*** gouthamr has joined #openstack-shade | 01:14 | |
*** gouthamr has quit IRC | 01:41 | |
*** stevemar has quit IRC | 02:21 | |
*** stevemar has joined #openstack-shade | 02:21 | |
*** gouthamr has joined #openstack-shade | 04:23 | |
*** yfried has joined #openstack-shade | 05:16 | |
*** yfried has quit IRC | 05:21 | |
*** abregman has joined #openstack-shade | 05:38 | |
*** yfried has joined #openstack-shade | 06:30 | |
*** greghaynes has quit IRC | 06:52 | |
*** Shrews has quit IRC | 06:53 | |
*** mordred has quit IRC | 06:53 | |
*** greghaynes has joined #openstack-shade | 06:59 | |
*** mordred has joined #openstack-shade | 07:02 | |
*** greghaynes has quit IRC | 07:04 | |
*** greghaynes has joined #openstack-shade | 07:15 | |
*** Shrews has joined #openstack-shade | 07:17 | |
*** gouthamr has quit IRC | 07:32 | |
*** abregman has quit IRC | 08:26 | |
*** abregman has joined #openstack-shade | 08:29 | |
*** Matias has quit IRC | 08:33 | |
*** Matias has joined #openstack-shade | 08:40 | |
*** abregman has quit IRC | 09:07 | |
*** abregman has joined #openstack-shade | 09:42 | |
*** kong has joined #openstack-shade | 09:53 | |
kong | hi, guys, do we want to implement vpnaas api in shade, given it was removed from big tent? | 10:34 |
---|---|---|
*** cdent has joined #openstack-shade | 10:55 | |
*** yolanda has quit IRC | 11:36 | |
*** yolanda has joined #openstack-shade | 11:36 | |
*** abregman has quit IRC | 12:33 | |
frickler | get_nat_destination() is returning a completely wrong network for my cloud, is there a way to override this? I've also failed to locate where this function is defined so I can debug it. it breaks assigning floating IPs | 12:55 |
Shrews | frickler: you can set nat_destination: true in your clouds.yaml (http://docs.openstack.org/developer/os-client-config/network-config.html). though i suspect mordred will be more interested in why it's wrong for your cloud | 13:03 |
Shrews | kong: probably not | 13:04 |
frickler | Shrews: hmm, sounds strange that this should be a boolean instead of being able to specify the correct network. anyway I'm digging through the code now, turned out was looking at a pretty old git checkout that did not have shade/openstackcloud.py | 13:10 |
*** abregman has joined #openstack-shade | 13:30 | |
*** yfried has quit IRC | 13:36 | |
frickler | Shrews: ah, I needed to read that doc more closely. it is one attribute per network and I need to list all of them. so this gets me one step further, thanks | 14:04 |
mordred | kong: hrm. it's a good question - but I think we may want to wait a bit before adding support ... my biggest question is going to wind up being supportability | 15:20 |
mordred | kong: I don't want to add vpnaas support to shade then have a _different_ unrelated vpnaas project with a different api show up | 15:20 |
Shrews | ++ | 15:21 |
Shrews | plus, i don't hear folks clammering for vpnaas support in shade atm | 15:21 |
mordred | yah. I'd rather spend neutron time on replacing python-neutronclient with rest calls currently | 15:23 |
mordred | although given how minimal of a wrapper python-neutronclient is - I think that's going to be _super_easy | 15:24 |
Shrews | famous. last. words. | 15:24 |
mordred | Shrews: "what could possibly go wrong?" | 15:24 |
*** cdent_ has joined #openstack-shade | 15:31 | |
*** cdent has quit IRC | 15:31 | |
*** cdent_ is now known as cdent | 15:31 | |
*** cdent has quit IRC | 15:36 | |
*** cdent_ has joined #openstack-shade | 15:37 | |
*** cdent_ has quit IRC | 15:42 | |
*** gouthamr has joined #openstack-shade | 15:47 | |
*** cdent has joined #openstack-shade | 15:50 | |
*** cdent has quit IRC | 16:00 | |
*** cdent has joined #openstack-shade | 16:08 | |
*** abregman is now known as abregman|afk | 17:38 | |
morgan | mornin | 17:48 |
openstackgerrit | Roberto Polli proposed openstack-infra/shade: Basic volume_type access https://review.openstack.org/414561 | 18:52 |
*** yfried has joined #openstack-shade | 18:53 | |
mordred | mornin morgan | 18:53 |
mordred | Shrews: ^^ I think that patch from Roberto is ready now | 18:54 |
mordred | Shrews: also - over the weekend I learned about https://github.com/metacloud/molecule ... which is for testing ansible roles. sure does use shade :) | 18:54 |
morgan | hah metacloud stuff still eh | 18:55 |
Shrews | yay consumers | 18:56 |
Shrews | mordred: what had we decided about the search/list stuff? weren't we deprecating one over the other? | 19:00 |
Shrews | been so many beers ago... | 19:01 |
Shrews | https://review.openstack.org/395868 | 19:03 |
Shrews | mordred: so i think we meant to deprecate the confusing search* stuff and just use list*. that review adds both | 19:06 |
Shrews | or just have all search* as alias, is maybe the path we intended. | 19:07 |
Shrews | ugh, we never documented that, either. we suck | 19:08 |
mordred | Shrews: yah - I started working on collapsing them, and it turns out to be WAY harder than you'd like | 19:10 |
* morgan lurks. | 19:10 | |
* morgan lurks harder. | 19:10 | |
mordred | Shrews: we have a bunch of snowflakes | 19:10 |
Shrews | mordred: so are we abandoning that idea? | 19:11 |
mordred | Shrews: I dunno - I still like it as an idea - but I think it's 4 things down on the list of long slogging efforts | 19:11 |
mordred | Shrews: (less important than RESTification and finishing making sure everything is normalized properly, for instance) | 19:12 |
Shrews | APIs are hard | 19:13 |
Shrews | let's become professional couch sitters | 19:13 |
mordred | Shrews: SOLD | 19:18 |
morgan | Shrews: but... i already put a deposit in on the goat farm? I need to do a deposit on the couch next?! =/ | 19:54 |
morgan | ;) | 19:54 |
Shrews | morgan: -1'd Roberto's patch. Some pretty serious issues found in testing. Requested functional tests, too. | 19:59 |
Shrews | err, mordred ^^^^ | 19:59 |
Shrews | morgan: goats are so 1990's | 19:59 |
mordred | Shrews: thank you | 20:00 |
morgan | Shrews: mo<tab> trap! | 20:02 |
Shrews | more like mor<tab>. i usually do 3 characters before tab. usually enough, except for you two | 20:04 |
kong | Shrews, mordred, actually, i 'm asking becuase we have the requirement in our cloud. it's a pity it's not going to be supported in shade | 20:20 |
*** jamielennox|away is now known as jamielennox | 20:59 | |
*** abregman|afk has quit IRC | 20:59 | |
*** cdent has quit IRC | 21:22 | |
*** Matias has quit IRC | 21:24 | |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: WIP Replace SwiftService with direct REST uploads https://review.openstack.org/415739 | 21:38 |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Add support to task manager for async tasks https://review.openstack.org/414758 | 21:38 |
mordred | Shrews: that WIP is still WIP - but I added in 2 tests and cleaned up a smidge in prep for actually testing actually uploading things | 21:39 |
mordred | Shrews: (my favorite thing in the world is testing uploads of large things) | 21:39 |
mordred | jeblair: ^^ | 21:39 |
Shrews | mordred: isn't it everyone's favorite thing? | 21:39 |
mordred | Shrews: so much favorite | 21:39 |
Shrews | mordred: where are we on the async patch? | 21:40 |
mordred | Shrews: I think it's ready to rock and roll | 21:41 |
Shrews | can't tell from the review if jeblair's concerns were met. i'll just wait for him to re-review first :) | 21:41 |
mordred | Shrews: I believe I have handled his review comments - but I agree that's a good plan | 21:42 |
mordred | Shrews: what's amazing to me is that the FileSegment objects pretty much worked the first time I tested them | 21:43 |
mordred | Shrews: that doesn't seem right | 21:43 |
Shrews | i was going to say something is off | 21:43 |
Shrews | our tests must be broken | 21:43 |
Shrews | or non-existent | 21:43 |
jeblair | my irc client scrolled at just the right time for me to read "<Shrews> mordred: where are we on the rock and roll" | 21:44 |
mordred | jeblair: that's the question we should be asking ourselves constantly | 21:45 |
Shrews | \oo/ \oo/ | 21:45 |
Shrews | rock on | 21:45 |
clarkb | there are really good midi programming tools for haskell | 21:46 |
jeblair | new gerrit review category: "Rock-and-roll: +2 Rock on" | 21:46 |
jeblair | mordred, Shrews: lgtm | 21:48 |
mordred | jeblair: woot | 21:48 |
*** dfflanders has joined #openstack-shade | 21:52 | |
*** yfried has quit IRC | 22:02 | |
*** htaccess has joined #openstack-shade | 22:12 | |
morgan | jeblair: hehe | 22:34 |
mordred | jeblair, Shrews: woot. the concurrent.futures stuff in async works | 22:50 |
mordred | https://object-store.example.com/v1/1c36b64c840a42cd9e9b931a369337f0/shade.tests.unit.test_object.TestObjectUploads.test_create_dynamic_large_object-2/shade.tests.unit.test_object.TestObjectUploads.test_create_dynamic_large_object-3/000025 | 22:50 |
mordred | I have a unit test trying (and failing) to upload that ^^ | 22:50 |
mordred | (it's failing because I need to register a crapton of urls with requests_mock | 22:50 |
mordred | but figuring out the requests_mock registration is a great problem to have - it means the other stuff worked right and got us to the ugly part | 22:51 |
openstackgerrit | Merged openstack-infra/shade: Add support to task manager for async tasks https://review.openstack.org/414758 | 23:06 |
openstackgerrit | Monty Taylor proposed openstack-infra/shade: Replace SwiftService with direct REST uploads https://review.openstack.org/415739 | 23:17 |
mordred | jeblair, Shrews: ^^ ok. that patch is mildly absurd | 23:18 |
mordred | but it passes the fake unit tests I wrote, so it must be correct, right? | 23:18 |
mordred | morgan: ^^ | 23:18 |
* mordred goes to cry in the corner | 23:19 | |
morgan | ok i'm looking | 23:28 |
mordred | morgan: enjoy the crazy | 23:42 |
morgan | mordred: ouch | 23:52 |
morgan | brain hurty | 23:52 |
mordred | morgan: yah | 23:53 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!