| *** VW has quit IRC | 00:01 | |
| *** VW_ has quit IRC | 00:02 | |
| *** VW has joined #craton | 00:27 | |
| *** VW has quit IRC | 00:32 | |
| *** VW has joined #craton | 00:42 | |
| *** VW has quit IRC | 00:43 | |
| *** VW has joined #craton | 00:43 | |
| *** Syed__ has quit IRC | 01:05 | |
| *** VW has quit IRC | 01:28 | |
| *** VW has joined #craton | 02:01 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/craton master: Updated from global requirements https://review.openstack.org/439212 | 04:05 |
|---|---|---|
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-cratonclient master: Updated from global requirements https://review.openstack.org/439231 | 04:15 |
| *** pwnall1337 is now known as zz_pwnall1337 | 04:25 | |
| *** VW has quit IRC | 05:13 | |
| openstackgerrit | Merged openstack/python-cratonclient master: Updated from global requirements https://review.openstack.org/439231 | 06:07 |
| *** sigmavirus has quit IRC | 06:52 | |
| *** sigmavirus has joined #craton | 06:53 | |
| *** sigmavirus is now known as Guest53153 | 06:54 | |
| sulo | o/ | 09:02 |
| openstackgerrit | sulochan-acharya proposed openstack/craton master: Allow resolved vars in details list calls https://review.openstack.org/439175 | 09:13 |
| sulo | hey git-harry, when you get some time can you take a look at https://review.openstack.org/#/c/439175/ please | 09:25 |
| git-harry | sulo: will do | 09:44 |
| openstackgerrit | sulochan-acharya proposed openstack/python-cratonclient master: Add filter params to host-list command https://review.openstack.org/439549 | 11:44 |
| *** Guest53153 is now known as sigmavirus | 12:10 | |
| *** sigmavirus has quit IRC | 12:10 | |
| *** sigmavirus has joined #craton | 12:10 | |
| *** ChanServ sets mode: +o sigmavirus | 12:10 | |
| *** VW has joined #craton | 12:44 | |
| *** VW has quit IRC | 12:48 | |
| *** VW has joined #craton | 12:49 | |
| sigmavirus | thomasem: sulo are we requiring that hosts be in a region or can they just be in a cloud? | 13:05 |
| sulo | sigmavirus: so far we are assuming region | 13:05 |
| sigmavirus | sulo: okay, just making sure | 13:06 |
| *** VW has quit IRC | 13:24 | |
| *** VW has joined #craton | 13:30 | |
| *** VW has quit IRC | 14:11 | |
| thomasem | Haha, yeah. Good question. | 14:17 |
| farid | morning guys | 14:22 |
| thomasem | hihi farid | 14:29 |
| *** tojuvone has quit IRC | 14:44 | |
| *** VW has joined #craton | 14:45 | |
| openstackgerrit | Ian Cordasco proposed openstack/python-cratonclient master: Relax requirement of region-id for hosts and cells https://review.openstack.org/439629 | 14:48 |
| sigmavirus | sulo: jimbaker git-harry ^ | 14:49 |
| *** VW has quit IRC | 14:52 | |
| *** tojuvone has joined #craton | 14:53 | |
| sigmavirus | How much do we hate people self-approving things? | 15:00 |
| sigmavirus | At this point, I honestly feel like a +1 from thomasem is good enough for me to self-approve things =P | 15:00 |
| sigmavirus | I'm also looking through our client reviews and seeing other things that could have been approved sooner | 15:01 |
| openstackgerrit | Merged openstack/python-cratonclient master: Add filter params to host-list command https://review.openstack.org/439549 | 15:04 |
| openstackgerrit | Merged openstack/python-cratonclient master: Fix help messages for resource-listing commands https://review.openstack.org/432354 | 15:09 |
| sigmavirus | thomasem: I'm confused by the current state of your client vars patch | 15:09 |
| thomasem | sigmavirus: I think it's been carte blanche until this deliverable. And, I appreciate your comment! :P | 15:13 |
| thomasem | sigmavirus: I'm a bit confused about it myself. I'm in the process of making it generic for other sub resources according to jimbaker's comments. I added a Variable class and extended Resource to understand that when it encounters "variables" when parsing "info" in _add_details, that it'll send that payload to a different handler to add a list of Variable objects to the resource at self._variables. | 15:15 |
| thomasem | I am just fixing that change up and then going to push it up. I believe that's more in-line with what you were suggesting? | 15:15 |
| sigmavirus | I just commented with what I was suggesting | 15:15 |
| thomasem | Looking! | 15:15 |
| sigmavirus | git-harry: ping | 15:15 |
| *** VW has joined #craton | 15:16 | |
| git-harry | sigmavirus: yep? | 15:16 |
| sigmavirus | https://review.openstack.org/438561 is this dependent on a server change? If so, you can add "Depends-On: <change-id>" so the tests reflect the true status of the patch | 15:16 |
| git-harry | No, it's not. That side of it has already merged. | 15:16 |
| sigmavirus | Ah okay | 15:17 |
| sigmavirus | Do you want a hand with those tests? | 15:17 |
| sigmavirus | I'm wary to refactor our client test suite as much as I'd like to with our deadline coming up and some large changes that need to be landed | 15:17 |
| * sigmavirus brb | 15:18 | |
| openstackgerrit | Merged openstack/python-cratonclient master: Relax requirement of region-id for hosts and cells https://review.openstack.org/439629 | 15:19 |
| git-harry | No, I'm fine thanks | 15:20 |
| thomasem | sigmavirus: excellent, that's pretty close to what I have now, minor tweak. | 15:20 |
| thomasem | Thanks for the clarification. | 15:21 |
| thomasem | I'll just go with name/value for now. We can add audit stuff later on. | 15:21 |
| thomasem | Since... well, we don't have anything for that. :P | 15:21 |
| thomasem | The patch I put up yesterday was more exploratory, so that's very helpful. I unraveled a lot of "how the hell does this work?" | 15:22 |
| openstackgerrit | Merged openstack/craton master: Updated from global requirements https://review.openstack.org/439212 | 15:24 |
| thomasem | jimbaker: sigmavirus: It seems there's some disagreement between y'all on how we should work with build_url regarding how generic we make this, whether we use variables_key, or some other method of specifying the subresource for the path? | 15:29 |
| thomasem | In my local patch, according to jimbaker's comment on the already known conventions, I actually did an enumeration of SubResourceTypes that map to the key to be added. | 15:29 |
| thomasem | Actually, let me clean it up a bit and push it up so y'all can see what I have. | 15:30 |
| 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 | 15:32 |
| 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 | 15:34 |
| 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 | 15:37 |
| thomasem | I honestly feel like we're overcomplicating this whole thing a bit, but that's probably due to lack of context. | 15:39 |
| thomasem | Especially for right now - trying to make the CRUDClient methods generic... when we implement labels, it's going to mean now we have to differentiate the arguments to set_subresource, since that'll be a list for labels, and a dict for variables | 15:40 |
| thomasem | I think delete/get will remain the same, though, as long as we have handlers for each type of return from the API. | 15:41 |
| thomasem | But, I do feel like I'm getting randomized on things not yet decided on for this patch, so I'd like for us to agree so I'm not just going back and forth. | 15:42 |
| thomasem | For me, personally, I'm more inclined to go with something simpler now and we can change as need arises. I do wonder if we're not trying to solve for problems we don't yet know we'll have. | 15:44 |
| thomasem | But, again, could be lack of context on my part. | 15:44 |
| sulo | thomasem: i thought you were already doing that | 15:44 |
| thomasem | Doing what? | 15:45 |
| sulo | thomasem: simplyfying the code | 15:45 |
| sigmavirus | thomasem: I'm already confused as to why you're adding methods to the formatters classes, that seems like a complication too =) | 15:45 |
| thomasem | sigmavirus: I removed those | 15:46 |
| sigmavirus | ah okay | 15:46 |
| * sigmavirus looks at new patch before commenting further | 15:46 | |
| thomasem | Yeah... I would prefer the formatter know how to parse the variables | 15:46 |
| thomasem | And, keep in mind, this is a pretty rough representation. I didn't see the need to polish when there seems to be a divide on some details. | 15:47 |
| * sigmavirus nods | 15:48 | |
| thomasem | So, 1) variables_key or subresource passed in (I'd prefer an Enum for this)? 2) path_arguments or regular arguments? 3) Generic get/set/delete_subresource, or more concrete? | 15:48 |
| farid | hey guys I have a meeting overlap with the sync today, which I think I can get out and come back to but if I don't show up please proceed as usual :) | 15:49 |
| sigmavirus | Okay | 15:49 |
| sigmavirus | I think I know where things are going awry | 15:49 |
| thomasem | Hmm? | 15:49 |
| sulo | farid: ok | 15:50 |
| thomasem | And we can hop on Vidyo or whatever helps, too. | 15:50 |
| farid | sulo: victor confirmed he'll come too, he mentioned being interested in knowing overview of the architecture of craton, which components it uses and such... if not during maybe after the demo | 15:53 |
| sulo | farid: rgr that | 15:53 |
| farid | I mentioned I think it's straight out python flask + mysql which he said if that's it that's all he needed to know | 15:53 |
| farid | think at some point he heard zookeeper and other stuff in there | 15:53 |
| farid | so wanted to find out. | 15:53 |
| sulo | farid: its in the diagram but we dont have in un use | 15:53 |
| sulo | farid: at this point its out of the picture | 15:54 |
| farid | ah gotcha, ok I understand where he got it from now then | 15:54 |
| sulo | so py + flask + sql is prettymuch it | 15:54 |
| farid | awesome, we'll mention it then | 15:55 |
| sigmavirus | thomasem: yeah, so in all candor I think there needs to be a bit more flexibility in the crud module rather than subclassing | 15:58 |
| sigmavirus | I'm trying to untangle that right now | 15:58 |
| sigmavirus | Actually a Vidyo call might make sense in Craton-Recordable | 15:58 |
| *** jovon has joined #craton | 16:00 | |
| thomasem | See you there | 16:00 |
| jimbaker | right, zookeeper is specifically for job board support in taskflow, which iirc, sulo demoed how to use in the summer. very much not required for cmdb | 16:15 |
| jimbaker | sigmavirus, thomasem, sounds good about getting that crud resource work done. also i do not want to delay current support for vars to attempt to future proof, just asking questions now in case they seem relevant | 16:17 |
| thomasem | jimbaker: yeah, we had a chat. Going to try going for something more generic in the resource itself with subresource_managers to build the URL that way. | 16:24 |
| jimbaker | thomasem, sounds good | 16:24 |
| thomasem | sigmavirus is sending over a little patch that he hacked on a bit to kind of point out what we're going for. | 16:25 |
| jimbaker | thomasem, also in terms of those specifics: the reason why not variables for everything is precisely because of differences like sets (for labels) and triples (for scoped assignments for rbac) | 16:26 |
| jimbaker | so good push back on the api considerations | 16:26 |
| thomasem | jimbaker: gotcha | 16:31 |
| openstackgerrit | sulochan-acharya proposed openstack/craton master: Enable vars resolution in cells and regions https://review.openstack.org/439695 | 16:33 |
| sulo | ^^ enabling resolved vars in regions and cells .. pls review | 16:34 |
| jimbaker | sulo, thanks | 16:40 |
| jimbaker | sulo, just to fix the minor pep8 | 16:41 |
| openstackgerrit | sulochan-acharya proposed openstack/craton master: Enable vars resolution in cells and regions https://review.openstack.org/439695 | 16:42 |
| sulo | i vidyo working for everyone ? its not for me. | 16:57 |
| farid | https://bugs.launchpad.net/python-cratonclient/+bug/1667109 | 16:57 |
| openstack | Launchpad bug 1667109 in Craton's Python Client "Allow environment export from client" [Undecided,New] | 16:57 |
| sigmavirus | jimbaker: would it help if I took over https://bugs.launchpad.net/craton/+bug/1668308 ? | 16:58 |
| openstack | Launchpad bug 1668308 in craton "orphan variables on host delete" [High,Confirmed] - Assigned to Jim Baker (jimbaker) | 16:58 |
| *** VW_ has joined #craton | 16:59 | |
| *** VW has quit IRC | 17:02 | |
| *** VW_ has quit IRC | 17:03 | |
| *** VW has joined #craton | 17:04 | |
| sigmavirus | farid: is that a priority? | 17:05 |
| sigmavirus | did vidyo die for everyone else? | 17:06 |
| sigmavirus | or just me? | 17:06 |
| farid | sigmavirus: it's in the milestone, had committed to and just noticed it's unassigned | 17:06 |
| git-harry | yup | 17:06 |
| farid | yeah died for me too | 17:06 |
| jimbaker | yes, died for me | 17:06 |
| sigmavirus | farid: I'll take it then | 17:06 |
| jimbaker | so maybe vidyo dying is a sign | 17:06 |
| farid | sigmavirus: I just pinged cloudnull about it and he's going to take a glance at it, maybe y'all can chat ? | 17:07 |
| sulo | vidyo hasnt worked for me for last 30 mins | 17:07 |
| jimbaker | i cannot rejoin | 17:07 |
| farid | since he (tmk) hasn't worked with craton in the past but is available to help this week | 17:07 |
| farid | yeah can't rejoin either | 17:07 |
| farid | wonder if aws is at it again | 17:07 |
| jimbaker | git-harry, sigmavirus, thomasem - let's revisit this question later about resolution | 17:08 |
| jimbaker | i think it needs a spec :) | 17:08 |
| thomasem | mmmkay | 17:08 |
| thomasem | Agreed. | 17:08 |
| thomasem | Also, Vidyo said it didn't want us. | 17:08 |
| farid | looks like it's working now | 17:08 |
| *** VW has quit IRC | 17:08 | |
| jimbaker | and that hurts :) | 17:08 |
| thomasem | It does. I feel the rejection. | 17:08 |
| jimbaker | so for those in that call just now, we are just to going keep in on #craton. or just work on our tasks | 17:09 |
| * sigmavirus is going to shovel his fiancee a path out of the garage | 17:11 | |
| farid | o/ | 17:11 |
| openstackgerrit | git-harry proposed openstack/python-cratonclient master: Add devices-list to support /v1/devices https://review.openstack.org/438561 | 17:13 |
| sulo | so is vidyo working for anyone ? | 17:18 |
| *** Syed__ has joined #craton | 17:18 | |
| *** VW has joined #craton | 17:34 | |
| *** VW has quit IRC | 17:43 | |
| *** VW has joined #craton | 17:45 | |
| openstackgerrit | git-harry proposed openstack/python-cratonclient master: Add devices-list to support /v1/devices https://review.openstack.org/438561 | 18:01 |
| *** VW has quit IRC | 18:12 | |
| thomasem | sigmavirus: mmm shoveling snow. How fun. | 18:13 |
| * thomasem has zero experience doing this. | 18:14 | |
| sigmavirus | I come from NJ where noreasters are common | 18:18 |
| sigmavirus | So Wisconsin snow is not that bad | 18:18 |
| sigmavirus | Went from 1-2 blizzards a year to snowing at sometimes once a week | 18:18 |
| sigmavirus | The worst part here is the temperatures that drop below -10F | 18:19 |
| thomasem | Ouch | 18:26 |
| thomasem | I can only complain about the other extremes here. :P | 18:28 |
| *** VW has joined #craton | 18:33 | |
| sigmavirus | https://blog.cloudflare.com/quantifying-the-impact-of-cloudbleed/ | 18:37 |
| sigmavirus | Oh we get temps over 90 but that's like a few days a year | 18:37 |
| sigmavirus | It's pretty great | 18:37 |
| sigmavirus | I'll take the extreme cold over the extreme heat | 18:37 |
| thomasem | I have to agree, but I say that having never had to deal with anything under 10F | 18:37 |
| *** VW has quit IRC | 18:38 | |
| thomasem | But this getting to 110F during the summer... yeah. | 18:38 |
| thomasem | Step outside and you have a heat stroke. | 18:38 |
| thomasem | Anyone know off the top of their head what's swallowing tracebacks in the CLI when things go wrong? Is it just a flag or something I can flip? | 18:39 |
| thomasem | tracking down problems with nothing but messages is a bit difficult. :P | 18:39 |
| sigmavirus | Yeah, that's another thing I need to work on (repeatable --debug flag) | 18:40 |
| thomasem | Lol | 18:40 |
| * thomasem throws it on the pile | 18:41 | |
| thomasem | https://bugs.launchpad.net/python-cratonclient/+bug/1666734 | 18:41 |
| openstack | Launchpad bug 1666734 in Craton's Python Client "Support debug logging for client/CLI" [High,New] | 18:41 |
| *** VW has joined #craton | 18:43 | |
| *** harlowja has quit IRC | 18:43 | |
| jimbaker | sigmavirus, yeah, that would be a nice thing to have for sure | 18:44 |
| jimbaker | --debug | 18:44 |
| *** harlowja has joined #craton | 18:46 | |
| jimbaker | as for the specifics in that bug report, that's merely a suggestion on what to do | 18:46 |
| openstackgerrit | Merged openstack/craton master: Enable vars resolution in cells and regions https://review.openstack.org/439695 | 18:50 |
| *** VW_ has joined #craton | 18:59 | |
| thomasem | https://bugs.launchpad.net/python-cratonclient/+bug/1666734 | 19:00 |
| thomasem | Whoops, sorry. | 19:00 |
| openstack | Launchpad bug 1666734 in Craton's Python Client "Support debug logging for client/CLI" [High,New] | 19:00 |
| *** VW_ has quit IRC | 19:01 | |
| *** VW_ has joined #craton | 19:01 | |
| *** VW has quit IRC | 19:03 | |
| thomasem | Hmmm... well now the VariableManager needs a resource_class | 19:04 |
| thomasem | Which leads us back to the implementation jimbaker had before | 19:05 |
| thomasem | somewhat | 19:05 |
| thomasem | sigmavirus: got a moment for Vidyo? | 19:06 |
| thomasem | We might want to expose variables in a more controlled manner rather than making them attributes of a resource class. | 19:14 |
| thomasem | variables = cc.projects.get(args.id).variables.get()... thinking about after that how we'd want to interact with variables. | 19:16 |
| thomasem | variables.items() is easy, since I can just access self._info | 19:17 |
| *** VW_ has quit IRC | 19:19 | |
| *** VW has joined #craton | 19:20 | |
| cloudnull | jimbaker: thomasem: antonym: sigmavirus: do any of you have a craton install I can point to for some of this client work ? | 19:22 |
| cloudnull | or should i kick my own? | 19:22 |
| cloudnull | IE is there a dev deploy I can beat on ? | 19:22 |
| thomasem | cloudnull: I would kick my own. You can pull the craton repo and run it locally: https://gist.github.com/jimbaker/e74a7b98bc60519033fd455a22163ad2 | 19:22 |
| thomasem | if you have Docker | 19:23 |
| thomasem | it's pretty straightforward | 19:23 |
| thomasem | There's also this little thing for setting up creds for use with a test deployment https://gist.github.com/jimbaker/9089e381af38d9db6bc4af59baed6d47#gistcomment-2008361 | 19:24 |
| thomasem | cloudnull: ^^ | 19:24 |
| cloudnull | do we have a db dump that I can use or should i run the "generate_fake_data.py" script ? | 19:25 |
| thomasem | cloudnull: generate_fake_data.py works decent enough for development. Might be useful to use the script that zz_pwnall1337 and sulo were using. | 19:26 |
| thomasem | With the inventory data | 19:26 |
| cloudnull | where's that ? | 19:26 |
| thomasem | But I'm unfamiliar with that set up at the moment. | 19:26 |
| cloudnull | ok | 19:26 |
| thomasem | hopefully one of them can point you to it. | 19:26 |
| cloudnull | kk | 19:26 |
| thomasem | But, to start, generate_fake_data.py should get you goign. | 19:26 |
| sigmavirus | thomasem: sorry, did not hear your ping | 19:26 |
| thomasem | going* | 19:27 |
| sigmavirus | I can vidyo | 19:27 |
| thomasem | sigmavirus: All good! | 19:27 |
| thomasem | Okay, excellent. Craton-recordable? | 19:27 |
| sigmavirus | omw | 19:28 |
| antonym | cloudnull: yeah, i have one you can poke on | 19:29 |
| *** VW has quit IRC | 19:31 | |
| *** VW has joined #craton | 19:32 | |
| *** VW has quit IRC | 19:37 | |
| *** VW has joined #craton | 19:37 | |
| *** VW has quit IRC | 19:38 | |
| *** VW has joined #craton | 19:38 | |
| farid | cloudnull: on importing ansible facts, since craton has a key-value variable feature, have any thoughts on nested properties the ansible hostvars has ? zz_pwnall1337 stopped by to ask and suggest importing all vars as a single key/value pair blob but I don't see that helping when trying to filter across multiple clouds | 19:42 |
| sigmavirus | oh thomasem that's the other thing ^ | 19:43 |
| sigmavirus | variables could be simple key=value | 19:43 |
| sigmavirus | or they could be more complex like key={"subkey": "subvalue"} | 19:43 |
| sigmavirus | so there's so much fun to be had | 19:43 |
| thomasem | Yeah... are we doing something more than top-level formatting? :) | 19:44 |
| thomasem | (except JSON, of course) | 19:44 |
| thomasem | Oh... right, so I'll have to actually account for that when assembling the dict, huh. | 19:44 |
| thomasem | Oh boy! | 19:45 |
| sigmavirus | thomasem: well, as it is currently formatted, it looks kind of awful too | 19:45 |
| sigmavirus | You could always do some kind of pretty-printing for the value | 19:45 |
| thomasem | That's true. | 19:46 |
| cloudnull | thomasem: sigmavirus: it's not uncommon to see something like this """{'other': {'stuff': [1, 2, 3, 4, 5, {"blah": 12, "stuff": ["haha12"]}]}, 'things': 'A'}""" coming from ansible facts and user defined variables. | 19:47 |
| cloudnull | especially if we take from core | 19:47 |
| cloudnull | in long format -- http://cdn.pasteraw.com/4wauhrlm3dj1z713seziozask3h70vu | 19:49 |
| sigmavirus | thomasem: to be clear, for the variable formatter we'll likely also want to rename the table column names | 19:49 |
| thomasem | sigmavirus: what to? | 19:50 |
| sigmavirus | cloudnull: you and I have that problem where we can write syntactically valid JSOn without problem | 19:50 |
| sigmavirus | thomasem: so right now it's "property" and "value" by default | 19:50 |
| thomasem | Right | 19:50 |
| sigmavirus | I'd think "Variable" and "Value" or something would be decent | 19:50 |
| thomasem | Ahhh, right. | 19:50 |
| sigmavirus | just because we can | 19:50 |
| cloudnull | sigmavirus: ++ i wrote it then looked at it and thought maybe i should expand it | 19:50 |
| thomasem | Gotcha lol | 19:50 |
| sigmavirus | so just .configure(dict_property="Variable") | 19:50 |
| thomasem | Cool | 19:51 |
| cloudnull | farid: I'm not super familiar with the json table scheme in mysql so im not 100% on it's capabilities. | 19:51 |
| thomasem | I'm assuming we want {"key": Variable("key", {"sub-key": Variable("sub-key", "value")})}? | 19:51 |
| cloudnull | if it can store nested data then great! | 19:51 |
| cloudnull | otherwise i don't know if there's an option other than using a blob | 19:51 |
| thomasem | And not {"key": Variable("key", Variable("sub-key", "value"))} | 19:52 |
| cloudnull | psql, cassandra, mongo, couchdb, etc dont have an issue with this. | 19:53 |
| cloudnull | being that we're using sqlalchemy it may be worth taking a look at psql ? | 19:53 |
| cloudnull | are we saying that craton does not work with complex variables at this moment in time? | 19:54 |
| * cloudnull not trolling, honestly asking | 19:54 | |
| *** VW_ has joined #craton | 19:55 | |
| *** VW_ has quit IRC | 19:58 | |
| *** VW_ has joined #craton | 19:58 | |
| *** VW_ has quit IRC | 19:58 | |
| *** VW_ has joined #craton | 19:59 | |
| *** VW has quit IRC | 19:59 | |
| *** VW_ has quit IRC | 19:59 | |
| *** VW has joined #craton | 20:00 | |
| *** VW has quit IRC | 20:01 | |
| *** VW has joined #craton | 20:01 | |
| jimbaker | cloudnull, right now this is more of a function of our usage of sqlalchemy_utils.types.json, which is a third party package for sqlalchemy | 20:02 |
| *** VW has quit IRC | 20:02 | |
| *** openstackgerrit has quit IRC | 20:03 | |
| jimbaker | it's support for json is better on postgresql. it could be made better on mysql, especially given mysql 8 when it gets released | 20:03 |
| *** VW has joined #craton | 20:03 | |
| jimbaker | relational db support for json has steadily been increasing, so that's a good long term trend for us. currently mysql 5.7 does have decent support for json indexing, but mysql 8 does improve that quite a bit | 20:04 |
| jimbaker | getting it closer to postgres | 20:05 |
| *** harlowja has quit IRC | 20:08 | |
| *** openstackgerrit has joined #craton | 20:29 | |
| openstackgerrit | Ian Cordasco proposed openstack/craton master: Update schema definition for network interfaces port https://review.openstack.org/439832 | 20:29 |
| * sigmavirus shakes fist at clouds | 20:30 | |
| sigmavirus | stop snowing damnit | 20:31 |
| thomasem | What'd I do? | 20:31 |
| thomasem | Ohhhhhhhh | 20:31 |
| thomasem | You mean actual clouds. | 20:31 |
| * sigmavirus thought that was vague enough to elicit entertaining responses | 20:32 | |
| * sigmavirus is happy to have been right | 20:32 | |
| thomasem | :P | 20:34 |
| * thomasem was happy to oblige | 20:34 | |
| thomasem | Man, you go so long without writing anything recursive and it's like an atrophied muscle. | 20:35 |
| thomasem | That took way longer than it should have. | 20:35 |
| sigmavirus | thomasem: yeahhhhhhh, I know what you mean | 20:35 |
| sigmavirus | I haven't made a tree in long enough that I can't remember how to balance them | 20:36 |
| * sigmavirus wouldn't make it through customs | 20:36 | |
| thomasem | ... had to restart. All of a sudden my keyboard was on the fritz? | 20:51 |
| cloudnull | jimbaker: so tl;dr no? | 20:52 |
| openstackgerrit | sulochan-acharya proposed openstack/craton master: Fake data: Generate child container per host https://review.openstack.org/439846 | 20:55 |
| *** harlowja has joined #craton | 21:01 | |
| *** harlowja has quit IRC | 21:01 | |
| *** harlowja has joined #craton | 21:01 | |
| openstackgerrit | sulochan-acharya proposed openstack/craton master: Fake data: Generate child container per host https://review.openstack.org/439846 | 21:11 |
| *** VW has quit IRC | 21:26 | |
| *** valw has joined #craton | 21:42 | |
| jimbaker | cloudnull, short answer, yes, it's just json. long answer, we can always improve how we work with json | 21:42 |
| cloudnull | so we do support complex vars? | 21:43 |
| cloudnull | IE http://cdn.pasteraw.com/4wauhrlm3dj1z713seziozask3h70vu | 21:43 |
| cloudnull | or are vars limited to key=value | 21:45 |
| cloudnull | where value is a string or int | 21:45 |
| jimbaker | value is arbitrary json | 21:47 |
| jimbaker | cloudnull, the only thing we do is that variables can be thought of as a dictionary of unicode keys mapping to json values | 21:49 |
| jimbaker | this allows us to do various things, including in the future, with respect to "top level keys" | 21:50 |
| 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 | 21:50 |
| jimbaker | such as the override model we have, based on ChainMap (and possibly extended, see https://bugs.launchpad.net/craton/+bug/1665050 | 21:51 |
| openstack | Launchpad bug 1665050 in craton "Support alternative variable overrides" [Undecided,New] | 21:51 |
| thomasem | errr... git review is blowing up :( | 21:54 |
| 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 | 21:54 |
| thomasem | there it goes | 21:54 |
| *** valw has quit IRC | 21:54 | |
| cloudnull | ^ that is what i was looking for . | 21:56 |
| cloudnull | seems like now it will support complex vars. | 21:56 |
| 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 | 21:56 |
| cloudnull | assuming that patch works. | 21:56 |
| * cloudnull has not pulled it yet | 21:56 | |
| cloudnull | thanks thomasem :) | 21:57 |
| farid | nice | 21:57 |
| thomasem | cloudnull: Are you referring to my VERY WIP patch? :D | 21:58 |
| cloudnull | yup | 21:58 |
| thomasem | Oh, you bet! | 21:58 |
| thomasem | I was fixing up the nested dict handling | 21:58 |
| cloudnull | that patch has been the question i've been trying to get an answer to. | 21:58 |
| thomasem | Ohhhhhhhhh | 21:58 |
| * cloudnull maybe was not using the right words | 21:59 | |
| thomasem | Yeah. It's been a bit of a journey getting this to work with the OpenStack client way of doing things. :P | 21:59 |
| thomasem | So, hoping we can put this one to bed in the next day or so. | 21:59 |
| thomasem | But, I'll spare you my rants. | 22:00 |
| thomasem | For now, at least. | 22:00 |
| thomasem | :P | 22:00 |
| thomasem | heyyy https://gist.github.com/thomasem/12cf2a299eb0ebc5cf06f59175555bf4 | 22:03 |
| thomasem | winning | 22:03 |
| thomasem | cloudnull: ^^ | 22:04 |
| thomasem | Is that what you were wondering | 22:04 |
| farid | nice thomasem , looks solid... will there be a capability to search/filter on thomas.what.on's value? | 22:05 |
| *** VW has joined #craton | 22:05 | |
| thomasem | I do not know, actually. I suspect we'll want to, do not know if that's implemented yet, though. | 22:06 |
| thomasem | farid: ^^ | 22:06 |
| farid | cool I think that's something that we need to figure out | 22:06 |
| thomasem | Yeah, I think that'll be important. Would appreciate jimbaker's thoughts | 22:06 |
| farid | since we'll be importing ansible facts, I think it's useful to be able to query say the list of hosts that have a particular disk manufacturer or things like that | 22:07 |
| farid | sure | 22:07 |
| thomasem | Oh, yep! | 22:07 |
| cloudnull | -qq https://github.com/openstack/craton/blob/master/etc/craton-api-conf.sample#L11 are we binding to 8080 as the default port ? | 22:07 |
| thomasem | cloudnull: yep | 22:08 |
| cloudnull | if so swift will not be happy. | 22:08 |
| cloudnull | https://github.com/openstack/swift/blob/master/etc/proxy-server.conf-sample#L3 | 22:08 |
| jimbaker | farid, yes for now; and more so once the var search takes in account resolution | 22:08 |
| cloudnull | is there a non-openstack conflicting port we indend to use ? | 22:08 |
| thomasem | cloudnull: Is this deployed alongside swift or something? | 22:08 |
| cloudnull | yea. i was trying to run in an osa cloud | 22:09 |
| cloudnull | which had swift | 22:09 |
| cloudnull | **has | 22:09 |
| jimbaker | farid, so right now, we can search for an entity if the var is set for that entity | 22:09 |
| jimbaker | eg in the gist thomasem posted, i can search for matching projects | 22:09 |
| thomasem | cloudnull: We can configure it however we want, furthermore if you're running in Docker, you can map however you like also. | 22:09 |
| jimbaker | with https://bugs.launchpad.net/craton/+bug/1661226, which is modestly difficult, but doable, anything under that project could be searched by that project var | 22:10 |
| openstack | Launchpad bug 1661226 in craton "Filter by variables lacks resolved variables" [Critical,Confirmed] - Assigned to Jim Baker (jimbaker) | 22:10 |
| cloudnull | essentially trying to get it running along side of all of the rest of the infra in a real env | 22:10 |
| cloudnull | sure. | 22:10 |
| jimbaker | farid, hope that helps! | 22:10 |
| cloudnull | just curious if we've thought about using a port that does not conflict with other openstack projects. | 22:10 |
| thomasem | cloudnull: Yeah, I mean, I'd change the default. 8080 is a pretty standard default port. | 22:10 |
| thomasem | By default? Probably not. But, I don't see a problem with doing so. | 22:10 |
| farid | it does thanks jimbaker | 22:11 |
| cloudnull | agreeded. just pointing out that as an openstack project we might want to ensure we're not conflicting with others. | 22:11 |
| cloudnull | 8080 may be fairly stock/standard. but its the default for the swift proxt | 22:12 |
| cloudnull | **proxy | 22:12 |
| antonym | probably should be a port registry for openstack projects heh | 22:13 |
| cloudnull | getting craton as part of https://docs.openstack.org/liberty/config-reference/content/firewalls-default-ports.html may be a good idea | 22:13 |
| antonym | hah | 22:13 |
| cloudnull | you're welcome :) | 22:14 |
| * cloudnull is a helper | 22:14 | |
| thomasem | cloudnull: gotcha. | 22:16 |
| thomasem | antonym: lol yep! | 22:16 |
| cloudnull | pretty https://docs.openstack.org/newton/config-reference/firewalls-default-ports.html | 22:16 |
| cloudnull | more up-to-date that is | 22:16 |
| thomasem | O_O so pretty | 22:16 |
| thomasem | :P | 22:17 |
| jimbaker | yeah, 8080 is one of those things that we assume, really you're going to change it, right? unless you are in docker and completely own your port space | 22:17 |
| cloudnull | adding craton somewhere there would likely be a good thing | 22:17 |
| cloudnull | jimbaker: not really. most folks will deploy with the defaults. like me. | 22:17 |
| cloudnull | and if they add craton to an openstack cloud they'll run into this issue | 22:18 |
| cloudnull | its easy to change | 22:18 |
| cloudnull | just pointing out the conflict | 22:18 |
| antonym | yeah, same, more collisions with 8080 | 22:18 |
| cloudnull | that is an "openstack cloud" as defined by defcore. so swift is present. | 22:19 |
| cloudnull | had i not had swift i would not be here typing this | 22:19 |
| 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 | 22:25 |
| jimbaker | uh oh, default passwords and all.... :) | 22:27 |
| jimbaker | yes, we can do a better deployment guide! | 22:27 |
| 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 | 22:27 |
| jimbaker | and agreed about adding our port in the openstack registry | 22:27 |
| cloudnull | are there default passwords within the app? | 22:28 |
| jimbaker | yes, 'craton' | 22:28 |
| cloudnull | :\ | 22:28 |
| jimbaker | but only used by docker startup | 22:28 |
| jimbaker | craton itself doesn't know about it | 22:29 |
| cloudnull | ok. | 22:29 |
| jimbaker | it's the responsibility of the deployer to do the right thing here when setting up mysql (or db of choice) | 22:29 |
| cloudnull | so the db-sync will use that as a default password ? | 22:30 |
| jimbaker | https://github.com/openstack/craton/blob/master/tools/docker_run.sh#L24 | 22:30 |
| cloudnull | ah , ok . only in the docker run | 22:30 |
| cloudnull | if I run craton in !docker i should not have to worry about that tomfoolery | 22:30 |
| jimbaker | correct | 22:31 |
| jimbaker | docker stuff is only for us devs. and possibly helpful as a working guide to deployment for others | 22:31 |
| 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 | 22:31 |
| thomasem | Yeah, I still haven't gotten to it, but I intend to write a compose setup and some initial documentation to iterate on for how we think this thing ought to be deployed. | 22:33 |
| jimbaker | right, that's a good longterm plan. something we can discuss with deployers to see what makes most sense | 22:34 |
| thomasem | Yeah. I'm really interested in the sort of feedback we're already getting from cloudnull regarding sane defaults for deployment | 22:35 |
| thomasem | Among other things that we'll likely get as we get some proper run time with this thing. | 22:36 |
| 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 | 22:38 |
| thomasem | I stripped out the peripheral refactoring in that change to focus on the issue at hand. | 22:39 |
| thomasem | It was getting in the way of focusing on the problems the patch is really trying to solve. | 22:39 |
| thomasem | Anywho, I'm burnt for the day. I'll catch y'all tomorrow. Have a lovely evening! | 22:42 |
| jimbaker | thomasem, thanks, i will take a look. do have a good evening, and some relaxation! | 22:52 |
| cloudnull | what is this about https://github.com/openstack/craton/blob/master/tools/docker_run.sh#L38-L46 | 23:06 |
| cloudnull | is this when keystone is not present ? | 23:06 |
| antonym | i think that's just for demoing stuff from the docs | 23:08 |
| cloudnull | so i should not need that irl | 23:08 |
| cloudnull | ? | 23:08 |
| antonym | https://github.com/openstack/craton/blob/master/doc/source/docker-install.rst | 23:08 |
| antonym | this doc references those | 23:08 |
| antonym | yeah, i'd imagine you wouldn't need it for a deployment | 23:09 |
| cloudnull | so the default docker file calls that | 23:09 |
| cloudnull | was just curious | 23:09 |
| antonym | i guess its to let you immediately play and test | 23:10 |
| antonym | did they change githubs colors or do i just need to take a breather | 23:11 |
| cloudnull | no they changed it | 23:11 |
| antonym | hah, no blog post or anything | 23:11 |
| antonym | it's like bugging my eyes | 23:12 |
| cloudnull | ok i think im online finally in an osa deployment. | 23:24 |
| cloudnull | ima write all of this up as a role/playbook | 23:25 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/craton master: Move craton to a non-conflicting https://review.openstack.org/439909 | 23:30 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/craton master: Move craton to a non-conflicting port https://review.openstack.org/439909 | 23:30 |
| cloudnull | sorry spam added tag | 23:30 |
| *** VW has quit IRC | 23:39 | |
| *** jovon has quit IRC | 23:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!