| *** Dusty has joined #craton | 00:08 | |
| *** jovon has quit IRC | 00:37 | |
| *** Dusty has quit IRC | 00:49 | |
| *** Dusty has joined #craton | 01:26 | |
| *** Dusty1 has joined #craton | 01:29 | |
| *** Dusty has quit IRC | 01:30 | |
| *** Dusty1 has quit IRC | 01:47 | |
| *** Dusty has joined #craton | 02:16 | |
| *** syed__ has quit IRC | 02:25 | |
| *** rainya has joined #craton | 02:57 | |
| *** rainya has quit IRC | 03:02 | |
| *** rainya has joined #craton | 03:03 | |
| *** rainya has quit IRC | 03:10 | |
| *** mhayden has quit IRC | 03:12 | |
| *** mhayden has joined #craton | 03:13 | |
| *** rainya has joined #craton | 03:16 | |
| *** Dusty has quit IRC | 03:23 | |
| *** rainya has quit IRC | 03:32 | |
| *** Dusty has joined #craton | 03:32 | |
| *** valw has joined #craton | 03:36 | |
| *** Dusty has joined #craton | 03:41 | |
| *** valw_ has joined #craton | 03:45 | |
| *** valw has quit IRC | 03:47 | |
| *** Dusty has joined #craton | 03:47 | |
| *** Dusty has joined #craton | 03:49 | |
| *** Dusty has joined #craton | 03:53 | |
| *** Dusty has joined #craton | 03:56 | |
| *** Dusty has joined #craton | 03:59 | |
| *** Dusty has quit IRC | 03:59 | |
| *** Dusty has joined #craton | 04:01 | |
| *** Dusty has quit IRC | 04:01 | |
| *** Dusty has joined #craton | 04:02 | |
| *** Dusty has joined #craton | 04:06 | |
| *** Dusty has joined #craton | 04:08 | |
| *** Dusty has joined #craton | 04:09 | |
| *** Dusty has quit IRC | 04:09 | |
| *** Dusty has joined #craton | 04:10 | |
| *** Dusty has quit IRC | 04:10 | |
| *** Dusty has joined #craton | 04:12 | |
| *** Dusty has joined #craton | 04:13 | |
| *** Dusty has joined #craton | 04:14 | |
| *** valw_ has quit IRC | 04:40 | |
| *** syed__ has joined #craton | 05:43 | |
| *** tojuvone has joined #craton | 06:21 | |
| *** tojuvone_ has joined #craton | 06:21 | |
| *** tojuvone_ has quit IRC | 06:21 | |
| *** Mudpuppy has quit IRC | 06:37 | |
| *** Mudpuppy has joined #craton | 07:08 | |
| *** Mudpuppy has quit IRC | 07:13 | |
| *** syed__ has quit IRC | 07:55 | |
| sulo | jimbaker: yeah i think we can do what we are doing .... i will need to spend time on the image a bit .. and bring the time down to low under 3 mins | 09:26 |
|---|---|---|
| sulo | so lets not worry about tempest or other stuff just yet | 09:26 |
| sulo | i am going to get the base setup done first .. then work on the image later | 09:27 |
| *** git-harry has quit IRC | 09:31 | |
| *** git-harry has joined #craton | 09:31 | |
| *** klindgren has quit IRC | 09:58 | |
| *** klindgren has joined #craton | 09:59 | |
| *** tojuvone has quit IRC | 10:54 | |
| *** VW has joined #craton | 12:41 | |
| *** VW has quit IRC | 12:57 | |
| *** VW has joined #craton | 12:57 | |
| *** VW has quit IRC | 13:04 | |
| *** VW has joined #craton | 13:04 | |
| *** VW has quit IRC | 13:18 | |
| sulo | soo... | 13:22 |
| sulo | for these functional tests .. i would like to use python-cratonclien ;) | 13:22 |
| sulo | is it doable right now ? I havent looked at it recently to know .. | 13:23 |
| sulo | actually looks doable .. atlest to start | 13:29 |
| *** tojuvone has joined #craton | 13:34 | |
| sigmavirus | sulo: totes | 13:36 |
| *** Dusty has joined #craton | 13:42 | |
| sulo | sigmavirus: the only thing is inventory is still per region though | 13:47 |
| sulo | rather than per project | 13:47 |
| sigmavirus | sulo: well it's both | 13:47 |
| sulo | sigmavirus: you are changing that right ? | 13:47 |
| sulo | both ? | 13:47 |
| sigmavirus | but an Inventory object is per-region | 13:48 |
| sulo | ah | 13:48 |
| sigmavirus | well you can't access another project's information right now | 13:48 |
| sigmavirus | that will only strengthen when we add in RBAC | 13:48 |
| sulo | sigmavirus: so that wasnt my concern, although thats very important | 13:49 |
| sulo | the way it is right now | 13:49 |
| sulo | regions has different object model than the rest .. i can do inventory.cells | 13:49 |
| sigmavirus | correct | 13:49 |
| sulo | but region get all for example would be different ? | 13:49 |
| sigmavirus | inventory.cells.list | 13:49 |
| sigmavirus | right, you'd create your Client(..) | 13:50 |
| sigmavirus | then do client.regions.list() | 13:50 |
| *** Mudpuppy has joined #craton | 13:50 | |
| sigmavirus | Honestly, I'm still not sure why jimbaker and chrisspencer wanted to not have cells/hosts/etc. hang off a region object but *shrug* | 13:50 |
| sulo | right .. thats my point .. if we have inventory | 13:50 |
| sulo | then inventory should work for all | 13:50 |
| sulo | yeah that makes no sense at all | 13:50 |
| sigmavirus | sulo: what makes no sense at all? | 13:51 |
| sulo | having inventory exclude region | 13:51 |
| sulo | because its not the top object in inventory | 13:51 |
| sulo | project is | 13:51 |
| sulo | anyway .. i thought you were changing that to have project as the top level item | 13:52 |
| sigmavirus | sulo: so something like project = client.projects.get('uuid'), project.regions.list(), project.hosts.list(region_id=1)? | 13:53 |
| sigmavirus | or something else? | 13:53 |
| sulo | yeah if we are doing that .. then that works because it gives a consistent way to access resources | 13:53 |
| sigmavirus | sulo: the thing is, you can't specify a particular project_id any longer for requests, we removed that | 13:54 |
| sigmavirus | so you can't access the regions/hosts for a project_id that you aren't authenticating with iirc | 13:54 |
| sulo | right, all i am saying is .. i dont understand why we have the inventory object | 13:55 |
| sigmavirus | sulo: I've been saying that for a while :) | 13:55 |
| sigmavirus | right now, caching seems to be a priority so I'm working on that first | 13:55 |
| sulo | gotcha | 13:55 |
| sigmavirus | I seemed to be the only person who disliked the current API design | 13:57 |
| sigmavirus | I think the sole purpose of the inventory object is to prevent you having to constantly specify the region id | 13:57 |
| sigmavirus | but that seems like an ergonomic consideration for the higher level API that jimbaker wants | 13:58 |
| sulo | ok, i dont see the benifit, but there must be a reason that chrisspencer and jimbaker wanted that | 14:00 |
| sulo | we can discuss that later today | 14:00 |
| sigmavirus | yep | 14:01 |
| *** valw has joined #craton | 14:41 | |
| *** valw has quit IRC | 14:43 | |
| *** valw has joined #craton | 14:43 | |
| *** valw has quit IRC | 15:04 | |
| *** valw has joined #craton | 15:07 | |
| *** syed__ has joined #craton | 15:10 | |
| jimbaker | sulo, sigmavirus - sounds good about finalizing that api. i'm confused at this point ;), hopefully today's meeting will clear things up | 15:25 |
| jimbaker | per our earlier discussion, we are having today's call at 1700 UTC. it's been current in the topic for a while here, but finally revised outlook. my earlier attempt, probably because i used to have my calendar in UTC | 15:25 |
| jimbaker | and outlook just pretends to know about timezones | 15:26 |
| jimbaker | my earlier attempt *failed* to update | 15:26 |
| jimbaker | so delete, insert vs update :) | 15:27 |
| jimbaker | sulo, re functional testing - yes do use the cratonclient. i think this should be the point of any such testing | 15:27 |
| jimbaker | btw, here's one functional test that i have been running as a smoke test https://gist.github.com/jimbaker/df47cabfefa6d6985af9029449b589df | 15:30 |
| jimbaker | "functional test" | 15:30 |
| jimbaker | but my assumption has been: we have the fixture data from fake data; and we are able to interact with it, then make assertions, using the craton client. simple enough | 15:31 |
| *** valw has quit IRC | 15:42 | |
| *** valw_ has joined #craton | 15:43 | |
| *** valw_ has quit IRC | 15:50 | |
| *** valw has joined #craton | 15:54 | |
| *** valw_ has joined #craton | 15:55 | |
| *** klindgren_ has joined #craton | 15:56 | |
| *** klindgren has quit IRC | 15:58 | |
| *** valw_ has quit IRC | 15:58 | |
| *** valw has quit IRC | 15:59 | |
| syed__ | Hi guys | 15:59 |
| syed__ | what time will be the meeting today ? | 15:59 |
| syed__ | 11 ? | 15:59 |
| *** valw has joined #craton | 16:01 | |
| *** valw has quit IRC | 16:02 | |
| *** valw has joined #craton | 16:04 | |
| *** jovon has joined #craton | 16:09 | |
| *** Mudpuppy_ has joined #craton | 16:23 | |
| *** Mudpuppy has quit IRC | 16:26 | |
| sigmavirus | syed__: 11 Central? Yes | 16:29 |
| *** valw has quit IRC | 16:38 | |
| *** valw has joined #craton | 16:40 | |
| *** valw has quit IRC | 16:42 | |
| *** valw has joined #craton | 16:42 | |
| *** valw has quit IRC | 16:43 | |
| *** valw has joined #craton | 16:44 | |
| jimbaker | syed__, 1700 UTC for tues/thurs. i like https://www.timeanddate.com/ to put this into a specific timezone (for you 11:00a CT), but this doesn't work so well as we scale out the team in different locales | 16:50 |
| jimbaker | hence using UTC for all times | 16:51 |
| *** Mudpuppy_ is now known as Mudpuppy | 16:54 | |
| sigmavirus | client caching is a turning out to be a rabbit hole xD | 16:56 |
| jimbaker | sigmavirus, oh. well let's really discuss that momentarily | 16:57 |
| sigmavirus | jimbaker: so much to talk about :P | 16:57 |
| jimbaker | yeah, it's good | 16:57 |
| jimbaker | sulo, client/ecosystem meeting on vidyo? | 17:01 |
| sigmavirus | sulo: http://specs.openstack.org/openstack/openstack-specs/specs/cors-support.html will be necessary for Horizon | 17:22 |
| *** valw has quit IRC | 17:33 | |
| *** valw has joined #craton | 17:43 | |
| *** rainya has joined #craton | 17:44 | |
| sigmavirus | sulo: for doing listings, we could probably just checkum the resource table | 18:01 |
| jovon | hey sigmavirus | 18:07 |
| sigmavirus | hey jovon | 18:07 |
| jovon | since you're working on API's would you you mind if "look over your shoulder" so i can get a better understanding of the process | 18:08 |
| sigmavirus | jovon: you mean working on the client or the craton API itself or ... ? | 18:08 |
| jovon | i was referring to craton API but learning both could never hurt if you don't mind | 18:09 |
| jovon | I really want to understand your methodology | 18:10 |
| jovon | to see how an expert would do it. am i making sense? | 18:10 |
| sigmavirus | I don't mind at all | 18:11 |
| sigmavirus | Yep you are :) | 18:11 |
| jovon | awesome! :) thank you | 18:12 |
| sulo | sigmavirus: jimbaker: git-harry: https://review.openstack.org/#/c/397872 its far from ready .. but wanted to pass this to see how y'all feel about the general direction of this patch | 18:14 |
| sulo | jovon: syed__^ | 18:14 |
| sulo | ideas/reviews are welcome ;) | 18:15 |
| sigmavirus | sulo: so, for what it's worth, "tearDownClass" (really tearDown*) are not guaranteed to run so you want to use addCleanup | 18:16 |
| sulo | sigmavirus: thanks, will do | 18:16 |
| sigmavirus | wait a second, addCleanup might only be to correspond to setUp | 18:17 |
| sigmavirus | not setUpClass | 18:17 |
| sulo | yeah .. i think we need on module level actually | 18:17 |
| sulo | even setupcalss is prolly not good enough | 18:18 |
| sulo | so ill change taht | 18:18 |
| sigmavirus | nevermind there's nothing similar | 18:18 |
| sigmavirus | http://specs.openstack.org/openstack/openstack-specs/specs/cors-support.html | 18:18 |
| sigmavirus | gah wrong link | 18:18 |
| sigmavirus | https://docs.python.org/3/library/unittest.html#class-and-module-fixtures | 18:18 |
| sigmavirus | Specifically: "Note that shared fixtures do not play well with [potential] features like test parallelization and they break test isolation. They should be used with care." | 18:19 |
| sigmavirus | and testr does do parallelization | 18:19 |
| sulo | i was going to switch to testr | 18:20 |
| sulo | :| | 18:20 |
| sigmavirus | you should switch to it :/ | 18:20 |
| jovon | Sulo will take a look after a quick lunch | 18:20 |
| sulo | sigmavirus: hehe, yeah | 18:21 |
| sulo | jovon: thanks | 18:21 |
| sigmavirus | sulo: jimbaker https://review.openstack.org/397875 | 18:22 |
| jimbaker | sulo, thanks i will take a look at that wip patch re functional testing | 18:23 |
| jimbaker | ian, thanks for the ci patch, looks good | 18:24 |
| *** Dusty has quit IRC | 18:30 | |
| *** harlowja has quit IRC | 18:38 | |
| *** Dusty has joined #craton | 18:39 | |
| *** harlowja has joined #craton | 19:06 | |
| *** valw has quit IRC | 19:13 | |
| *** valw has joined #craton | 19:26 | |
| *** valw has quit IRC | 19:31 | |
| *** rainya has quit IRC | 19:33 | |
| *** valw has joined #craton | 19:50 | |
| *** valw_ has joined #craton | 20:00 | |
| *** valw_ has quit IRC | 20:02 | |
| *** valw__ has joined #craton | 20:02 | |
| *** valw has quit IRC | 20:03 | |
| *** Dusty1 has joined #craton | 20:03 | |
| *** Dusty has quit IRC | 20:04 | |
| sigmavirus | jimbaker: sulo so, in our functional tests, we can configure MySQL to log any queries that don't use an index and analyze them for full table scans (if we want to be paranoid) | 20:08 |
| jimbaker | sigmavirus, +1, great idea | 20:08 |
| sigmavirus | I worked on a project as a consultant that did this in their tests | 20:08 |
| sigmavirus | I didn't set it up though, so we should probably put that off for a while | 20:08 |
| sigmavirus | jimbaker: better yet | 20:13 |
| sigmavirus | there are mysql status counters for stuff like this | 20:13 |
| sigmavirus | if we just restart mysql between tests then it'll reset | 20:13 |
| jimbaker | sigmavirus, so collect at end as a report | 20:13 |
| sigmavirus | "Select_scan" will count how many full table scans there are and "Select_full_join" will count full table scans against joined tables which have even worse performance | 20:14 |
| jimbaker | makes sense. i fully expect we may have some queries (rare) that should be table scans | 20:14 |
| jimbaker | but if we see variance, that's what we care about | 20:14 |
| *** valw__ has quit IRC | 20:22 | |
| *** valw has joined #craton | 20:24 | |
| sigmavirus | jimbaker: sure, those we exclude from this kind of chec | 20:38 |
| sigmavirus | *check | 20:38 |
| jimbaker | yeah, we can just reset counters around our expectations | 20:38 |
| sigmavirus | yarp | 20:41 |
| jimbaker | right, using flush status, http://dev.mysql.com/doc/refman/5.7/en/flush.html | 20:41 |
| jimbaker | yeah, so this should just work | 20:42 |
| *** valw has quit IRC | 21:00 | |
| *** valw has joined #craton | 21:01 | |
| *** Dusty1 has quit IRC | 21:03 | |
| *** Dusty has joined #craton | 21:07 | |
| sigmavirus | https://review.openstack.org/397957 fwiw | 21:28 |
| sigmavirus | later y'all | 21:28 |
| *** valw has quit IRC | 21:30 | |
| *** valw has joined #craton | 21:32 | |
| *** valw has quit IRC | 21:51 | |
| *** harlowja has quit IRC | 21:55 | |
| *** jovon has quit IRC | 21:57 | |
| *** valw has joined #craton | 22:01 | |
| *** valw has quit IRC | 22:03 | |
| *** valw has joined #craton | 22:03 | |
| *** harlowja has joined #craton | 22:13 | |
| *** Dusty has quit IRC | 22:14 | |
| *** Dusty has joined #craton | 22:28 | |
| *** Dusty has quit IRC | 22:28 | |
| *** Dusty has joined #craton | 22:30 | |
| *** valw has quit IRC | 22:42 | |
| *** valw has joined #craton | 22:43 | |
| *** valw has quit IRC | 22:45 | |
| *** valw has joined #craton | 22:45 | |
| *** Dusty has joined #craton | 23:00 | |
| *** valw has quit IRC | 23:05 | |
| *** valw has joined #craton | 23:07 | |
| jimbaker | Dusty, sigmavirus, sulo, so i will get in a submission in for rax.io on craton, given our multicloud discussion at today's meeting. deadline is nov 30 | 23:09 |
| jimbaker | Dusty, i expect to also bring this up with the platform architecture team this friday. one of my peers on the group is working on supporting FAWS, so this will give me further insight into technical reqs | 23:10 |
| jimbaker | although i expect it's going to be mostly, how do we model AWS on craton, vs really much in the way of extension needed for craton. so anything we are collecting can be captured as variables; but the real interesting thing is when we look at aggregating aws service info using our virtualized variable model; and what audit/remediation looks like | 23:12 |
| *** valw has quit IRC | 23:15 | |
| *** valw has joined #craton | 23:16 | |
| *** valw has quit IRC | 23:19 | |
| *** Dusty has quit IRC | 23:25 | |
| *** valw has joined #craton | 23:30 | |
| *** valw has quit IRC | 23:34 | |
| *** Mudpuppy has quit IRC | 23:41 | |
| *** valw has joined #craton | 23:49 | |
| *** valw has quit IRC | 23:50 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!