| *** peristeri has quit IRC | 00:00 | |
| *** tsekiyama has quit IRC | 00:18 | |
| *** david-lyle has quit IRC | 00:32 | |
| *** spzala has joined #openstack-glance | 01:20 | |
| *** david-lyle has joined #openstack-glance | 01:20 | |
| *** annashen has joined #openstack-glance | 01:39 | |
| *** annashen has quit IRC | 01:42 | |
| *** dims_ is now known as dims | 01:52 | |
| *** harlowja is now known as harlowja_away | 02:03 | |
| *** david-lyle has quit IRC | 02:30 | |
| *** david-lyle has joined #openstack-glance | 02:31 | |
| *** ayoung has joined #openstack-glance | 02:36 | |
| *** david-lyle has quit IRC | 02:36 | |
| openstackgerrit | Yongfeng Du proposed openstack/glance-specs: Clarify the "add" operation behavior https://review.openstack.org/178491 | 03:11 |
|---|---|---|
| *** david-lyle has joined #openstack-glance | 03:32 | |
| *** changbl has joined #openstack-glance | 03:33 | |
| *** wayne__ has quit IRC | 03:43 | |
| *** sudipto has joined #openstack-glance | 03:43 | |
| *** sudipto has quit IRC | 05:04 | |
| *** nlevinki has joined #openstack-glance | 05:25 | |
| *** ajayaa has joined #openstack-glance | 05:38 | |
| *** annashen has joined #openstack-glance | 05:45 | |
| *** abhishekk has joined #openstack-glance | 05:48 | |
| *** tshefi has joined #openstack-glance | 05:56 | |
| *** annashen has quit IRC | 05:57 | |
| *** sgotliv has quit IRC | 06:01 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/glance: Imported Translations from Transifex https://review.openstack.org/176633 | 06:01 |
| *** Longgeek has joined #openstack-glance | 06:03 | |
| *** sudipto has joined #openstack-glance | 06:06 | |
| *** Longgeek has quit IRC | 06:08 | |
| *** pcaruana has joined #openstack-glance | 06:09 | |
| *** groen692 has joined #openstack-glance | 06:10 | |
| *** zul has joined #openstack-glance | 06:17 | |
| *** pkoniszewski has joined #openstack-glance | 06:19 | |
| *** zul has quit IRC | 06:24 | |
| *** zul has joined #openstack-glance | 06:25 | |
| *** sudipto has quit IRC | 06:45 | |
| *** sudipto has joined #openstack-glance | 06:50 | |
| *** Longgeek has joined #openstack-glance | 06:56 | |
| *** Longgeek has quit IRC | 07:02 | |
| *** Longgeek has joined #openstack-glance | 07:06 | |
| *** aarefiev has quit IRC | 07:24 | |
| *** bfic has joined #openstack-glance | 07:25 | |
| *** zul has quit IRC | 07:32 | |
| *** zul has joined #openstack-glance | 07:32 | |
| *** sgotliv has joined #openstack-glance | 07:34 | |
| *** aarefiev has joined #openstack-glance | 07:40 | |
| *** sgotliv has quit IRC | 07:57 | |
| *** jistr has joined #openstack-glance | 07:58 | |
| *** Longgeek has quit IRC | 08:05 | |
| *** zul has quit IRC | 08:09 | |
| *** dims has quit IRC | 08:10 | |
| *** zul has joined #openstack-glance | 08:25 | |
| openstackgerrit | Kamil Rykowski proposed openstack/python-glanceclient: Extend unittests coverage for v2 tasks module https://review.openstack.org/165436 | 08:29 |
| openstackgerrit | Kamil Rykowski proposed openstack/python-glanceclient: Use assertIn instead of assertTrue in tests https://review.openstack.org/165379 | 08:30 |
| *** sgotliv has joined #openstack-glance | 08:53 | |
| *** bfic has quit IRC | 08:54 | |
| *** annashen has joined #openstack-glance | 08:57 | |
| *** MattMan has quit IRC | 09:00 | |
| *** MattMan has joined #openstack-glance | 09:01 | |
| *** annashen has quit IRC | 09:02 | |
| *** bfic has joined #openstack-glance | 09:04 | |
| *** bfic has quit IRC | 09:04 | |
| *** e0ne has joined #openstack-glance | 09:06 | |
| *** Longgeek has joined #openstack-glance | 09:06 | |
| *** e0ne is now known as e0ne_ | 09:06 | |
| *** Longgeek has quit IRC | 09:12 | |
| *** Longgeek has joined #openstack-glance | 09:13 | |
| *** e0ne_ is now known as e0ne | 09:14 | |
| *** zul has quit IRC | 09:27 | |
| *** krykowski has joined #openstack-glance | 09:35 | |
| *** sudipto has quit IRC | 09:36 | |
| openstackgerrit | Kamil Rykowski proposed openstack/glance: Image member name validation https://review.openstack.org/175415 | 09:43 |
| *** aswadr has joined #openstack-glance | 09:44 | |
| *** zul has joined #openstack-glance | 09:51 | |
| *** dims has joined #openstack-glance | 09:51 | |
| *** sudipto has joined #openstack-glance | 09:52 | |
| *** sgotliv has quit IRC | 09:55 | |
| *** ajayaa has quit IRC | 09:57 | |
| *** bfic has joined #openstack-glance | 10:02 | |
| *** bfic has quit IRC | 10:02 | |
| *** e0ne is now known as e0ne_ | 10:03 | |
| *** e0ne_ is now known as e0ne | 10:05 | |
| *** mtreinish has quit IRC | 10:07 | |
| *** ajayaa has joined #openstack-glance | 10:10 | |
| *** sgotliv has joined #openstack-glance | 10:13 | |
| *** zul has quit IRC | 10:19 | |
| *** arif-ali has quit IRC | 10:20 | |
| *** bfic has joined #openstack-glance | 10:56 | |
| *** bfic has quit IRC | 10:57 | |
| *** bfic has joined #openstack-glance | 10:58 | |
| *** krykowski has quit IRC | 11:03 | |
| *** krykowski_ has joined #openstack-glance | 11:04 | |
| *** dims has quit IRC | 11:18 | |
| *** sudipto has quit IRC | 11:20 | |
| *** e0ne is now known as e0ne_ | 11:30 | |
| *** ajayaa has quit IRC | 11:30 | |
| *** aix has joined #openstack-glance | 11:35 | |
| *** sudipto has joined #openstack-glance | 11:35 | |
| *** krykowski_ has quit IRC | 11:38 | |
| *** bfic has quit IRC | 11:40 | |
| *** arif-ali has joined #openstack-glance | 11:40 | |
| *** cdelatte has joined #openstack-glance | 11:42 | |
| *** delattec has joined #openstack-glance | 11:42 | |
| *** ajayaa has joined #openstack-glance | 11:42 | |
| *** delatte has joined #openstack-glance | 11:44 | |
| *** krykowski has joined #openstack-glance | 11:45 | |
| *** cdelatte has quit IRC | 11:47 | |
| *** delattec has quit IRC | 11:47 | |
| *** cdelatte has joined #openstack-glance | 11:47 | |
| *** delatte has quit IRC | 11:52 | |
| *** cdelatte has quit IRC | 11:52 | |
| *** haomaiwa_ has quit IRC | 11:52 | |
| *** zul has joined #openstack-glance | 11:53 | |
| *** dims has joined #openstack-glance | 11:56 | |
| *** bfic has joined #openstack-glance | 12:01 | |
| *** krykowski has quit IRC | 12:03 | |
| *** krykowski has joined #openstack-glance | 12:03 | |
| *** openstackgerrit has quit IRC | 12:07 | |
| *** openstackgerrit has joined #openstack-glance | 12:07 | |
| *** krykowski has quit IRC | 12:08 | |
| openstackgerrit | Merged openstack/glance_store: Drop use of 'oslo' namespace package https://review.openstack.org/178336 | 12:10 |
| *** sudipto has quit IRC | 12:10 | |
| *** bfic has quit IRC | 12:26 | |
| *** e0ne_ is now known as e0ne | 12:26 | |
| *** EmilienM is now known as EmilienM|afk | 12:41 | |
| *** ajayaa has quit IRC | 12:41 | |
| *** raildo has joined #openstack-glance | 12:41 | |
| *** lifeless has quit IRC | 12:43 | |
| *** rodrigods has joined #openstack-glance | 12:43 | |
| *** chlong has joined #openstack-glance | 12:44 | |
| *** krykowski has joined #openstack-glance | 12:45 | |
| *** krykowski has quit IRC | 12:52 | |
| *** e0ne is now known as e0ne_ | 12:54 | |
| *** bfic has joined #openstack-glance | 12:59 | |
| *** pdb_ has quit IRC | 13:00 | |
| *** pdb_ has joined #openstack-glance | 13:00 | |
| *** annegentle has joined #openstack-glance | 13:03 | |
| *** julim has joined #openstack-glance | 13:05 | |
| *** spzala has quit IRC | 13:05 | |
| *** krykowski has joined #openstack-glance | 13:06 | |
| *** e0ne_ is now known as e0ne | 13:06 | |
| *** julim has quit IRC | 13:06 | |
| *** julim has joined #openstack-glance | 13:08 | |
| openstackgerrit | Cindy Pallares proposed openstack/glance-specs: Complete refactor to glance-store's API https://review.openstack.org/126550 | 13:13 |
| *** julim has quit IRC | 13:18 | |
| *** peristeri has joined #openstack-glance | 13:25 | |
| *** zeddii has quit IRC | 13:27 | |
| *** ajayaa has joined #openstack-glance | 13:34 | |
| *** zeddii has joined #openstack-glance | 13:36 | |
| *** EmilienM|afk is now known as EmilienM | 13:37 | |
| *** stpierre has joined #openstack-glance | 13:37 | |
| *** e0ne is now known as e0ne_ | 13:38 | |
| *** lifeless has joined #openstack-glance | 13:38 | |
| *** e0ne_ is now known as e0ne | 13:39 | |
| *** abhishekk has quit IRC | 13:42 | |
| *** mtreinish has joined #openstack-glance | 13:54 | |
| *** zz_jgrimm is now known as jgrimm | 14:01 | |
| *** mtreinish has quit IRC | 14:05 | |
| *** bfic has quit IRC | 14:06 | |
| *** ericpeterson has joined #openstack-glance | 14:08 | |
| *** cbits has joined #openstack-glance | 14:08 | |
| *** cbits has left #openstack-glance | 14:09 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 14:09 | |
| *** haomaiwang has joined #openstack-glance | 14:10 | |
| *** pkoniszewski has quit IRC | 14:11 | |
| *** annegentle has quit IRC | 14:12 | |
| *** annegentle has joined #openstack-glance | 14:17 | |
| *** tshefi has quit IRC | 14:21 | |
| *** asrangne has joined #openstack-glance | 14:26 | |
| *** aswadr has quit IRC | 14:28 | |
| *** julim has joined #openstack-glance | 14:31 | |
| openstackgerrit | Inessa Vasilevskaya proposed openstack/glance: Fix HTTP 500 on NotAuthenticated in registry https://review.openstack.org/178697 | 14:34 |
| *** dims has quit IRC | 14:42 | |
| *** bkopilov has quit IRC | 14:42 | |
| *** dims has joined #openstack-glance | 14:42 | |
| *** vijendar has joined #openstack-glance | 14:42 | |
| *** mtreinish has joined #openstack-glance | 14:45 | |
| *** bkopilov has joined #openstack-glance | 14:45 | |
| *** marcusvrn_ has joined #openstack-glance | 14:49 | |
| *** mtreinish has quit IRC | 14:49 | |
| *** nlevinki has quit IRC | 14:50 | |
| *** ajayaa has quit IRC | 14:52 | |
| *** pkoniszewski has joined #openstack-glance | 14:53 | |
| *** peristeri has quit IRC | 15:04 | |
| *** ajayaa has joined #openstack-glance | 15:05 | |
| *** peristeri has joined #openstack-glance | 15:33 | |
| *** julim has quit IRC | 15:43 | |
| *** mtreinish has joined #openstack-glance | 15:45 | |
| *** groen692 has quit IRC | 15:51 | |
| *** annegentle has quit IRC | 15:53 | |
| *** asrangne has quit IRC | 15:53 | |
| *** sudipto has joined #openstack-glance | 16:11 | |
| *** jistr has quit IRC | 16:11 | |
| *** zul has quit IRC | 16:15 | |
| *** julim has joined #openstack-glance | 16:17 | |
| *** chlong has quit IRC | 16:19 | |
| *** pcaruana has quit IRC | 16:22 | |
| *** sudipto has quit IRC | 16:44 | |
| *** annegentle has joined #openstack-glance | 16:54 | |
| *** marcusvrn_ is now known as marcusvrn | 16:55 | |
| *** annegentle has quit IRC | 17:01 | |
| *** pkoniszewski has quit IRC | 17:01 | |
| *** SpamapS has joined #openstack-glance | 17:01 | |
| SpamapS | Hello everybody. I'm poking at some glance v2 stuff and wondering if anybody can point me to where glance v2 is tested in the gate. | 17:02 |
| *** annegentle has joined #openstack-glance | 17:03 | |
| *** greghaynes has joined #openstack-glance | 17:05 | |
| *** aix has quit IRC | 17:10 | |
| *** sgotliv has quit IRC | 17:13 | |
| *** marcusvrn has quit IRC | 17:19 | |
| SpamapS | mtreinish: so I'm looking at tempests's v2 API tests and I'm a bit confused as to how this differs from the way RAX's glance v2 works.. | 17:23 |
| SpamapS | mtreinish: http://git.openstack.org/cgit/openstack/tempest/tree/tempest/services/image/v2/json/image_client.py#n136 | 17:23 |
| SpamapS | mtreinish: It's my understanding RAX forbids that upload | 17:24 |
| greghaynes | SpamapS: I have a rax test setup I can give you some specifics | 17:24 |
| * greghaynes makes an image | 17:24 | |
| mtreinish | SpamapS: that could be, I've never tried running it against rax | 17:24 |
| SpamapS | greghaynes: crazy idea, run just those tempest tests against rax? | 17:24 |
| SpamapS | the task code is in trunk, so we should probably have tests for it if we're not going to kick it out. | 17:25 |
| greghaynes | hah, I dont have any kind of tempest setup, not sure how much work that is to get going? | 17:25 |
| SpamapS | greghaynes: you _might_ be able to convince mtreinish to help you get one up quickly. :) | 17:25 |
| mtreinish | greghaynes: http://docs.openstack.org/developer/tempest/configuration.html#tempest-configuration is a start | 17:25 |
| SpamapS | faster than I can get my RAX account activated surely. :) | 17:25 |
| greghaynes | yes, that requires a phone call | 17:25 |
| SpamapS | phone? | 17:25 |
| SpamapS | whats that? | 17:25 |
| mtreinish | but you don't need to do much to setup just running the glance api tests serially | 17:25 |
| * SpamapS is just a caveman | 17:25 | |
| greghaynes | SpamapS: exactly | 17:25 |
| greghaynes | mtreinish: awesome | 17:26 |
| mtreinish | greghaynes, SpamapS: I'd do it myself but my free credit has like a week left on it | 17:26 |
| SpamapS | noooooo | 17:26 |
| SpamapS | should be infinite for openstack-devs ;) | 17:26 |
| mtreinish | SpamapS: yes, I agree fully :) | 17:26 |
| greghaynes | I still have to battle with getting the rax expense report to go through :( | 17:27 |
| mtreinish | greghaynes: to run it you should just need to put the keystone endpoint, a set of credentials to run the tests with, and probably set the catalog name for glance (because I doubt they use the default tempest value) | 17:28 |
| mtreinish | oh, and probably a random http address to use as an "image" for the tests | 17:29 |
| SpamapS | greghaynes: while you're at it, if you can try and run the tests from here against rax, I can help you too: https://review.openstack.org/178774 | 17:32 |
| greghaynes | ok, ill mess with that now | 17:32 |
| greghaynes | heres what I get from glanceclient http://paste.ubuntu.com/10942747/ | 17:32 |
| SpamapS | greghaynes: 'tox -efunctional test_image' with OS_xxx set properly should do it | 17:32 |
| SpamapS | greghaynes: v1 should obviously fail... :) | 17:32 |
| mtreinish | greghaynes: heh, so they're rate limiting you? | 17:32 |
| greghaynes | hrm, I wonder if that X-Auth-Token is a problem in a paste ; | 17:32 |
| greghaynes | mtreinish: they say that | 17:33 |
| SpamapS | quota for images? | 17:33 |
| SpamapS | greghaynes: yeah you should invalidate that token | 17:33 |
| greghaynes | I think its not actually a rate limit though | 17:33 |
| greghaynes | they say that persistently | 17:33 |
| greghaynes | which might be how they "disable" support for that, setting a rate of 0 somewhere | 17:34 |
| mtreinish | greghaynes: heh, that makes it super clear :) | 17:34 |
| SpamapS | greghaynes: you may have to ask for image upload? | 17:35 |
| mtreinish | that's like neutron 503s when you try to create a tenant network in a vlan provider network config | 17:35 |
| SpamapS | mtreinish: no, neutron knows: you're doing it wrong | 17:35 |
| SpamapS | 503 is just the way to tell you that you doing it wrong, broke it's brain | 17:35 |
| SpamapS | its | 17:35 |
| greghaynes | SpamapS: I did, and they confirmed I have it. I think that means I can task-create | 17:35 |
| SpamapS | greghaynes: ok, so this sort of proves it doesn't it? :) | 17:36 |
| SpamapS | greghaynes: can you try the shade tests? | 17:36 |
| mtreinish | SpamapS: heh, I know: service unavailable == you're doing it wrong | 17:36 |
| SpamapS | What I'm seeing is that when I run that against devstack, it fails because of a missing field, import_from_format | 17:36 |
| greghaynes | SpamapS: is there a patch I need to make it hit a remote cloud? | 17:37 |
| SpamapS | mtreinish: It's the REST equivalent of "I'm not even going to respond to that." | 17:37 |
| SpamapS | greghaynes: no it respects OS_* | 17:37 |
| greghaynes | oh, neat | 17:37 |
| SpamapS | I think... | 17:37 |
| greghaynes | I could see how that would work via the os-cloud-cfg magic | 17:38 |
| mtreinish | SpamapS: or you could just give a 4xx response with a real error message | 17:38 |
| SpamapS | mtreinish: politeness is overrated. | 17:38 |
| mtreinish | SpamapS: definitely :) | 17:38 |
| SpamapS | mtreinish: http://cdn.meme.am/instances/500x/61733074.jpg | 17:40 |
| mtreinish | SpamapS: lol | 17:41 |
| greghaynes | python dep installing for tests is slow... | 17:41 |
| greghaynes | youall must depend on either lxml or pyopenssl | 17:42 |
| SpamapS | yep | 17:42 |
| SpamapS | both | 17:42 |
| SpamapS | takes my hpcloud instance 8 minutes to build the venv | 17:42 |
| mtreinish | greghaynes: if you're using any of the python-*clients both get pulled in | 17:43 |
| greghaynes | awesome! | 17:43 |
| SpamapS | so if I get around the import_from_format issue by adding that field in shade, the next problem I have is that glance-api wants to fetch from http:// and I'm just giving it a swift container name. | 17:43 |
| greghaynes | SpamapS: ok, tests passed with OS_* set to rax.. I suspect it didnt respect them | 17:46 |
| greghaynes | actually, your tests are mostly fake clients? | 17:46 |
| SpamapS | greghaynes: -e functional | 17:47 |
| SpamapS | greghaynes: 'tox -efunctional test_image' | 17:47 |
| SpamapS | greghaynes: I don't care about the unit tests. :) | 17:47 |
| greghaynes | ok, well, dep installing again then | 17:48 |
| SpamapS | greghaynes: :( | 17:48 |
| greghaynes | hehe | 17:48 |
| greghaynes | says passed... | 17:50 |
| greghaynes | im not sure it did anything | 17:50 |
| greghaynes | ah, the arg messed it up | 17:51 |
| *** e0ne has quit IRC | 17:57 | |
| SpamapS | greghaynes: the arg should limit to just test_image | 18:00 |
| *** harlowja_away is now known as harlowja | 18:00 | |
| SpamapS | greghaynes: oh, maybe it has to be tox -efunctional ? | 18:00 |
| SpamapS | greghaynes: ? | 18:03 |
| mtreinish | SpamapS: sigh I'm trying to run tempest against rax, it's exploding on the date format in the tokens... | 18:03 |
| mtreinish | iterop ftw | 18:03 |
| *** krykowski has quit IRC | 18:04 | |
| *** dhellmann has quit IRC | 18:05 | |
| mtreinish | ugh, there is are microseconds in the response... | 18:09 |
| SpamapS | mtreinish: microseconds are microuseful | 18:12 |
| * harlowja didn't do it, lol | 18:15 | |
| mtreinish | SpamapS, greghaynes: http://paste.openstack.org/show/212486/ so I get the 413 on the image create | 18:15 |
| greghaynes | sounds right | 18:16 |
| *** harlowja has quit IRC | 18:16 | |
| *** harlowja has joined #openstack-glance | 18:16 | |
| mtreinish | SpamapS: heh, especially for an expiration time, you need as much precision as possible :) | 18:16 |
| SpamapS | alright, so perhaps whats needed is a devstack-gate that has that configured, and tests that use tasks/imports | 18:16 |
| *** harlowja has quit IRC | 18:17 | |
| mtreinish | SpamapS: configure devstack to 413 on these requests? I would assume the tasks/imports stuff would just work if we added tests for them | 18:18 |
| SpamapS | mtreinish: well I'm kind of the mindset that since the point of the tasks is that some people don't want glance-api to be in the data-plane .... testing that tasks work when 413 happens seems like the right way to go? | 18:20 |
| *** harlowja has joined #openstack-glance | 18:20 | |
| SpamapS | mtreinish: but isolating it into the two things is fine too :-P | 18:20 |
| mtreinish | SpamapS: oh, I don't care one way or the other, I was just confused on exactly you were proposing. I do feel that if we want glance to behave that way it should be the glance default | 18:23 |
| mtreinish | because our job config permutation count is high enough as it is :) | 18:23 |
| *** harlowja has quit IRC | 18:29 | |
| *** harlowja has joined #openstack-glance | 18:29 | |
| *** e0ne has joined #openstack-glance | 18:31 | |
| *** e0ne is now known as e0ne_ | 18:31 | |
| SpamapS | mtreinish: good point | 18:31 |
| *** e0ne_ is now known as e0ne | 18:32 | |
| *** e0ne is now known as e0ne_ | 18:32 | |
| *** e0ne_ is now known as e0ne | 18:32 | |
| *** e0ne has quit IRC | 18:32 | |
| *** annegentle has quit IRC | 18:36 | |
| *** e0ne has joined #openstack-glance | 18:48 | |
| *** annegentle has joined #openstack-glance | 18:49 | |
| *** annegentle has quit IRC | 18:52 | |
| *** annegentle has joined #openstack-glance | 18:53 | |
| *** annashen has joined #openstack-glance | 18:55 | |
| *** annashen has quit IRC | 19:01 | |
| *** annashen has joined #openstack-glance | 19:02 | |
| *** ericpeterson has quit IRC | 19:06 | |
| *** dhellmann has joined #openstack-glance | 19:06 | |
| *** annashen has quit IRC | 19:06 | |
| *** jaypipes has quit IRC | 19:06 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/glance: Updated from global requirements https://review.openstack.org/176827 | 19:16 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-glanceclient: Updated from global requirements https://review.openstack.org/155581 | 19:23 |
| *** jaypipes has joined #openstack-glance | 19:23 | |
| *** ajayaa has quit IRC | 19:23 | |
| *** david-lyle has quit IRC | 19:44 | |
| *** spzala has joined #openstack-glance | 19:44 | |
| *** david-lyle has joined #openstack-glance | 19:44 | |
| *** ajayaa has joined #openstack-glance | 19:48 | |
| *** ericpeterson has joined #openstack-glance | 19:53 | |
| *** arif-ali has quit IRC | 19:56 | |
| harlowja | is there going to be a session on taskflow + glance in the summit, just wondering? | 20:11 |
| * harlowja should probably attend that one if so :) | 20:13 | |
| *** annegentle has quit IRC | 20:15 | |
| *** annegentle has joined #openstack-glance | 20:23 | |
| *** e0ne has quit IRC | 20:26 | |
| SpamapS | harlowja: is this thing on? | 20:29 |
| * SpamapS taps | 20:29 | |
| harlowja | hollaa | 20:29 |
| harlowja | SpamapS sup dawg | 20:31 |
| lifeless | SpamapS: HOWWWDEEEEEEE | 20:34 |
| SpamapS | you are both wonderful, but I'm more wondering where the regulars are. ;) | 20:35 |
| SpamapS | I have glancey questiones | 20:35 |
| *** e0ne has joined #openstack-glance | 20:36 | |
| lifeless | SpamapS: don't ask to ask :) | 20:38 |
| SpamapS | lifeless: ^^^ quite a while ago, asking if anybody here knows if glance v2 is tested in the gate | 20:38 |
| SpamapS | lifeless: and the ensuing discovery has been... interesting. :) | 20:38 |
| SpamapS | basically the glance v2 in the gate is not hte glance v2 in RAX | 20:38 |
| lifeless | have you asked treinish or sdague ? | 20:38 |
| SpamapS | lifeless: treinish wrote the v2 tests :) | 20:39 |
| SpamapS | and yes | 20:39 |
| lifeless | great | 20:39 |
| lifeless | I'm surprised they didn't know | 20:39 |
| SpamapS | the deal is, those test the one where glance is still accepting PUT's of images. | 20:39 |
| SpamapS | which RAX returns 413 for | 20:39 |
| SpamapS | you have to instead create an import task in RAX | 20:39 |
| SpamapS | and that code exists in upstream | 20:39 |
| SpamapS | but is zomg different | 20:39 |
| SpamapS | and currently I'm making an attempt to see if I can get it to work at all | 20:40 |
| lifeless | may the force be with you :) | 20:41 |
| *** e0ne has quit IRC | 20:44 | |
| *** mtreinish_ has joined #openstack-glance | 20:45 | |
| kragniz | SpamapS: nikhil_k_ is the person to talk to about tasks | 20:47 |
| kragniz | (also rax) | 20:47 |
| SpamapS | kragniz: indeed I do see some reference to that name in the commits | 20:50 |
| *** mtreinish has quit IRC | 20:51 | |
| *** mtreinish_ is now known as mtreinish | 20:51 | |
| SpamapS | currently getting 500's now for missing fields | 20:52 |
| kragniz | :/ | 20:54 |
| kragniz | I'm afraid I don't know much about tasks | 20:55 |
| SpamapS | and when I add them, even more fun.. | 20:55 |
| SpamapS | DBError: (ProgrammingError) (1064, "You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '), NULL, NULL, 'queued', 0, NULL, 0, 0, '19572f9748f24bab8c3ed76592f49151', 0)' at line 1") 'INSERT INTO images (created_at, updated_at, deleted_at, deleted, id, name, disk_format, container_format, size, virtual_size, status, is_public, checksum, min_disk, min_ram, owner, | 20:55 |
| *** belmoreira has joined #openstack-glance | 20:55 | |
| SpamapS | kragniz: it's ok, it's just that we have to stop merging API's without proper tests | 20:56 |
| *** raildo has quit IRC | 20:58 | |
| kragniz | oh dear... | 21:01 |
| kragniz | SpamapS: I agree with that | 21:01 |
| *** peristeri has quit IRC | 21:09 | |
| sigmavirus24 | SpamapS: flaper87 can help with tasks too | 21:10 |
| SpamapS | Cool | 21:11 |
| SpamapS | I'm still nailing down why it might be so different when talking to RAX vs. a devstack. | 21:12 |
| *** ajayaa has quit IRC | 21:13 | |
| sigmavirus24 | Well then you definitely want nikhil_k_ for that | 21:15 |
| sigmavirus24 | He's on our public cloud team that manages our glance deployment sort of | 21:15 |
| sigmavirus24 | Also rosmaita | 21:15 |
| sigmavirus24 | And Hemmanth but I don't see him around | 21:15 |
| sigmavirus24 | Oh jcook may be able to help you too | 21:15 |
| *** arif-ali has joined #openstack-glance | 21:19 | |
| *** annashen has joined #openstack-glance | 21:21 | |
| SpamapS | so my biggest concern is that backtraces during tasks are not leading to the tasks marked as failed | 21:21 |
| SpamapS | they just sit as 'processing' forever | 21:21 |
| nikhil_k_ | sigmavirus24: SpamapS : hi | 21:22 |
| SpamapS | nikhil_k_: greetings | 21:23 |
| nikhil_k_ | :-) | 21:23 |
| nikhil_k_ | how can I help? | 21:23 |
| nikhil_k_ | trying to scroll back | 21:23 |
| nikhil_k_ | and read fast | 21:23 |
| SpamapS | nikhil_k_: So we have some logic in the shade library which works well for uploading images to RAX | 21:23 |
| SpamapS | nikhil_k_: Simpler to summarize here. | 21:23 |
| SpamapS | nikhil_k_: I'm trying to write functional tests to test this logic against devstack. | 21:23 |
| SpamapS | and that does not work. :-/ | 21:24 |
| nikhil_k_ | huh | 21:24 |
| nikhil_k_ | uploading as in POST/PUT or tasks? | 21:24 |
| nikhil_k_ | I see tasks mentioned above | 21:25 |
| SpamapS | nikhil_k_: http://git.openstack.org/cgit/openstack-infra/shade/tree/shade/__init__.py#n1146 is the logic | 21:25 |
| SpamapS | nikhil_k_: uploading as in tasks | 21:25 |
| nikhil_k_ | ohk | 21:25 |
| nikhil_k_ | yep, that's not upstream yet | 21:25 |
| nikhil_k_ | issue with glance_store not completely compatible | 21:25 |
| nikhil_k_ | and we need to copy a bunch of swift driver onto the tasks script for doing something that can be 2 liner | 21:26 |
| SpamapS | ok, so that is RAX-private still then? | 21:26 |
| SpamapS | as in, that API hasn't stabilized yet? | 21:27 |
| nikhil_k_ | I try to avoid useing that word, prefer "Rack" the other one is our stock symbol. Also anything I say here isn't related to stocks ;-) | 21:27 |
| nikhil_k_ | But yes | 21:27 |
| nikhil_k_ | The API is all good | 21:27 |
| SpamapS | Except there's no testing of the API.. so... ;) | 21:27 |
| nikhil_k_ | tasks allow some elements ni the body request | 21:28 |
| nikhil_k_ | that can be deployer specific | 21:28 |
| nikhil_k_ | if you are seeing that error then something's really wrong | 21:28 |
| SpamapS | Yeah that's entirely fine, and as long as the task is rejected with a 4xx because you didn't have the right deployer-specific data that's good. | 21:28 |
| SpamapS | Currently I get 500, or an immortal processing task. | 21:28 |
| *** Longgeek has quit IRC | 21:29 | |
| nikhil_k_ | Oh I see you are worried about the DB error specifically | 21:29 |
| SpamapS | nikhil_k_: so I'm going to back away. If you say it doesn't work upstream, I'm not going to try and brute force around that. | 21:29 |
| nikhil_k_ | Please do not | 21:29 |
| SpamapS | nikhil_k_: the DB error is at the end of a long chain of other fields that were required in devstack that weren't required for an upload to the rackspace cloud. | 21:30 |
| nikhil_k_ | Upstream is still fragile. Hopefully I get enough time this cycle to fix this. | 21:30 |
| nikhil_k_ | SpamapS: yep, the import script in master isn't nice yet | 21:30 |
| *** julim has quit IRC | 21:30 | |
| nikhil_k_ | Basically, the validation at API level is minimal by design | 21:30 |
| nikhil_k_ | because, these are supposed to be asynchronous | 21:31 |
| SpamapS | And that would be fine.. if they produced errors.. or were retried.. | 21:31 |
| nikhil_k_ | So, we need to tighten valiadtion in the import script for it to work nice in devstack :-) All bugs welcome! | 21:31 |
| nikhil_k_ | Agree | 21:31 |
| SpamapS | nikhil_k_: that brings up another point. The async bits are realy poorly documented. It took me 2 hours of digging to figure out that glance-api just executes things immediately. | 21:31 |
| SpamapS | I was looking everywhere for an async worker. | 21:32 |
| nikhil_k_ | SpamapS: even that's configurable. Depending on the type of executor implmented :) | 21:32 |
| SpamapS | see 'executor implemented' ... ??? | 21:32 |
| nikhil_k_ | ohk, yes the default one is just an eventlet thread >.< | 21:32 |
| nikhil_k_ | err | 21:32 |
| nikhil_k_ | "executor deployed" | 21:32 |
| SpamapS | I have the result of git clone glance. I see no executor other than the one that gets pulled into glance-api in very mysterious ways. | 21:32 |
| nikhil_k_ | heh | 21:33 |
| nikhil_k_ | taskflow is pulled in | 21:33 |
| *** annashen has quit IRC | 21:33 | |
| jcook | sigmavirus24, SpamapS: hmm? | 21:34 |
| sigmavirus24 | jcook: nevermind. nikhil_k_ helped SpamapS out | 21:34 |
| sigmavirus24 | sorry for the ping | 21:34 |
| jcook | sigmavirus24: ha, procrastination ftw | 21:34 |
| sigmavirus24 | ^5 jcook | 21:34 |
| jcook | nikhil_k_, sigmavirus24, SpamapS: glad I could be of help ;) | 21:35 |
| nikhil_k_ | we are too :) | 21:35 |
| jcook | nikhil_k_: >_>: | 21:35 |
| nikhil_k_ | SpamapS: So, how can I really help? | 21:35 |
| jcook | >_>; * | 21:35 |
| nikhil_k_ | :) | 21:35 |
| nikhil_k_ | jcook: :P | 21:35 |
| greghaynes | nikhil_k_: Is there a patch stack upstream for the swift import stuff? | 21:36 |
| SpamapS | So, I'll say again, I'm going to back away from this. I don't actually think the task API is the right way to go, I think interoperability trumps scalability here, and it's worth the effort to make a scalable glance upload endpoint so end-users don't have to understand how to write to all the backends.... | 21:36 |
| greghaynes | nikhil_k_: something we could pull and test against that might resemble what rax is doing | 21:36 |
| *** harlowja is now known as harlowja_away | 21:36 | |
| SpamapS | That, combined with the fact that this doesn't work upstream yet, and isn't tested even minimally, makes it a huge time sink that I can't afford. | 21:36 |
| jcook | those statements perked my interest | 21:37 |
| nikhil_k_ | greghaynes: I can share stuff. It's no biggie as I had shared in a mini-summit anyways :) | 21:37 |
| jcook | "interoperability trumps scalability" | 21:37 |
| *** harlowja_away is now known as harlowja | 21:37 | |
| greghaynes | nikhil_k_: yes please | 21:37 |
| nikhil_k_ | SpamapS: The reason tasks were implemented is because glance cannot handle objects as well as something like swift | 21:38 |
| SpamapS | jcook: Indeed. We're already bending over backward to have nodepool be able to upload to hpcloud (which uses glance v1) and rackspace (which uses v2+tasks). When it lands upstream, we might have _three_ ways we have to upload... :-P | 21:38 |
| mtreinish | if someone could take a look at https://bugs.launchpad.net/glance/+bug/1450209 glance is spewing garbage in it's log files on every gate run | 21:38 |
| openstack | Launchpad bug 1450209 in Glance "Glance API logs a never ending stream of 404 tracebacks from glance-registry " [Undecided,New] | 21:38 |
| jcook | SpamapS: not disagreeing about statement, just curious what we're talking about since scalability is on my radar | 21:39 |
| *** stpierre has quit IRC | 21:39 | |
| greghaynes | not to mention the only way were figuring out how to make it work with these clouds is lots of info were getting directly from the cloud operators - making this work against an arbitraty cloud would be almost hopeless | 21:39 |
| SpamapS | Whereas if you can just take the swift proxy that does all this heavy PUT lifting, and put it in front of glance at the right spot, I think you'd be able to just let users PUT to /images/foo | 21:39 |
| jcook | SpamapS: oooh, now you are talking about a problem I think I am looking to tackle | 21:40 |
| SpamapS | nikhil_k_: No disrespect is meant. I completely understand why a real, heavily loaded cloud, might choose the tasks interface. | 21:40 |
| SpamapS | nikhil_k_: I'm more concerned for the ecosystem which will need to deal with this problem, and may not all be able to store their images in swift, at which point, the interop matrix gets _really_ tricky. | 21:40 |
| jcook | SpamapS: I have a cross-project summit session proposed that might interest you | 21:40 |
| SpamapS | jcook: yes? | 21:41 |
| jcook | SpamapS: what is the problem you are trying to solve? upload reliability, upload performance, download reliability, download performance, or something else | 21:41 |
| SpamapS | jcook: so, the only problem I want to solve is that the tasks systme that Rackspace has chosen to facilitate uploads to their clouds has a hidden interoperability problem in it as it moves upstream. | 21:42 |
| jcook | SpamapS: ah, perhaps not then. I have to admin that my knowledge of tasks is limited. For us, upload reliability is more important than performance right now. | 21:43 |
| SpamapS | jcook: well are you experiencing unreliable uploads because glance-api wont' scale? | 21:44 |
| jcook | SpamapS: I need to spend a little time looking at this subject before I can talk to it | 21:44 |
| SpamapS | That seems like it could be a real problem. | 21:44 |
| sigmavirus24 | mtreinish: I'll take a look at what's landed recently | 21:44 |
| jcook | SpamapS: I'd say unreliable in that Glance doesn't do retries well | 21:44 |
| SpamapS | Likely one nikhil_k_ and the Rackspace team spent quite a bit of time looking at before choosing the path they took. | 21:44 |
| mtreinish | sigmavirus24: thanks, I don't know when that started tbh, I just noticed it a couple of days go | 21:44 |
| * mtreinish checks logstash | 21:45 | |
| jcook | SpamapS: moving an immutable object should not result in 409 on full try or fail midstream and not retry bit that failed | 21:45 |
| SpamapS | jcook: indeed. | 21:45 |
| sigmavirus24 | mtreinish which jobs are causing this? Jobs from master or kilo? | 21:45 |
| mtreinish | sigmavirus24: I've seen it on master, I haven't checked kilo | 21:46 |
| mtreinish | sigmavirus24: looks like just on master: http://goo.gl/7hol8Y | 21:49 |
| SpamapS | jcook: I mean, another option which we could do is just require that all clouds have a swift API and let the image API just be a catalog of pointers to swift objects. | 21:49 |
| SpamapS | jcook: if that swift API happens to write to CEPH, or gluster, or sheepdog, or NFS so be it. | 21:50 |
| SpamapS | jcook: then your image upload reliability would be as good as swift's. | 21:50 |
| jcook | jcook: that seems like an implementation that could work in the context of what I had in mind | 21:51 |
| SpamapS | jcook: Lets make sure to connect at the summit and discuss this. :) | 21:51 |
| SpamapS | nikhil_k_: you too. :) | 21:51 |
| *** annegentle has quit IRC | 21:51 | |
| jcook | SpamapS: https://etherpad.openstack.org/p/liberty-caching | 21:52 |
| jcook | Line item 15: https://docs.google.com/spreadsheets/d/1vCTZBJKCMZ2xBhglnuK3ciKo3E8UMFo5S5lmIAYMCSE/edit#gid=827503418 | 21:52 |
| jcook | SpamapS: that session is mostly about downloads and caching. However, if you look at reference architecture, it is very related to what you are saying here | 21:52 |
| SpamapS | jcook: didn't you hear? CEPH solved that. Just give everything 4x 10Gbe and sacrifice a goat to the RADOS gods. | 21:53 |
| jcook | SpamapS: Upload Recovery proposal is a bit terse atm, but here: https://etherpad.openstack.org/p/liberty-glance-summit-topics | 21:53 |
| jcook | SpamapS: lol | 21:53 |
| *** ericpeterson has quit IRC | 21:54 | |
| SpamapS | jcook: I took issue with your statement about LRU btw ;) | 21:55 |
| jcook | SpamapS: did you write it >_>; | 21:56 |
| * jcook gives SpamapS a virtual hug. | 21:56 | |
| sigmavirus24 | mtreinish: this is interesting. I'm not seeing much that should/could be causing this | 21:57 |
| *** ericpeterson has joined #openstack-glance | 21:57 | |
| SpamapS | http://en.wikipedia.org/wiki/Cache_algorithms#Examples hah fascinating read | 21:57 |
| jcook | SpamapS: oh, I see your comment. I should probably clarify that it doesn't use the column that is basically hit count | 21:57 |
| SpamapS | I had never heard of "the clairvoyant algorithm" | 21:58 |
| SpamapS | jcook: that would be LFU | 21:58 |
| jcook | SpamapS: I think we need something that considers both | 21:59 |
| SpamapS | jcook: really I think glance would benefit from one of the LRU+LFU approaches. | 22:00 |
| jcook | SpamapS: LRU makes sense in processor cache | 22:00 |
| jcook | SpamapS: +1 | 22:00 |
| lifeless | caching what data? | 22:00 |
| SpamapS | jcook: LFU, IIRC, does better with higher-cost misses. | 22:00 |
| jcook | lifeless: images in glance | 22:01 |
| jcook | lifeless: goal is to improve build times | 22:01 |
| lifeless | jcook: on hypervisors? | 22:01 |
| SpamapS | honestly if you want fast builds though, CEPH, like not even joking, is the way to go. | 22:01 |
| jcook | lifeless: was talking about in Glance | 22:01 |
| lifeless | jcook: yes, I'm confused and seeking to resolve the confusion | 22:01 |
| SpamapS | worth the extra overhead of a dedicated storage network. | 22:01 |
| jcook | SpamapS: I should probably look into CEPH | 22:02 |
| lifeless | SpamapS: sheepdog should be equivalent | 22:02 |
| jcook | download times from glance cache are actually pretty quick | 22:02 |
| lifeless | same basic principle - scaled spindles, COW | 22:02 |
| SpamapS | lifeless: I don't know how sheepdog works, but probably worth glancing at (no pun intended) | 22:02 |
| jcook | at least from my tests on rax prod | 22:02 |
| lifeless | jcook: so the cache is in addition to the hypervisor image cache | 22:02 |
| lifeless | jcook: its holding a copy of the compressed image in a glance server? | 22:02 |
| jcook | lifeless: yeah, glance has cache on api nodes right now | 22:03 |
| jcook | lifeless: it uses LRU (which actually had a bug that made it MRU until recently fixed) | 22:03 |
| SpamapS | doohhhhhh | 22:03 |
| openstackgerrit | Kent Wang proposed openstack/python-glanceclient: Support add multiple images/tenants in members https://review.openstack.org/178857 | 22:03 |
| lifeless | jcook: so, why isn't that done via e.g. squid or varnish ? | 22:03 |
| SpamapS | IIRC MRU is only advisable when you have completely random access patterns | 22:03 |
| lifeless | jcook: e.g. dedicated memory caching tools | 22:03 |
| jcook | lifeless: ding ding ding, we have a winner | 22:03 |
| jcook | lifeless: +1 | 22:04 |
| lifeless | SpamapS: also sequential apparently.. | 22:04 |
| SpamapS | lifeless: right because you'd at least get some hits. | 22:04 |
| lifeless | jcook: I mean, I can advise based on the data we have in the squid project about caching algorithms... | 22:04 |
| SpamapS | lifeless: broken clock problem :) | 22:04 |
| lifeless | jcook: but honestly just using squid would be better :) | 22:04 |
| SpamapS | varnish is so hip tho | 22:04 |
| lifeless | same same | 22:04 |
| jcook | lifeless: we are actively testing that, we have some interesting results show that glance can be ever so slightly faster but still want to move to squid | 22:05 |
| SpamapS | squid == Zoolander, Varnish == Hansel | 22:05 |
| lifeless | jcook: if you want me to have a look over your squid conf, let me know | 22:05 |
| lifeless | SpamapS: oh I like | 22:05 |
| SpamapS | http://2.bp.blogspot.com/-cDs50EboctY/VQCCNT6qzDI/AAAAAAAACvc/STbSETJD3rI/s1600/tumblr_nl08iwtDZf1qlvwnco2_500.gif | 22:05 |
| jcook | lifeless: ultimately, though I want small chunks cached on determistically selected node with random order parallel downloads | 22:05 |
| lifeless | SpamapS: I'll get practicing my blue velvet | 22:05 |
| jcook | lifeless: I should have common sync up with you | 22:06 |
| SpamapS | lifeless: you'll need Magnum if they rewrite varnish in Rust... | 22:06 |
| lifeless | SpamapS: whatshisname would never | 22:06 |
| sigmavirus24 | mtreinish: are you still around? | 22:06 |
| SpamapS | true true | 22:07 |
| lifeless | jcook: small chunks for scatter gather reads? | 22:07 |
| lifeless | jcook: I suspect you'd be better off with very optimised linear on single nodes, for everything but extreme scales - and there just redundant copies | 22:07 |
| jcook | lifeless: use case, build n thousand instances simultaneously off uncachedi mage | 22:07 |
| lifeless | jcook: multicast! | 22:07 |
| lifeless | jcook: I wrote a spec for that for glance | 22:08 |
| lifeless | specish I should say - etherpad, before in-gerrit specs | 22:08 |
| SpamapS | a few past-lives ago I used multicast-rsync to deploy images. :) | 22:08 |
| SpamapS | works quite well even at medium-scale of 40 nodes. | 22:09 |
| jcook | lifeless: maybe, I think deterministically caching in selected cache (either glance or squid vs using peer caching) and random ordering your downloads means object moves from swift on average 1 time when builds are less than segments, and chunks can be pulled in parallel from glance without wasting bandwidth or space | 22:09 |
| mtreinish | sigmavirus24: yep | 22:10 |
| jcook | lifeless: I am intrigued by your multicast option though, I should look at this | 22:10 |
| * SpamapS needs to step away from the super fun caching talk for a bit. :) | 22:10 | |
| mtreinish | sigmavirus24: sorry, yeah it looks like it's been logging that for >= 10days it goes back our full logstash history | 22:10 |
| sigmavirus24 | mtreinish: I'm not seeing what's caused these to suddenly start popping up so much | 22:10 |
| lifeless | jcook: oh look https://blueprints.launchpad.net/glance/+spec/multicast-image-transfer | 22:10 |
| sigmavirus24 | mtreinish: ah that makes more sense | 22:10 |
| lifeless | and the etherpad https://etherpad.openstack.org/p/glance-multicast-image-transfer | 22:11 |
| lifeless | I think we ended up saying that frisbee looked super interesting | 22:11 |
| lifeless | in terms of working, being fairly modern | 22:11 |
| sigmavirus24 | mtreinish: most of the changes in the apparently affected areas are from 2014 | 22:11 |
| jcook | lifeless: I'll have to read through that and think about the trade offs of the different solutions. Seems to be similar use case | 22:12 |
| *** vijendar has quit IRC | 22:12 | |
| lifeless | jcook: cool | 22:12 |
| sigmavirus24 | I'll have to dig into this more tomorrow | 22:13 |
| lifeless | jcook: I'm happy to be a sounding board! | 22:13 |
| lifeless | jcook: I'm very keen on using the right tool for a given problem :> | 22:13 |
| jcook | lifeless: ^_^ | 22:14 |
| jcook | lifeless: will you be at the summit? | 22:14 |
| harlowja | I'm back, had food, SpamapS nikhil_k_ poke me if u need me | 22:14 |
| lifeless | jcook: I shall | 22:14 |
| harlowja | *actually was back a little while ago, but just remembered about this talk/chat, lol | 22:14 |
| jcook | lifeless, SpamapS: we should sync up at summit on the caching stuff / quick build times | 22:15 |
| SpamapS | jcook: yes indeed | 22:15 |
| mtreinish | sigmavirus24: hmm, ok | 22:15 |
| SpamapS | jcook: you could also stick your nova-compute cache on AFS.. ;) | 22:16 |
| jcook | SpamapS: Andrew File System | 22:18 |
| *** belmoreira has quit IRC | 22:18 | |
| jcook | ? | 22:18 |
| lifeless | yes, SpamapS is getting indoctrinated by infra atm | 22:18 |
| mtreinish | sigmavirus24: I mean it could have been going on that long, people normally don't look at the log files unless the tests fail | 22:18 |
| mtreinish | but I think someone would have noticed it long before if it was introduced last year | 22:19 |
| jcook | yeah DFS is definitely a possibility for the HV cache. Reduces chance of cache miss | 22:19 |
| sigmavirus24 | mtreinish: yeah that's what I'm thinking | 22:19 |
| sigmavirus24 | so i'm going to look elsewhere | 22:19 |
| sigmavirus24 | and try to replicate this | 22:19 |
| SpamapS | jcook: realistically putting it on AFS just means you're slave to AFS's caching :) | 22:22 |
| SpamapS | which is overly generic and likely won't be as optimized as a multicast solution. | 22:22 |
| jcook | SpamapS: yeah, I'd be curious how fast it is vs pulling from Glance / squid / multicat / w/e | 22:23 |
| jcook | SpamapS: Glance is actually pretty quick depending how close it is to HV, cache misses are a real problem though | 22:24 |
| nikhil_k_ | It prolly depends on the kind of glance node | 22:28 |
| nikhil_k_ | for a generic solution like this glance supporting bit torrent would help | 22:28 |
| *** ericpeterson has quit IRC | 22:44 | |
| *** bkopilov has quit IRC | 23:10 | |
| *** bkopilov has joined #openstack-glance | 23:25 | |
| *** ericpeterson has joined #openstack-glance | 23:45 | |
| *** dims has quit IRC | 23:50 | |
| *** chlong has joined #openstack-glance | 23:50 | |
| *** ericpeterson has quit IRC | 23:51 | |
| *** annashen has joined #openstack-glance | 23:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
