| *** zz_pwnall1337 is now known as pwnall1337 | 00:07 | |
| openstackgerrit | Merged openstack/python-cratonclient master: Adds basic Cloud resource CRUD to CLI https://review.openstack.org/437098 | 00:21 |
|---|---|---|
| openstackgerrit | Merged openstack/craton master: Require region_id and cloud_id for POST /networks https://review.openstack.org/438265 | 00:24 |
| *** VW_ has joined #craton | 03:30 | |
| *** VW has quit IRC | 03:30 | |
| *** pwnall1337 is now known as zz_pwnall1337 | 03:31 | |
| *** VW_ has quit IRC | 03:31 | |
| *** VW has joined #craton | 03:31 | |
| *** VW has quit IRC | 03:57 | |
| *** VW has joined #craton | 04:16 | |
| *** valw has joined #craton | 05:12 | |
| *** valw has quit IRC | 05:17 | |
| *** tojuvone has joined #craton | 05:54 | |
| openstackgerrit | Jim Baker proposed openstack/python-cratonclient master: WIP: CLI and client support for get/set/delete of resource vars https://review.openstack.org/427032 | 05:56 |
| openstackgerrit | Jim Baker proposed openstack/python-cratonclient master: WIP: CLI and client support for get/set/delete of resource vars https://review.openstack.org/438339 | 06:13 |
| openstackgerrit | Jim Baker proposed openstack/python-cratonclient master: WIP: CLI and client support for get/set/delete of resource vars https://review.openstack.org/427032 | 06:18 |
| *** VW has quit IRC | 06:34 | |
| *** valw has joined #craton | 07:13 | |
| *** valw has quit IRC | 07:17 | |
| *** tojuvone has quit IRC | 08:11 | |
| *** tojuvone has joined #craton | 08:11 | |
| *** tojuvone has quit IRC | 09:44 | |
| *** tojuvone has joined #craton | 11:05 | |
| sulo | https://bugs.launchpad.net/craton/+bug/1668251 | 12:27 |
| openstack | Launchpad bug 1668251 in craton "Including project in var resolution order causes error" [Critical,New] | 12:27 |
| thomasem | sulo: are you getting that bug on master? | 13:04 |
| sulo | thomasem: yeah | 13:04 |
| sulo | also our cli (client) sucks | 13:05 |
| thomasem | Could be a set up problem... if the first project doesn't get a variable_association entry, it'll have that aerror. | 13:05 |
| thomasem | error* | 13:05 |
| thomasem | Which is why I added https://github.com/openstack/craton/blob/master/tools/docker_run.sh#L41 | 13:05 |
| thomasem | sulo: was this from a manual set up? | 13:05 |
| thomasem | or docker_run.sh set up? | 13:05 |
| sulo | thomasem: manual | 13:05 |
| thomasem | Yeah, you MUST add a variable_association for the project | 13:06 |
| sulo | huh ? | 13:06 |
| sulo | i didnt update .. i did a fresh install | 13:06 |
| thomasem | In the setup where you're seeing that error, do you have a row in the variable_association for that project, and the ID of that row referenced in the Project's record? | 13:06 |
| sulo | i dont | 13:07 |
| thomasem | That's what causes that error, I believe. | 13:08 |
| thomasem | In the bug you listed | 13:08 |
| sulo | hrm .. i am pretty sure i nuked everything and started fresh | 13:08 |
| thomasem | Okay, and what process did you go through to start it? | 13:08 |
| sulo | thomasem: create db / run db sync / start api | 13:09 |
| thomasem | Okay, and starting the API throw that error? | 13:09 |
| sulo | no | 13:09 |
| thomasem | At what point did you get that error? | 13:09 |
| sulo | its when you do vars for say hosts | 13:09 |
| thomasem | Did you manually add a project? | 13:09 |
| sulo | resolved-vars | 13:09 |
| sulo | thomasem: yes | 13:10 |
| thomasem | Without giving it a variable_association reference? | 13:10 |
| thomasem | Part of the bootstrap process now must be adding a variable_association reference for the first project, too. | 13:10 |
| thomasem | Because that's required to reference the variables for that object. No variable_association, no variables (it returned None, I think) and therefore no __dict__ attribute. | 13:11 |
| thomasem | What I'm positing is that's not a bug, but expected behavior given the implemented design. | 13:12 |
| sulo | so we are expecting users to do project create and go into the db and add something to make it work ? | 13:12 |
| thomasem | I don't know... that's been the premise from the start here. I think it makes more sense to start with some bootstrapped project that users can then delete once they make their own root user. | 13:13 |
| thomasem | I don't like this bootstrap process as it is at all. | 13:13 |
| sulo | thats what we have ? | 13:13 |
| sulo | right i see what you are syaing | 13:13 |
| thomasem | You said you bootstrapped manually | 13:13 |
| sulo | yeah ... but i am expecting that user to be temp | 13:14 |
| sulo | once i start creating projects i dont want projects to go in and do anything on the db | 13:14 |
| thomasem | I think that's perfectly fine. My point is that if you're going to manually bootstrap, you must follow the same process as docker_run.sh right now. | 13:14 |
| *** valw has joined #craton | 13:14 | |
| thomasem | I think the better approach would be some default set up via the migration. | 13:15 |
| sulo | yeah i agree | 13:15 |
| thomasem | So, let's do that in that bug you linked? | 13:16 |
| sulo | sounds good | 13:16 |
| thomasem | Cool. I was wanting to explain why that bug was happening all of a sudden, if you didn't know, is all. | 13:18 |
| *** valw has quit IRC | 13:19 | |
| thomasem | sulo: why do you say the CLI sucks, btw? | 13:26 |
| sulo | thomasem: heh, mostly because its missing a lot of things | 13:28 |
| thomasem | Ohhh, gotcha. FYI, the CRUD for clouds patch merged, it looks like, soooo the API should match up again. | 13:28 |
| thomasem | And, yeah... it's pretty rough. | 13:29 |
| sulo | yeah, i mean mostly for demo purposes ... i dont think i can use cli today | 13:29 |
| thomasem | Gotcha :( | 13:29 |
| sulo | it just doest have the things we need to show | 13:29 |
| sulo | search filters are missing ... vars | 13:29 |
| sulo | etc etc | 13:30 |
| sulo | sorting is not right | 13:30 |
| thomasem | Agreed. | 13:30 |
| thomasem | I think the most important things are that the API supports the things and (hopefully) some real inventory data in there... at least that's what I've gathered from our chats over the last several weeks. | 13:30 |
| thomasem | s/several/couple of/ | 13:31 |
| sulo | thomasem: yeah i got now .. i am going to send everyone their users in a sec to play with | 13:31 |
| thomasem | Oh, excellent. | 13:31 |
| thomasem | After the demo, let's get all of the CLI bugs entered in so we can start working them. | 13:37 |
| * sigmavirus forgot to log onto IRC this morning | 13:40 | |
| thomasem | Missed much party, you have. | 13:41 |
| * thomasem kids | 13:41 | |
| *** VW has joined #craton | 13:54 | |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Reorganize shell based integration tests https://review.openstack.org/437658 | 13:54 |
| *** VW has quit IRC | 14:00 | |
| *** VW has joined #craton | 14:00 | |
| *** VW has quit IRC | 14:05 | |
| *** VW has joined #craton | 14:06 | |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Reorganize shell based integration tests https://review.openstack.org/437658 | 14:07 |
| sulo | also i get this now | 14:08 |
| sulo | https://bugs.launchpad.net/python-cratonclient/+bug/1668221 | 14:08 |
| openstack | Launchpad bug 1668221 in Craton's Python Client "Random NoneType() < int() error from cli" [High,New] | 14:08 |
| sigmavirus | sulo: odd | 14:11 |
| farid | hi guys | 14:12 |
| sulo | yeah, it works the second or the third try though | 14:12 |
| thomasem | hey farid | 14:19 |
| sigmavirus | sulo: huh? | 14:20 |
| sigmavirus | sulo: what sha are you installing from? | 14:20 |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Add --format to the client shell https://review.openstack.org/437659 | 14:20 |
| thomasem | sigmavirus: did you not see my comments on your review here https://review.openstack.org/437658? Or do you disagree with them? | 14:21 |
| sigmavirus | For --format or for the first one? The first one, I thought I addressed them. | 14:22 |
| sigmavirus | If the latter, I missed them | 14:22 |
| thomasem | the first one... lemme double-check | 14:22 |
| thomasem | Mmm, nope. I was suggesting adding __init__.py to the new directories. | 14:22 |
| thomasem | Surprised it's working, actually? | 14:22 |
| thomasem | Considering where all you import from. | 14:23 |
| thomasem | from cratonclient.tests.integration.shell import base | 14:23 |
| thomasem | There's no __init__.py in shell/ | 14:23 |
| sigmavirus | ah, was missing it in shell, but not integration | 14:24 |
| * sigmavirus adds it and re-runs tests | 14:24 | |
| thomasem | Yeah... I think I misspoke on one of my comments. But, yeah, it looks like shell/ needs one. | 14:25 |
| sigmavirus | Adding it | 14:28 |
| thomasem | +1 | 14:28 |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Reorganize shell based integration tests https://review.openstack.org/437658 | 14:29 |
| * sigmavirus is all too used to py.test | 14:29 | |
| thomasem | Oh, that's right. | 14:30 |
| thomasem | With py.test it's actually recommended you don't package your tests. | 14:30 |
| thomasem | Because :magic: | 14:30 |
| *** valw has joined #craton | 14:33 | |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Add --format to the client shell https://review.openstack.org/437659 | 14:37 |
| *** valw_ has joined #craton | 14:37 | |
| sigmavirus | thomasem: yeah, but fixtures and parametrization are so much better in py.test | 14:37 |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Reorganize shell based integration tests https://review.openstack.org/437658 | 14:39 |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Add --format to the client shell https://review.openstack.org/437659 | 14:39 |
| thomasem | I guess I never ran into the case that required so much magic? | 14:39 |
| *** valw has quit IRC | 14:40 | |
| thomasem | sigmavirus: So, testing our your --format patch... I shouldn't be expecting it to work yet, according to the commit message? | 14:40 |
| thomasem | s/our/out/ | 14:40 |
| sigmavirus | Correct | 14:40 |
| sigmavirus | to replace existing formatting would have made that change double in size | 14:41 |
| sigmavirus | As it is it is over 848 lines of code =/ | 14:41 |
| sigmavirus | I should probably split it up further | 14:41 |
| thomasem | I see... so are we going to leave this WIP until we can get that change in? Looks like it breaks table formatting, too | 14:41 |
| thomasem | $ craton region-list | 14:42 |
| thomasem | ERROR: 'Extension' object is not callable | 14:42 |
| thomasem | rather, that change up for review | 14:42 |
| thomasem | not "in" | 14:42 |
| sigmavirus | thomasem: you're on an old version | 14:42 |
| thomasem | Oh | 14:42 |
| thomasem | Pulling | 14:42 |
| thomasem | There it goes | 14:43 |
| thomasem | Thanks! | 14:43 |
| sigmavirus | You're welcome :) | 14:43 |
| *** valw_ has quit IRC | 14:46 | |
| *** valw has joined #craton | 14:46 | |
| *** valw has quit IRC | 14:54 | |
| sulo | what does the descendants=true flag do ? | 14:55 |
| sulo | git-harry: ^ | 14:55 |
| sulo | heh want to get a quick anser without reading the code :) | 14:56 |
| git-harry | sulo: if you specify a parent_id, you get the children by default | 14:56 |
| git-harry | sulo: if you set descendants=true you get all descendants | 14:56 |
| sulo | git-harry: ah in the list itself ? | 14:56 |
| git-harry | yes | 14:56 |
| sulo | gotcha | 14:56 |
| sulo | thx | 14:56 |
| sulo | git-harry: so it addes network-devices on the list is that the thing | 14:59 |
| *** valw has joined #craton | 14:59 | |
| sulo | i am trying to figure out how this is differnt fromjust doing say host list | 14:59 |
| git-harry | sulo: it returns both network-devices and hosts | 14:59 |
| git-harry | it was added to allow more convenient queries against devices especially where descendants are involved | 15:00 |
| sulo | but this does not group that does it | 15:01 |
| sulo | what i mean is | 15:01 |
| thomasem | Are we still having our Monday meeting, or skipping for the demo? | 15:01 |
| sulo | if i do /hosts get | 15:02 |
| sulo | i get all hosts including parent or any child | 15:02 |
| sulo | minus network-devices | 15:02 |
| *** valw_ has joined #craton | 15:02 | |
| sulo | bar network-device that no differnt to doing /devices with children | 15:02 |
| *** valw has quit IRC | 15:03 | |
| git-harry | sulo: /devices provides two things not supported by /hosts or /network-devices, homogeneous collections and optionally all descendants. | 15:04 |
| farid | thomasem: I'm there just in case, whatever works better for y'all. | 15:04 |
| sulo | git-harry: so all network-devcies + any host child for that network device (but in a separate hosts list) ? | 15:05 |
| git-harry | yes | 15:05 |
| thomasem | Oh... jimbaker isn't on? | 15:06 |
| sulo | git-harry: yeah thats whats confusing me .. it would be useful to get (parent/child) for a device right ? we can almost already do this with 2 queries .. if we do /host and /network-devices we get everything we need to form that .. actually not even | 15:07 |
| sulo | because there is no descendants property | 15:07 |
| *** jimbaker has joined #craton | 15:08 | |
| *** ChanServ sets mode: +o jimbaker | 15:08 | |
| sulo | i think if we can add this for a particular device .. then its a really cool feature | 15:08 |
| sulo | anyway i guess we can discuss later | 15:09 |
| jimbaker | sorry, network connectivity issues | 15:09 |
| git-harry | sulo: yes it could be added to specific device types as well although what was added was specifically to address the current short-term requirements | 15:10 |
| git-harry | and adding it to specific device types doesn't get you everything /devices gives | 15:11 |
| sulo | git-harry: i guess what i am saying is i am failing to see how this is any different than doing /hosts /network-device and forming a parent child from that | 15:11 |
| sulo | because there is no descendants property | 15:11 |
| sulo | so you'd have to from it though parent_id ? | 15:11 |
| jimbaker | sulo, it was added | 15:11 |
| sulo | jimbaker: what was ? | 15:12 |
| jimbaker | we now compute the descendants property from the parent_id, via children iirc the specific name | 15:12 |
| git-harry | sulo: ignoring descendants there is no difference | 15:13 |
| jimbaker | reviewing the code in models.py, i'm correct in that statement | 15:13 |
| sulo | ok i havent looked at the code yet | 15:14 |
| sulo | just tying to understand this to make some examples for demo | 15:15 |
| sulo | so here https://gist.github.com/sulochan/d8f36481216a02491b0e608d2c93fdd4 | 15:15 |
| sulo | i am trying to make sense of that output | 15:15 |
| jimbaker | sulo, basically this parallels the code for ancestors (which could just use the code from descendants as-is once we implement parents) | 15:15 |
| jimbaker | (it's just a BFS traversal) | 15:16 |
| sulo | so in that output .. what iam i looking for | 15:16 |
| jimbaker | btw, we have another meeting in #openstack-meeting-4 | 15:16 |
| *** valw_ has quit IRC | 15:28 | |
| *** tojuvone_ has joined #craton | 15:30 | |
| *** valw has joined #craton | 15:37 | |
| openstackgerrit | git-harry proposed openstack/python-cratonclient master: Add support for /v1/devices https://review.openstack.org/438561 | 15:38 |
| git-harry | sulo: ^ | 15:39 |
| git-harry | that should allow you to use /devices in the client | 15:39 |
| sulo | git-harry: awesome ! thanks | 15:39 |
| sulo | git-harry: ill try to pull that in ... although i will have to mix cli with curl for demo | 15:40 |
| sulo | its not the end of the world if we cant get this to show | 15:40 |
| sulo | much of the demo is curl | 15:40 |
| git-harry | ah, I thought it was all supposed to be in the client | 15:41 |
| sulo | yeah it would be nice but cli is not ready | 15:41 |
| sulo | its missing a whole lot of things | 15:41 |
| *** Syed__ has joined #craton | 15:47 | |
| jimbaker | sulo, but ideally show some of the client in the demo | 15:49 |
| jimbaker | at least whatever there fully works with the rest of the rest api :) | 15:49 |
| sulo | jimbaker: yeah the cli is missing most of the stuff we want to show | 15:49 |
| sulo | which is why i am saying client should be the priority work this week | 15:50 |
| jimbaker | stuff like network-devices | 15:50 |
| jimbaker | etc | 15:50 |
| jimbaker | agreed about prioritizing client | 15:50 |
| jimbaker | thomasem, ^^^ | 15:51 |
| *** valw has quit IRC | 15:51 | |
| thomasem | +1 on it | 15:51 |
| sulo | git-harry: your change looks good to me | 15:53 |
| sulo | works as expected | 15:54 |
| git-harry | good to hear | 15:54 |
| jimbaker | sulo, i assume you are using some wrappers like my craton-get around curl | 16:02 |
| jimbaker | demos better imho | 16:03 |
| sulo | jimbaker: whats that ? i am not | 16:03 |
| jimbaker | https://gist.github.com/jimbaker/9089e381af38d9db6bc4af59baed6d47 | 16:05 |
| jimbaker | so craton-get v1/hosts?region_id=... | 16:05 |
| tojuvone_ | Evening | 16:05 |
| jimbaker | tojuvone_, hi | 16:05 |
| thomasem | sigmavirus: how does the --format=json support in jimbaker's CLI vars patch impact your --format patch? Are we thinking we want to drop JSON format support in jimbaker's patch in favor of yours? | 16:06 |
| thomasem | https://review.openstack.org/#/c/427032 | 16:06 |
| jimbaker | thomasem, i would assume that we should favor sigmavirus' more general impl | 16:06 |
| thomasem | That's what I figured... shall I go ahead and drop that support from the patch, then? | 16:06 |
| jimbaker | +1 | 16:06 |
| jimbaker | it's t | 16:06 |
| sigmavirus | thomasem: wasn't aware that was even there | 16:07 |
| thomasem | Surprise! | 16:07 |
| thomasem | :waves hands: | 16:07 |
| * thomasem waves hands* | 16:07 | |
| jimbaker | we will figure it out :) | 16:07 |
| thomasem | Undoubtedly | 16:07 |
| sigmavirus | eventually =P | 16:09 |
| jimbaker | eventually consistent coding | 16:09 |
| jimbaker | as we process the version graph | 16:09 |
| *** tojuvone_ has quit IRC | 16:15 | |
| *** tojuvone_ has joined #craton | 16:15 | |
| *** wirehead_ has quit IRC | 16:19 | |
| *** valw has joined #craton | 16:28 | |
| openstackgerrit | Thomas Maddox proposed openstack/python-cratonclient master: WIP: CLI and client support for get/set/delete of resource vars https://review.openstack.org/427032 | 16:35 |
| *** jovon has joined #craton | 16:49 | |
| thomasem | jimbaker: Do you mind explaining to me why if info['variables'] is None, you want to force 'loaded' to be False here: https://review.openstack.org/#/c/427032/10/cratonclient/v1/variables.py (line 29)? | 16:51 |
| jimbaker | thomasem, so this is with respect to lazy loading | 16:52 |
| jimbaker | at this point, you have not loaded variables. you don't know if it's empty {}, or not. you just have None | 16:53 |
| thomasem | Ahhhhh, okay. I see. | 16:53 |
| jimbaker | a comment on that aspect sounds like a very good idea to me! | 16:53 |
| thomasem | Yep. I also simplified it a bit. | 16:53 |
| jimbaker | fresh eyes! | 16:53 |
| jimbaker | ok, let me take a look at that review | 16:54 |
| thomasem | https://gist.github.com/thomasem/4b35997b820ace0fe8e69d3d5fdfc786 | 16:54 |
| thomasem | Oh, the patch isn't up yet. About to push one up. | 16:54 |
| thomasem | jimbaker: ^^ | 16:54 |
| openstackgerrit | Thomas Maddox proposed openstack/python-cratonclient master: WIP: CLI and client support for get/set/delete of resource vars https://review.openstack.org/427032 | 16:57 |
| jimbaker | thomasem, so that looks like a reasonable simplification of what's there. my only concern is | 16:57 |
| jimbaker | whether we properly capture that this is lazy or not | 16:57 |
| jimbaker | but we can verify that logic | 16:58 |
| thomasem | jimbaker: it's the same logic, just took out the repeated code. We can test for it. | 16:59 |
| thomasem | At least, I'm expecting it to be the same logic - hence testing being important. :P | 17:00 |
| thomasem | That's why I specifically didn't default variable to {} when it is None, but actually check for it not existing. | 17:00 |
| jimbaker | thomasem, +1 | 17:03 |
| thomasem | That's why I specifically didn't default variable to {} when it is None, but actually check for it not existing. | 17:05 |
| thomasem | oops, mt | 17:05 |
| thomasem | Meant to up arrow and enter in my shell | 17:05 |
| sigmavirus | thomasem: heh | 17:24 |
| *** valw has quit IRC | 17:41 | |
| *** david-lyle has joined #craton | 17:44 | |
| *** david-lyle has quit IRC | 17:45 | |
| *** david-lyle has joined #craton | 17:46 | |
| *** david-lyle has quit IRC | 17:58 | |
| *** david-lyle has joined #craton | 18:01 | |
| *** valw has joined #craton | 18:02 | |
| sulo | git-harry: your patch seems to have broken host-list | 18:03 |
| * thomasem backs slowly away and out of the room | 18:05 | |
| git-harry | sulo: okay, I'll take a look | 18:06 |
| sulo | thanks | 18:06 |
| *** zz_pwnall1337 is now known as pwnall1337 | 18:10 | |
| *** valw has quit IRC | 18:10 | |
| jimbaker | sulo, https://gist.github.com/jimbaker/9089e381af38d9db6bc4af59baed6d47 | 18:11 |
| jimbaker | latest craton wrappers, now based on httpie. which is a very fine package, highly worthwhile, commendable! | 18:11 |
| jimbaker | :) | 18:11 |
| jimbaker | thomasem, ^^^ | 18:11 |
| thomasem | Thanks | 18:12 |
| openstackgerrit | git-harry proposed openstack/python-cratonclient master: Add support for /v1/devices https://review.openstack.org/438561 | 18:16 |
| git-harry | sulo: that should do it for now, I'll clean it up properly tomorrow. | 18:16 |
| sulo | git-harry: thanks | 18:17 |
| openstackgerrit | Thomas Maddox proposed openstack/python-cratonclient master: WIP: CLI and client support for get/set/delete of resource vars https://review.openstack.org/427032 | 18:24 |
| *** valw has joined #craton | 18:44 | |
| *** pwnall1337 is now known as zz_pwnall1337 | 18:49 | |
| sigmavirus | jimbaker: httpie is terrible. It uses that awful Requests library | 18:53 |
| jimbaker | sigmavirus, yeah, i know. such a wrong headed decision ;) | 18:54 |
| * sigmavirus wouldn't touch that library with a 10ft pole | 18:55 | |
| jimbaker | i was going to say 3 meters, but could use the extra distancing | 18:56 |
| jimbaker | (all ~4.8 cm...) | 18:57 |
| thomasem | Will be like... 1 minute late | 18:59 |
| openstackgerrit | Thomas Maddox proposed openstack/python-cratonclient master: WIP: CLI and client support for get/set/delete of resource vars https://review.openstack.org/427032 | 18:59 |
| sigmavirus | farid: mute ;) | 19:05 |
| sigmavirus | jimbaker: thomasem what do you think about sorting those field names? | 19:17 |
| sigmavirus | especially on listing | 19:18 |
| sigmavirus | two subsequent calls should be identical | 19:18 |
| *** tojuvone_ has quit IRC | 19:18 | |
| jimbaker | yes, that's an outstanding bug | 19:18 |
| sigmavirus | ah okay | 19:18 |
| jimbaker | it should be stable. maybe logical, vs lexicographic. but regardless consistent | 19:18 |
| jimbaker | also we need to be able to show labels, etc | 19:18 |
| jimbaker | but that's a next pass | 19:19 |
| sulo | sigmavirus: +1 i've created a bug for that iirc .. also whats displayed is odd for some caes | 19:37 |
| sulo | like devices i think is showuing cell but not region etc | 19:38 |
| sulo | where cel was optional etc | 19:38 |
| sulo | also id's werent asc or desc | 19:38 |
| sigmavirus | sulo: I'm fixing that | 19:38 |
| sigmavirus | lol | 19:38 |
| sulo | ah nice | 19:38 |
| sigmavirus | (Side-benefit of fixing --format) | 19:38 |
| sulo | ok heading out for today, cya tomorrow | 19:49 |
| jimbaker | sulo, cool, thanks so much for a great demo! | 19:49 |
| *** valw has quit IRC | 19:51 | |
| *** valw has joined #craton | 19:53 | |
| cloudnull | so there was some talk in that last meeting about export formats. can we simply standardize on a single format (not ansible, not puppet) ? Asking because as a downstream consumer I'd like to simply retrieve and parse the data in a way that makes sense for me. I'd rather not be forced to adopt an output format that changes parts of my runtime. | 19:53 |
| cloudnull | IE: export of devices comes out as craton format? | 19:53 |
| sulo | cloudnull: +1 that makes the most sense tbh | 19:55 |
| cloudnull | ++ | 19:55 |
| cloudnull | this would make life so much simpler. IMHO. | 19:56 |
| cloudnull | simply tell me how to format the data for input and what I can expect on export. | 19:57 |
| cloudnull | building a plugin within ansible, puppet, chef, ect would be on the consumer and should be fairly simple. | 19:57 |
| sigmavirus | I suspect that if we can get some good APIs defined around that | 19:58 |
| sigmavirus | we could even get a craton inventory plugin into upstream ansible | 19:58 |
| jimbaker | +100 | 19:59 |
| farid | well that's a new, I saw the mute request an hour late | 20:01 |
| farid | between vidyo, slack and IRC I need a better highlight system than what I have. | 20:02 |
| sigmavirus | farid: :) | 20:06 |
| sigmavirus | your keyboard is loud | 20:07 |
| sigmavirus | Which is a complaiint I get often | 20:07 |
| sigmavirus | Sometimes even from my fiancée's coworkers | 20:07 |
| farid | hehe | 20:13 |
| farid | yeah I hear you :) | 20:13 |
| jimbaker | farid's keyboard is loud. very vintage | 20:13 |
| jimbaker | i'm thinking it must be an ibm model m - https://en.wikipedia.org/wiki/Model_M_keyboard | 20:14 |
| farid | it's a das, I've it had for almost 5 years now thinking about it... | 20:14 |
| farid | they were a rackspace customer and gave us a racker discount at some point so I got one | 20:15 |
| farid | it's the closest to a model m I could find though :) | 20:15 |
| jimbaker | nice! | 20:15 |
| farid | wow they're expensive now, they used to go for 100 and now I see 133 as the lower price | 20:16 |
| farid | think I ended up paying like 85-90 with the discount back then | 20:16 |
| farid | ah no they have some for 99... | 20:16 |
| farid | allright gonna transfer home bbiab | 20:16 |
| jimbaker | later. biab myself, time for a run | 20:17 |
| sigmavirus | farid: I am a mechanical keyboard junky myself. We should chat sometime | 20:18 |
| sigmavirus | =) | 20:18 |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Add --format to the client shell https://review.openstack.org/437659 | 20:33 |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Update hosts commands to use pluggable formatters https://review.openstack.org/438693 | 20:33 |
| sigmavirus | jimbaker: thomasem ^ will do the needful formatting =) | 20:42 |
| sigmavirus | but just for hosts | 20:42 |
| sigmavirus | going to work on the other commands tomorrow. Looks like some of that test refactoring I did ended up benefiting me instead of biting me in the arse. That said, I'm confused why some of the tests are passing so | 20:42 |
| *** zz_pwnall1337 is now known as pwnall1337 | 21:24 | |
| *** pwnall1337 is now known as zz_pwnall1337 | 21:33 | |
| *** VW has quit IRC | 21:40 | |
| *** valw has quit IRC | 21:42 | |
| *** zz_pwnall1337 is now known as pwnall1337 | 21:42 | |
| thomasem | sigmavirus: Yay! | 21:49 |
| thomasem | Oh, man. I just can't deal with how audible mechanical keyboards are. Tried one for a short bit and just got tired of the sound, and I was using one intended to be fairly quiet. | 21:51 |
| *** VW has joined #craton | 22:08 | |
| *** VW has quit IRC | 22:13 | |
| openstackgerrit | Thomas Maddox proposed openstack/python-cratonclient master: WIP: CLI and client support for get/set/delete of resource vars https://review.openstack.org/427032 | 22:17 |
| *** jovon has quit IRC | 22:22 | |
| thomasem | Thanks for stitching that demo together. I know that was probably a bit frustrating with all of the stuff that hadn't been shored up yet. It looked like it was received really well and I just wanted to thank you for doing that. | 22:23 |
| jimbaker | big kudos to pwnall1337 and sulo! | 22:24 |
| jimbaker | of course a big team effort all around | 22:24 |
| jimbaker | thomasem, re https://review.openstack.org/#/c/427032/, still WIP, right? | 22:24 |
| thomasem | Yep. Fleshing out some functional tests | 22:25 |
| thomasem | jimbaker: ^^ | 22:25 |
| jimbaker | cool, just checking | 22:25 |
| thomasem | Sure thing | 22:25 |
| thomasem | I did some clean up to fix up a bunch of pep8 and other little errors I ran into using it, so | 22:25 |
| thomasem | And added in clouds, etc. Should just be tests now | 22:26 |
| jimbaker | thomasem, btw, i was thinking for vars, we should support this format for nested https://github.com/jkbrzt/httpie#non-string-json-fieldsjson: | 22:26 |
| jimbaker | vs the hackiness that is done now | 22:27 |
| thomasem | jimbaker: Is that not going to be a part of the generic JSON formatting stuff? | 22:27 |
| jimbaker | thomasem, no this is the CLI side of things | 22:27 |
| thomasem | jimbaker: I'm talking about sigmavirus's patches for the CLI | 22:27 |
| thomasem | That's intended to allow us to format any output from the CLI as JSON | 22:28 |
| jimbaker | yeah, just want to keep it separated. so this is for invocations like so: $ craton cell-vars-set 13 x= y=42 | 22:28 |
| jimbaker | x is deleted, y is set to 42 | 22:28 |
| jimbaker | but what about craton cell-vars-set 13 x:='{arbitrary-json-goes-here...}' | 22:29 |
| jimbaker | obviously we can do this now by consuming from stdin | 22:29 |
| thomasem | How about we get tests and such written for what's already supported first? | 22:30 |
| jimbaker | :) | 22:30 |
| thomasem | I agree that it'd be nice, but I also don't want to snowball this patch | 22:30 |
| jimbaker | thomasem, yes, i agree with that sentiment 100% | 22:30 |
| thomasem | Excellent | 22:30 |
| jimbaker | sorry, didn't want to derail getting that important stuff done first | 22:30 |
| thomasem | No worries. I definitely agree that it'd be nice. I've used that format quite a bit. | 22:31 |
| jimbaker | just wanted to capture this, and see if it made sense. given i had just experience with using httpie for real as of today | 22:31 |
| thomasem | Ohhhhhhhhh | 22:31 |
| jimbaker | definitely makes for a better experience\ | 22:31 |
| thomasem | Yeah, it's nice. I've used that for a couple years now. | 22:31 |
| thomasem | Came in SO handy | 22:31 |
| thomasem | Let's capture it in the LP | 22:31 |
| jimbaker | cool. it seems that the wrappers i put together just give it some reasonable defaults for headers | 22:31 |
| jimbaker | brb | 22:32 |
| *** VW has joined #craton | 22:33 | |
| thomasem | Added to the LP | 22:34 |
| *** VW has quit IRC | 22:51 | |
| *** VW has joined #craton | 22:51 | |
| *** VW_ has joined #craton | 23:07 | |
| *** VW has quit IRC | 23:10 | |
| *** VW_ has quit IRC | 23:11 | |
| *** VW has joined #craton | 23:21 | |
| *** VW has quit IRC | 23:25 | |
| farid | sigmavirus: nice do you have the model m? sorry was telling thomas I got back home and jumped on slack and for some reason I Didn't open irc | 23:54 |
| farid | the fancyness | 23:54 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!