! | 23:55 |
| Haomeng | yes | 23:55 |
| WillOchanda | devananda: thanks! trying it now | 23:56 |
| Haomeng | ok, we keep current behavior | 23:56 |
| Haomeng | will not change client code | 23:56 |
| NobodyCam | devananda: was that ever -D'd | 23:56 |
| devananda | ? | 23:57 |
| NobodyCam | I don't see a "Uploaded patch set 1." comemnt | 23:57 |
| devananda | there was a p1 | 23:58 |
| devananda | i just uploaded p2 really really quic | 23:58 |
| devananda | no -D | 23:58 |
| NobodyCam | oh well | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!