| asalkeld | so s34n change your devstack to rabbit | 00:00 |
|---|---|---|
| *** halfss has joined #openstack-dev | 00:00 | |
| asalkeld | (not qpidd) | 00:00 |
| s34n | asalkeld: it was rabbit by default | 00:00 |
| asalkeld | I changed mine | 00:00 |
| asalkeld | so go with default | 00:00 |
| s34n | asalkeld: you changed to qpid, right? | 00:01 |
| asalkeld | yip | 00:01 |
| asalkeld | fedora default | 00:01 |
| asalkeld | for openstack | 00:01 |
| s34n | asalkeld: ok. beam.smp seams to be there for rabbit | 00:03 |
| s34n | I stopped rabbit, but beam is lingering. I'm not sure how to make it go away nicely | 00:04 |
| *** steveb_ has quit IRC | 00:05 | |
| jkoelker | the shutdown script from rabbit should stop it | 00:05 |
| jkoelker | any epmd will linger | 00:05 |
| jkoelker | s/any/only/ | 00:05 |
| jkoelker | if you're not concerned about messages in it, you can just kill it | 00:06 |
| *** mikal has quit IRC | 00:09 | |
| *** halfss has quit IRC | 00:11 | |
| *** mikal has joined #openstack-dev | 00:11 | |
| *** novas0x2a|laptop has quit IRC | 00:12 | |
| *** Aaton is now known as Aaton_off | 00:16 | |
| *** mnewby has quit IRC | 00:16 | |
| *** spiffxp has quit IRC | 00:17 | |
| *** ecarlin has quit IRC | 00:21 | |
| *** zhuadl has joined #openstack-dev | 00:22 | |
| *** Ryan_Lane has quit IRC | 00:25 | |
| *** littleidea has quit IRC | 00:26 | |
| *** halfss has joined #openstack-dev | 00:28 | |
| *** jakedahn_zz is now known as jakedahn | 00:28 | |
| *** johnpostlethwait has quit IRC | 00:29 | |
| *** anniec has quit IRC | 00:32 | |
| *** anniec has joined #openstack-dev | 00:32 | |
| *** koolhead17 has joined #openstack-dev | 00:36 | |
| *** halfss has quit IRC | 00:37 | |
| *** maoy has joined #openstack-dev | 00:39 | |
| *** littleidea has joined #openstack-dev | 00:41 | |
| *** pixelbeat has quit IRC | 00:42 | |
| *** jakedahn is now known as jakedahn_zz | 00:45 | |
| *** nati_uen_ has joined #openstack-dev | 00:48 | |
| *** nati_uen_ has quit IRC | 00:48 | |
| *** nati_uen_ has joined #openstack-dev | 00:48 | |
| *** amotoki_ has joined #openstack-dev | 00:48 | |
| *** spiffxp has joined #openstack-dev | 00:49 | |
| *** sandywalsh has quit IRC | 00:49 | |
| *** nati_ueno has quit IRC | 00:50 | |
| *** amotoki has quit IRC | 00:50 | |
| *** edygarcia has quit IRC | 00:51 | |
| *** spiffxp has quit IRC | 00:52 | |
| *** hemna has quit IRC | 00:53 | |
| *** mikal has quit IRC | 00:56 | |
| *** mikal has joined #openstack-dev | 00:58 | |
| *** steveb_ has joined #openstack-dev | 01:02 | |
| *** Mandell_ has quit IRC | 01:07 | |
| *** ncode has joined #openstack-dev | 01:08 | |
| *** Dr_Who has joined #openstack-dev | 01:13 | |
| *** ncode has quit IRC | 01:15 | |
| *** cloudvirt has joined #openstack-dev | 01:18 | |
| *** mfer has joined #openstack-dev | 01:19 | |
| *** johnpur has quit IRC | 01:19 | |
| *** mfer has quit IRC | 01:20 | |
| *** zhuadl has quit IRC | 01:22 | |
| *** cloudvirt has quit IRC | 01:23 | |
| *** danwent has joined #openstack-dev | 01:24 | |
| *** PotHix has quit IRC | 01:26 | |
| *** jtran has quit IRC | 01:32 | |
| *** rkukura has joined #openstack-dev | 01:42 | |
| *** nati_ueno has joined #openstack-dev | 01:42 | |
| *** issackelly has quit IRC | 01:45 | |
| *** roge has joined #openstack-dev | 01:45 | |
| *** nati_uen_ has quit IRC | 01:45 | |
| *** littleidea has quit IRC | 01:46 | |
| *** Dr_Who has quit IRC | 01:47 | |
| *** danwent has quit IRC | 01:49 | |
| *** nunosantos has quit IRC | 01:49 | |
| *** primeministerp has quit IRC | 01:52 | |
| *** blamar has joined #openstack-dev | 01:53 | |
| *** nati_ueno has quit IRC | 01:55 | |
| *** anniec has quit IRC | 01:59 | |
| *** primeministerp has joined #openstack-dev | 01:59 | |
| *** roge has quit IRC | 02:00 | |
| *** jog0 has quit IRC | 02:01 | |
| *** jdurgin has quit IRC | 02:02 | |
| *** roge has joined #openstack-dev | 02:03 | |
| *** gongys has quit IRC | 02:06 | |
| *** bencherian has quit IRC | 02:07 | |
| *** amotoki_ has quit IRC | 02:11 | |
| *** littleidea has joined #openstack-dev | 02:12 | |
| *** sunxin has joined #openstack-dev | 02:15 | |
| *** bencherian has joined #openstack-dev | 02:15 | |
| *** zhuadl has joined #openstack-dev | 02:18 | |
| *** Dr_Who has joined #openstack-dev | 02:19 | |
| *** Dr_Who has joined #openstack-dev | 02:20 | |
| *** Dr_Who has quit IRC | 02:22 | |
| *** littleidea has left #openstack-dev | 02:23 | |
| *** halfss has joined #openstack-dev | 02:24 | |
| *** halfss has quit IRC | 02:26 | |
| *** nati_ueno has joined #openstack-dev | 02:30 | |
| *** troytoman is now known as troytoman-away | 02:31 | |
| *** dolphm has joined #openstack-dev | 02:31 | |
| *** nati_ueno has quit IRC | 02:31 | |
| *** nati_ueno has joined #openstack-dev | 02:32 | |
| *** Dr_Who has joined #openstack-dev | 02:33 | |
| *** koolhead17 has quit IRC | 02:35 | |
| *** n0ano has quit IRC | 02:38 | |
| *** rkukura has quit IRC | 02:42 | |
| *** jakedahn_zz is now known as jakedahn | 02:43 | |
| *** nati_ueno has quit IRC | 02:43 | |
| *** bencherian has quit IRC | 02:46 | |
| *** koolhead17 has joined #openstack-dev | 02:48 | |
| *** blamar has quit IRC | 02:50 | |
| *** bencherian has joined #openstack-dev | 02:52 | |
| *** blamar has joined #openstack-dev | 02:57 | |
| *** koolhead17 has quit IRC | 02:57 | |
| *** novas0x2a|laptop has joined #openstack-dev | 02:59 | |
| *** sunxin has quit IRC | 03:02 | |
| *** blamar has quit IRC | 03:09 | |
| *** bencherian has quit IRC | 03:10 | |
| *** Mandell has joined #openstack-dev | 03:10 | |
| *** blamar has joined #openstack-dev | 03:10 | |
| *** amotoki has joined #openstack-dev | 03:10 | |
| *** sunxin has joined #openstack-dev | 03:12 | |
| *** blamar has quit IRC | 03:18 | |
| *** blamar has joined #openstack-dev | 03:19 | |
| *** bcwaldon has quit IRC | 03:28 | |
| *** bcwaldon has joined #openstack-dev | 03:28 | |
| *** Dr_Who has quit IRC | 03:31 | |
| *** nati_ueno has joined #openstack-dev | 03:32 | |
| *** blamar has quit IRC | 03:34 | |
| *** nati_ueno has quit IRC | 03:35 | |
| *** blamar has joined #openstack-dev | 03:36 | |
| *** nati_ueno has joined #openstack-dev | 03:36 | |
| *** roge has quit IRC | 03:36 | |
| *** mdomsch has joined #openstack-dev | 03:39 | |
| *** blamar has quit IRC | 03:42 | |
| *** blamar has joined #openstack-dev | 03:43 | |
| *** blamar has quit IRC | 03:45 | |
| *** blamar has joined #openstack-dev | 03:46 | |
| *** blamar has joined #openstack-dev | 03:47 | |
| *** danwent has joined #openstack-dev | 03:47 | |
| *** blamar has quit IRC | 03:47 | |
| *** danwent has quit IRC | 03:50 | |
| *** lzyeval has quit IRC | 03:51 | |
| *** danwent has joined #openstack-dev | 03:53 | |
| *** jaypipes has quit IRC | 04:01 | |
| *** jakedahn is now known as jakedahn_zz | 04:15 | |
| *** jakedahn_zz is now known as jakedahn | 04:17 | |
| *** jakedahn is now known as jakedahn_zz | 04:20 | |
| *** nati_uen_ has joined #openstack-dev | 04:24 | |
| *** nati_uen_ has quit IRC | 04:24 | |
| *** nati_uen_ has joined #openstack-dev | 04:24 | |
| *** thingee is now known as thingee_zz | 04:25 | |
| *** nati_ueno has quit IRC | 04:27 | |
| *** mokas has quit IRC | 04:27 | |
| *** sacharya has quit IRC | 04:31 | |
| *** mokas has joined #openstack-dev | 04:37 | |
| *** Ryan_Lane has joined #openstack-dev | 04:48 | |
| *** nati_ueno has joined #openstack-dev | 05:12 | |
| *** nati_ueno has quit IRC | 05:12 | |
| *** nati_ueno has joined #openstack-dev | 05:13 | |
| *** anniec has joined #openstack-dev | 05:14 | |
| *** nati_uen_ has quit IRC | 05:15 | |
| *** mnewby has joined #openstack-dev | 05:30 | |
| *** mnewby has quit IRC | 05:30 | |
| *** spiffxp has joined #openstack-dev | 05:37 | |
| *** littleidea has joined #openstack-dev | 05:37 | |
| *** johnpur has joined #openstack-dev | 05:41 | |
| *** ChanServ sets mode: +v johnpur | 05:41 | |
| *** johnpur has quit IRC | 05:42 | |
| *** avishay has joined #openstack-dev | 05:45 | |
| *** hemna has joined #openstack-dev | 05:52 | |
| *** steveb_ has quit IRC | 05:54 | |
| *** nati_uen_ has joined #openstack-dev | 06:01 | |
| *** hattwick has quit IRC | 06:01 | |
| *** nati_ueno has quit IRC | 06:04 | |
| *** spiffxp has quit IRC | 06:05 | |
| *** sunxin has quit IRC | 06:08 | |
| *** nati_ueno has joined #openstack-dev | 06:09 | |
| *** nati_uen_ has quit IRC | 06:11 | |
| *** nati_ueno has quit IRC | 06:21 | |
| *** nati_ueno has joined #openstack-dev | 06:22 | |
| *** steveb_ has joined #openstack-dev | 06:38 | |
| *** steveb_ has quit IRC | 06:47 | |
| amotoki | danwent: hello! | 06:56 |
| danwent | hello | 06:57 |
| amotoki | I could not attend the IRC network meeting yesterday. | 06:57 |
| amotoki | In the meeting agenda, my patch about quantum keystone support in is listed. | 06:57 |
| amotoki | Is there any action to work on it? | 06:58 |
| *** xchu_ has quit IRC | 06:58 | |
| danwent | amotoki: i'm actually testing it out right now... | 06:58 |
| danwent | but ran into an unrelated issue | 06:59 |
| danwent | i should be able to +2 it tonight | 06:59 |
| amotoki | thanks. | 06:59 |
| danwent | then we'll need to ask dean or some other devstack core to approve it. | 06:59 |
| amotoki | I understand the situation. | 07:00 |
| amotoki | BTW, I have two question about subnet creation in v2 API. | 07:00 |
| danwent | ok, i have to run for a sec, but I will be back in 5 minutes. | 07:00 |
| danwent | feel free to ask, and I will respond then | 07:00 |
| amotoki | In API spec on wiki, we can specify subnet info in create_network(). Does it plan to be supported in F-3? | 07:02 |
| amotoki | In horizon work, I plan to specify network and subnet info in a single operation. So I am interested in the status of create_network(). | 07:04 |
| amotoki | The other is related to multiplen subnets for a network. | 07:05 |
| amotoki | In the current implementation two ipv4 subnets can be created on a network. But when create_port() only one ipv4 address is assigned for a port. | 07:06 |
| amotoki | Is it a bug? I think is is confusing for a user. | 07:06 |
| *** nati_uen_ has joined #openstack-dev | 07:10 | |
| *** reidrac has joined #openstack-dev | 07:12 | |
| *** m4xmr has joined #openstack-dev | 07:12 | |
| *** m4xmr has left #openstack-dev | 07:12 | |
| danwent | on first question | 07:13 |
| amotoki | hi | 07:13 |
| *** nati_ueno has quit IRC | 07:13 | |
| danwent | we had originally planned on letting someone specify a list of full subnet dictionary values during network create, and quantum would automatically create the subnets on their behalf. | 07:13 |
| danwent | we never implemented that, and currently, there's no bug to do so. | 07:14 |
| danwent | though I think from a UI perspective, that's the right workflow | 07:14 |
| danwent | but we can always have the UI make it a single step, which breaks down into multiple API calls | 07:14 |
| *** shang has quit IRC | 07:14 | |
| amotoki | My current implementation call multiple API calls. | 07:14 |
| danwent | i think that's fine, personally | 07:15 |
| danwent | on the second question | 07:15 |
| *** erikzaadi has joined #openstack-dev | 07:15 | |
| danwent | if no list of fixed_ips is specified, the port gets only a single IP. | 07:16 |
| danwent | if we want the port to get an IP on each subnet, we could specify a fixed_ip list of [ { subnet_id: XXX}, { subnet_id: YYY} ] | 07:16 |
| danwent | and it would be allocated an IP on both subnets. | 07:17 |
| *** alex88 has joined #openstack-dev | 07:17 | |
| *** alex88 has joined #openstack-dev | 07:17 | |
| danwent | the reason it was designed this way is that the common use case for having multiple subnets for a network is that you run out of space on a subnet (e.g., a public network). In this scenario, you really just want the port to get a single avaialble IP, not an IP from each subnet. | 07:18 |
| amotoki | I see. That sounds reasonable. | 07:18 |
| danwent | both very good questions, thanks for asking | 07:19 |
| amotoki | It sounds multiple IP subnet pools for a single network | 07:19 |
| amotoki | Thanks very much. I am happy to know the backgrond of the design. | 07:20 |
| danwent | btw, gabriel said that things are coming along well on the horizon work. I'm very happy to hear that :) | 07:20 |
| amotoki | i will send a status update on Horizon work. I think it is comming near review ready. | 07:22 |
| *** davidha has quit IRC | 07:28 | |
| *** nati_uen_ has quit IRC | 07:29 | |
| *** nati_ueno has joined #openstack-dev | 07:29 | |
| *** EmilienM has joined #openstack-dev | 07:29 | |
| *** Ryan_Lane has quit IRC | 07:40 | |
| *** dachary has joined #openstack-dev | 07:48 | |
| *** mancdaz has quit IRC | 07:52 | |
| *** mancdaz has joined #openstack-dev | 07:55 | |
| *** sulochan has joined #openstack-dev | 07:55 | |
| alex88 | hello, someone tried openstack with kernel > 3.4? i've tried with that to get O_DIRECT working with fuse, and in my configuration, vm weren't getting the ip, rolling back to 3.2 works fine. | 07:57 |
| *** sulochan has quit IRC | 07:57 | |
| *** sulochan has joined #openstack-dev | 08:01 | |
| *** darraghb has joined #openstack-dev | 08:01 | |
| *** nati_ueno has quit IRC | 08:04 | |
| *** Mandell has quit IRC | 08:06 | |
| *** pixelbeat has joined #openstack-dev | 08:17 | |
| *** erikzaadi has quit IRC | 08:24 | |
| *** kzhen has left #openstack-dev | 08:31 | |
| *** derekh has joined #openstack-dev | 08:34 | |
| *** hattwick has joined #openstack-dev | 08:43 | |
| *** derekh has quit IRC | 08:48 | |
| *** davidha has joined #openstack-dev | 09:08 | |
| *** eglynn has joined #openstack-dev | 09:15 | |
| *** johngarbutt has quit IRC | 09:17 | |
| *** derekh has joined #openstack-dev | 09:18 | |
| *** derekh has quit IRC | 09:21 | |
| *** derekh has joined #openstack-dev | 09:22 | |
| *** troytoman-away has quit IRC | 09:28 | |
| *** troytoman-away has joined #openstack-dev | 09:31 | |
| *** mcaway is now known as markmc | 09:31 | |
| *** dolphm_ has joined #openstack-dev | 09:33 | |
| *** dolphm has quit IRC | 09:36 | |
| *** halfss has joined #openstack-dev | 10:03 | |
| *** apevec has joined #openstack-dev | 10:04 | |
| *** cyberbob has joined #openstack-dev | 10:14 | |
| *** halfss has quit IRC | 10:24 | |
| *** danwent has quit IRC | 10:30 | |
| *** zhuadl has quit IRC | 10:31 | |
| *** alexpilotti has joined #openstack-dev | 10:59 | |
| *** mancdaz has quit IRC | 11:06 | |
| *** GheRivero has quit IRC | 11:09 | |
| *** alexpilotti has quit IRC | 11:10 | |
| *** troytoman-away is now known as troytoman | 11:16 | |
| *** rods has joined #openstack-dev | 11:16 | |
| *** GheRivero has joined #openstack-dev | 11:22 | |
| *** cyberbob has quit IRC | 11:23 | |
| *** sunxin has joined #openstack-dev | 11:27 | |
| *** wiliam has joined #openstack-dev | 11:30 | |
| *** gargya has joined #openstack-dev | 11:47 | |
| *** chrisfer has joined #openstack-dev | 11:47 | |
| *** huats has quit IRC | 11:54 | |
| *** maurosr has joined #openstack-dev | 11:57 | |
| *** davidha has quit IRC | 12:07 | |
| *** davidha has joined #openstack-dev | 12:07 | |
| *** rkukura has joined #openstack-dev | 12:14 | |
| *** sandywalsh has joined #openstack-dev | 12:14 | |
| *** zhuadl has joined #openstack-dev | 12:22 | |
| *** cloudvirt has joined #openstack-dev | 12:23 | |
| *** Shrews has joined #openstack-dev | 12:24 | |
| *** dprince has joined #openstack-dev | 12:31 | |
| *** lts has joined #openstack-dev | 12:35 | |
| *** salgado has joined #openstack-dev | 12:38 | |
| *** salgado has joined #openstack-dev | 12:38 | |
| *** huats has joined #openstack-dev | 12:40 | |
| *** huats has joined #openstack-dev | 12:40 | |
| *** sacharya has joined #openstack-dev | 12:41 | |
| *** sunxin has quit IRC | 12:42 | |
| *** sunxin has joined #openstack-dev | 12:44 | |
| *** lorin1 has joined #openstack-dev | 12:45 | |
| dprince | ayoung: I assigned https://bugs.launchpad.net/keystone/+bug/1031022 to you sir. | 12:47 |
| uvirtbot | Launchpad bug 1031022 in keystone "update auth_token to default signing_dir w/ os USER as suffix" [High,In progress] | 12:47 |
| *** GheRivero has quit IRC | 12:54 | |
| *** sacharya has quit IRC | 12:55 | |
| *** timello has joined #openstack-dev | 12:58 | |
| *** glenc_ has joined #openstack-dev | 13:00 | |
| *** shang has joined #openstack-dev | 13:02 | |
| *** glenc has quit IRC | 13:03 | |
| *** thovden has joined #openstack-dev | 13:14 | |
| *** tgall_foo has joined #openstack-dev | 13:15 | |
| *** roge has joined #openstack-dev | 13:15 | |
| *** ayoung has quit IRC | 13:16 | |
| *** ayoung has joined #openstack-dev | 13:17 | |
| eglynn | russellb: if you get a chance, can you give https://review.openstack.org/10130 another look? (corresponding tempest change has landed) | 13:19 |
| *** kbringard has joined #openstack-dev | 13:20 | |
| *** markmcclain has joined #openstack-dev | 13:20 | |
| *** thovden has quit IRC | 13:22 | |
| *** halfss has joined #openstack-dev | 13:23 | |
| asalkeld | dhellmann, you up and about? | 13:29 |
| *** sacharya has joined #openstack-dev | 13:29 | |
| asalkeld | (about to head off to bed - thought I'd catch you ...) | 13:29 |
| *** garyk has quit IRC | 13:30 | |
| mtaylor | morning chmouel | 13:30 |
| chmouel | mtaylor: hello | 13:30 |
| *** flaviamissi has joined #openstack-dev | 13:30 | |
| mtaylor | chmouel: sorry for using your email to complain :) | 13:30 |
| chmouel | mtaylor: ah no I totally agree with you heh :) | 13:30 |
| mtaylor | chmouel: hehe. good! :) | 13:31 |
| mtaylor | chmouel: I sent you a pkg_resources-based patch (it's not a large patch) | 13:31 |
| chmouel | mtaylor: yeah saw that it's missing a import :) | 13:32 |
| mtaylor | chmouel: BAH. imports should import themselves! | 13:32 |
| chmouel | heh | 13:33 |
| mtaylor | me hides | 13:33 |
| *** koolhead17 has joined #openstack-dev | 13:33 | |
| chmouel | but I definitively like the idea | 13:33 |
| mtaylor | chmouel: I can make a POC plugin that uses it to install too if you like | 13:33 |
| chmouel | I am trying to find a way so HP/RAX can provide a auth_url endpoint directly in plugin without having the user needing to specify it | 13:34 |
| mtaylor | chmouel: ++ | 13:34 |
| dansmith | eglynn: I like it! thanks for that.. I was actually punting on how that got reported (see my other comments).. I'll work on that change | 13:34 |
| chmouel | mtaylor: for the plugin yes please if you can do this quickly | 13:34 |
| eglynn | dansmith: cool, thanks! | 13:35 |
| chmouel | mtaylor: i mean if that's not something that would take you too much time | 13:35 |
| mtaylor | chmouel: and I think, different from random env vars in the code, we _do_ have a hisory of having plugins in the tree... | 13:38 |
| mtaylor | chmouel: it'll just be a few minutes... | 13:38 |
| *** dolphm_ has quit IRC | 13:38 | |
| chmouel | mtaylor: you mean having the rax/hp auth plugin directly in tree ? i guess this would be then needs to be in openstack-common as this would be shared across | 13:39 |
| * chmouel think his english is getting worse but he's multitasking by watching france-sweden women's handball at the same time | 13:41 | |
| *** maoy has quit IRC | 13:41 | |
| *** zhuadl has quit IRC | 13:42 | |
| *** alex88 has quit IRC | 13:43 | |
| *** matiu has joined #openstack-dev | 13:45 | |
| mtaylor | chmouel: https://github.com/emonty/python-novaclient/commit/6af2c914c3eb40bfcd6d35286bf691d8a7c5f4ad | 13:53 |
| mtaylor | chmouel: ah, hadn't thought about htat - yeah, inside of novaclient would be silly :) | 13:54 |
| mtaylor | chmouel: I stuck it in tree for now for sake of argument | 13:54 |
| chmouel | yeah | 13:54 |
| dhellmann | asalkeld: looks like we're on different schedules, my day is just starting | 13:54 |
| mtaylor | chmouel: in fact, ooh! instead of my original patch to remove the plugin, we could make one patch first which moves rax auth to a plugin but would not break end users | 13:55 |
| dhellmann | asalkeld: I'll reply to your emails | 13:55 |
| mtaylor | chmouel: and then we could work on making an externally installable version of the plugin with the same results, and then remove it from the tree | 13:55 |
| mtaylor | chmouel: so that people depending on the current behavior don't get screwed | 13:55 |
| chmouel | well the nova under rax cloud has just being released today | 13:56 |
| chmouel | so it doesn't matter to break compatibility | 13:56 |
| mtaylor | I've been using that nova for openstack build farm for 6 months | 13:56 |
| * mtaylor points out he was submitting a patch that was going to make his own life hell | 13:57 | |
| chmouel | yeah | 13:57 |
| chmouel | was going to says that :) | 13:57 |
| *** glenc has joined #openstack-dev | 13:58 | |
| chmouel | we would just need to have people move from env varibale from NOVA_AUTH_RAX to OS_AUTH_SYSTEM=rackspace2_0 (or whatever it's called) | 13:58 |
| mtaylor | chmouel: I like your idea about having plugins be able to provide the auth_url too ... that actually makes it nicer than the previous scheme, where you had to know the url, but you also had to know to do NOVA_RAX_AUTH | 13:58 |
| mtaylor | yea | 13:58 |
| chmouel | it should not be too hard to handle NOVA_RAX_AUTH and auth_url would always be overridable | 13:59 |
| mtaylor | ++ | 13:59 |
| chmouel | I can package all this and works with your patches and propose for a review if you don't mind | 14:00 |
| mtaylor | sounds great to me | 14:00 |
| chmouel | keep everything in python-novaclient for now | 14:00 |
| chmouel | and would move them to openstack.common when i'll get that in swiftclient after | 14:01 |
| mtaylor | yeah. it's a smoother starting point, I think | 14:01 |
| mtaylor | yeah | 14:01 |
| *** glenc_ has quit IRC | 14:01 | |
| mtaylor | well, moving to openstack.common will be hard before openstack.common becomes a library... | 14:01 |
| chmouel | yeah, it sucks we probably would have to duplicate until we have library | 14:02 |
| *** mnewby has joined #openstack-dev | 14:05 | |
| *** PotHix has joined #openstack-dev | 14:05 | |
| *** mdomsch has quit IRC | 14:08 | |
| mtaylor | chmouel: actually, that might get problematic with how entry points work. we might need to either just go ahead and make external hp and rackspace auth plugin packages | 14:12 |
| mtaylor | chmouel: or make an openstack-auth-plugins package | 14:13 |
| chmouel | personally I would say that should be the job from the rax/hp to do that | 14:13 |
| *** Shrews has quit IRC | 14:13 | |
| mtaylor | totally | 14:13 |
| mtaylor | as soon as we're happy with the entrypoints interface we want to use, it should be easy to hand that off - I'm happy to write the first version of code for each of them :) | 14:14 |
| *** ayoung has quit IRC | 14:14 | |
| chmouel | I don't think we need more authenticate | 14:16 |
| chmouel | maybe a post or a pre_authenticate hook for some use case | 14:16 |
| *** edygarcia has joined #openstack-dev | 14:16 | |
| *** ayoung has joined #openstack-dev | 14:18 | |
| *** dolphm has joined #openstack-dev | 14:19 | |
| *** mnewby has quit IRC | 14:21 | |
| *** zhuadl has joined #openstack-dev | 14:21 | |
| *** andrewbogott_afk is now known as andrewbogott | 14:21 | |
| *** mnewby has joined #openstack-dev | 14:23 | |
| *** Shrews has joined #openstack-dev | 14:23 | |
| mtaylor | kk | 14:25 |
| *** AlanClark has joined #openstack-dev | 14:25 | |
| mtaylor | chmouel: so, this may be getting ahead of ourselves, but setuptools made a naming convention for setuptools plugins that they asked people to follow - so the git plugin is "setuptools-git" | 14:26 |
| mtaylor | chmouel: I think we don't want these to start with openstack-, because that's what all of our official stuff does... what about "hp-auth-openstack" and "rackspace-auth-openstack" as the python packages | 14:27 |
| chmouel | yeah i like that | 14:27 |
| *** rnirmal has joined #openstack-dev | 14:28 | |
| chmouel | I think people would take whatever we will give them tbh | 14:28 |
| *** mnewby has quit IRC | 14:29 | |
| mtaylor | yup | 14:29 |
| *** mnewby has joined #openstack-dev | 14:29 | |
| *** amotoki has quit IRC | 14:31 | |
| *** cloudvirt has quit IRC | 14:33 | |
| *** maurosr is now known as maurosr_meeting | 14:33 | |
| *** maoy has joined #openstack-dev | 14:33 | |
| *** maurosr_meeting is now known as maurosr | 14:33 | |
| *** cloudvirt has joined #openstack-dev | 14:33 | |
| ayoung | dprince, I think you missed the point on https://review.openstack.org/#/c/10627/ | 14:36 |
| ayoung | Obviosuly, the right value for the cache dir is in /var/cache | 14:36 |
| ayoung | but that will not work for development, etc. The default value is one that will work, securely, if left unset | 14:36 |
| *** Shrews has quit IRC | 14:39 | |
| dprince | ayoung: Okay. I think that should be fine then. | 14:41 |
| *** Shrews has joined #openstack-dev | 14:41 | |
| ayoung | dprince, thanks. | 14:41 |
| dprince | ayoung: looks like it failed for SmokeStack though. Any idea why? | 14:42 |
| ayoung | dprince, no. It might be a permissions issue | 14:42 |
| ayoung | I just triggered a recheck | 14:42 |
| dprince | ayoung: My user for keystone uses /var/lib/keystone. | 14:42 |
| ayoung | lets see if it happens again. | 14:42 |
| dprince | as HOM | 14:42 |
| dprince | HOME | 14:42 |
| ayoung | dprince, what sets that up? | 14:42 |
| dprince | packages. | 14:42 |
| *** dspano has joined #openstack-dev | 14:43 | |
| ayoung | well, it probably is close enough to correct that we can use it. TEchnically, that should be for files generated by your user, and /var/cache if for cached files, but I think I am OK with not splitting that hair. THe thing is, we need a dir for every service that usese Keystone | 14:43 |
| ayoung | Keystone itself is actually the exception | 14:44 |
| ayoung | as it reads the files out of /etc/keystone/ | 14:44 |
| dprince | Yep. | 14:44 |
| ayoung | dprince, torpedo error was Error: test_get(Torpedo::Compute::Flavors) | 14:45 |
| ayoung | OpenStack::Compute::Exception::Connection: Unable to connect to localhost | 14:45 |
| ayoung | /usr/local/share/gems/gems/openstack-compute-1.1.10/lib/openstack/compute/connection.rb:333:in `rescue in start_http' | 14:45 |
| *** guitarzan_ has quit IRC | 14:46 | |
| *** Gordonz has joined #openstack-dev | 14:47 | |
| ayoung | dprince, my concern is that http://smokestack.openstack.org/?go=/jobs/27143 error is coming from trying to create or read a dir in $HOME | 14:47 |
| dprince | ayoung: Doesn't auth_token try to create it though? | 14:47 |
| ayoung | dprince, yes, and if it fails, it would explain the error. But I thought I would see it in the log | 14:48 |
| ayoung | 2012-07-31 20:55:12 INFO keystone.middleware.auth_token [-] Starting keystone auth_token middleware | 14:48 |
| ayoung | looks OK to me | 14:48 |
| dprince | ayoung: Yeah. Well. Keep in mind *all* SmokeStack runs started failing yesterday due to a Nova bug just fixed this morning. So that could have been the issue. | 14:49 |
| dprince | ayoung: let me have a look at it... | 14:49 |
| mtaylor | chmouel: https://github.com/emonty/rackspace-auth-openstack <-- there it is split out into its own repo | 14:50 |
| *** timello has quit IRC | 14:50 | |
| bcwaldon | dprince: can you run through stuart's patches today? | 14:50 |
| bcwaldon | dprince: should be super easy | 14:50 |
| dprince | bcwaldon: yes. | 14:51 |
| bcwaldon | dprince: and you've got comments on your next-to-last swift storage patch | 14:52 |
| *** alex88 has joined #openstack-dev | 14:52 | |
| bcwaldon | dprince: I would love to get it in today | 14:52 |
| *** cloudvirt has quit IRC | 14:53 | |
| chmouel | mtaylor: cool, so I have fixed a bit your load_plugin in https://github.com/chmouel/python-novaclient so we basically need to remove the setup.py entry_point part from novaclient? | 14:53 |
| *** sunxin has quit IRC | 14:54 | |
| mtaylor | chmouel: yup | 14:54 |
| mtaylor | chmouel: well, the plugin part, but yeah | 14:54 |
| chmouel | mtaylor: cool, will remove that | 14:55 |
| mtaylor | chmouel: and for testing before we publish, just doing "python setup.py develop" in the rackspace-auth-openstack should get things install right - I'll grab your tree and test it out locally when you've done that | 14:55 |
| *** sunxin has joined #openstack-dev | 14:55 | |
| chmouel | mtaylor: cool thanks | 14:56 |
| chmouel | mtaylor: I have tested already with my latest github branch and this works perfectly now let's see with your package | 14:56 |
| mtaylor | chmouel: cool - did you add an env var yet? | 14:58 |
| *** datsun180b has joined #openstack-dev | 14:58 | |
| mtaylor | like, does OS_AUTH_SYSTEM work? | 14:58 |
| chmouel | mtaylor: yeah | 14:58 |
| *** hemna has quit IRC | 14:58 | |
| mtaylor | chmouel: w00t! | 14:58 |
| mtaylor | chmouel: it worked for me | 14:58 |
| chmouel | hey hey lucky :) | 14:59 |
| *** alex88 has quit IRC | 14:59 | |
| chmouel | mtaylor: btw: you want to make sure to use --no_cache | 15:00 |
| chmouel | mtaylor: it has been bothering me with my tests | 15:01 |
| *** salgado has quit IRC | 15:02 | |
| mtaylor | chmouel: ok. cool | 15:03 |
| mtaylor | chmouel: btw- we might need to refactor something more... | 15:03 |
| mtaylor | chmouel: we're currently requiring that this plugin call cls._authenticate which is a private method | 15:03 |
| mtaylor | chmouel: we might want to make there be an actual api for this plugin to call to :) | 15:03 |
| *** reidrac has quit IRC | 15:04 | |
| chmouel | mtaylor: yeah agreed... it is such a fine line between private and public methos in python :) | 15:04 |
| mtaylor | ++ | 15:04 |
| *** nunosantos has joined #openstack-dev | 15:05 | |
| *** salgado has joined #openstack-dev | 15:06 | |
| *** salgado has joined #openstack-dev | 15:06 | |
| *** jaypipes has joined #openstack-dev | 15:07 | |
| *** mdomsch has joined #openstack-dev | 15:09 | |
| *** dubsquared has joined #openstack-dev | 15:09 | |
| *** Gordonz has quit IRC | 15:10 | |
| *** Gordonz has joined #openstack-dev | 15:11 | |
| *** winston-d has joined #openstack-dev | 15:13 | |
| *** heckj has joined #openstack-dev | 15:13 | |
| *** thingee_zz is now known as thingee | 15:14 | |
| *** cloudvirt has joined #openstack-dev | 15:15 | |
| *** andrewbogott is now known as andrewbogott_afk | 15:17 | |
| *** e1mer has quit IRC | 15:19 | |
| *** alex88 has joined #openstack-dev | 15:20 | |
| *** andrewbogott_afk is now known as andrewbogott | 15:21 | |
| mtaylor | chmouel: one more pull req your way - ability to leave auth_url blank if you're using an auth_system plugin | 15:24 |
| *** jtran has joined #openstack-dev | 15:24 | |
| chmouel | mtaylor: awesome i was just about to test that | 15:24 |
| chmouel | mtaylor: using your repo and remoing the entry_point from novaclient works perfectly | 15:25 |
| mtaylor | chmouel: yah. works for me too | 15:25 |
| dprince | ayoung: So the reason https://review.openstack.org/#/c/10627/ fails for me is because nova-api failed to startup due to a permissions issue. For Nova we set $HOME to /home/nova via Fedora packages... but we don't actually create /home/nova. So fail. | 15:25 |
| mtaylor | chmouel: do you think I should just publish that repo to pypi myself and then hand the keys to someone later? | 15:25 |
| *** andrewbogott is now known as andrewbogott_afk | 15:26 | |
| *** nati_ueno has joined #openstack-dev | 15:26 | |
| *** sunxin has quit IRC | 15:26 | |
| chmouel | mtaylor: yeah I think that reasonable | 15:26 |
| dprince | ayoung: I'm not sure offhand why we aren't using /var/lib/nova for HOME by default. So maybe that is the fix for my setup. Anyway. I left you some more feedback that logging the signing directory would be nice. Otherwise I think this looks good. | 15:27 |
| chmouel | mtaylor: from history I was owning the python-cloudfiles repo in pypi for the last 4 years and nobody wanted to take it from me :) so expect to keep it for a while :) | 15:27 |
| *** sulochan has quit IRC | 15:27 | |
| *** panpengjun has joined #openstack-dev | 15:27 | |
| *** danwent has joined #openstack-dev | 15:30 | |
| *** nati_ueno has quit IRC | 15:30 | |
| *** andrewbogott_afk is now known as andrewbogott | 15:33 | |
| *** gargya has quit IRC | 15:36 | |
| mtaylor | chmouel: hehe. ok. well, I've pushed up a small change to rackspace_auth_openstack and have also pushed up an hpcloud_auth_openstack repo | 15:43 |
| chmouel | mtaylor: cool, so for uk rax there is a twist | 15:43 |
| chmouel | mtaylor: more suckage to be honest but the auth_url is a change | 15:44 |
| mtaylor | chmouel: wow, really? | 15:44 |
| chmouel | mtaylor: yeah believe me i did yell on that a few years ago | 15:44 |
| *** zhuadl has quit IRC | 15:44 | |
| chmouel | mtaylor: but lost the fight | 15:44 |
| mtaylor | chmouel: should that be a different auth_system? | 15:44 |
| *** ayoung has quit IRC | 15:45 | |
| mtaylor | chmouel: or some other mechanism | 15:45 |
| chmouel | mtaylor: no it's exactly the same as the us but the endpoint url is starting by lon. | 15:45 |
| mtaylor | chmouel: wow. | 15:45 |
| mtaylor | chmouel: it's almost like we're one step away from making a general cloud provider plugin here ... ;) | 15:46 |
| chmouel | mtaylor: ahah we just made a new jcloud | 15:46 |
| mtaylor | chmouel: yay! :) | 15:46 |
| chmouel | mtaylor: I can make the uk plugin available | 15:47 |
| chmouel | mtaylor: as it's not really a big change and would not expect the interface would change anyway | 15:47 |
| *** halfss has quit IRC | 15:48 | |
| *** samkottler has joined #openstack-dev | 15:49 | |
| annegentle | bcwaldon: or jaypipes: do you know if all extensions in the Compute API are controlled by policy.json listings? Or are some controlled in code, others not pre-controlled for the user/admin? | 15:51 |
| mtaylor | chmouel: https://github.com/emonty/rackspace-auth-openstack/commit/8c9c3d0efe9520d4ec49a77f0082faf654b1db13 | 15:51 |
| mtaylor | chmouel: what about that? | 15:51 |
| jaypipes | annegentle: they should all be controlled by plicy.json by this point | 15:52 |
| chmouel | mtaylor: perfect! | 15:52 |
| jaypipes | annegentle: the issue I was referring to was just calling things "admin extensions" when they technically could be non-admin extensions by simply modiofying the policy.json file | 15:52 |
| annegentle | ok, at least in folsom, but who knows in essex? | 15:52 |
| chmouel | mtaylor: so people would just do pip install the_package_name;export OS_AUTH_SYSTEM=rackspace_$REGION and that would just work with their username/passwords | 15:52 |
| annegentle | jaypipes: yeah, totally get what you mean, just wasn't sure if the code reflects that, happy to change the docs though | 15:53 |
| annegentle | jaypipes: probably the docs should say "access controlled by policy.json" | 15:53 |
| mtaylor | chmouel: should do! | 15:53 |
| jaypipes | annegentle: ++ | 15:53 |
| chmouel | mtaylor: cool I have added a small fix to your last patch and will add tomorrow the tests and stuff and would give that for review/docs etc.. | 15:54 |
| chmouel | mtaylor: and get the RAX guys aware of updating their documentation | 15:54 |
| jaypipes | sdague: FYI, made excellent progress on the whitebox stuff last night... pushing a new patch for review shortly. | 15:54 |
| mtaylor | chmouel: awesome. I'll publish the plugins so that it will work as soon as your thing lands | 15:54 |
| *** samkottler has quit IRC | 15:55 | |
| *** danwent has quit IRC | 15:57 | |
| sdague | jaypipes: awesome | 15:58 |
| *** dprince has quit IRC | 15:59 | |
| *** alex88 has quit IRC | 15:59 | |
| mtaylor | chmouel: ok. I have pushed them both up to pypi | 16:01 |
| *** gargya has joined #openstack-dev | 16:03 | |
| *** issackelly has joined #openstack-dev | 16:03 | |
| *** jdurgin has joined #openstack-dev | 16:03 | |
| *** ayoung has joined #openstack-dev | 16:06 | |
| *** salgado is now known as salgado-lunch | 16:07 | |
| *** lloydde has joined #openstack-dev | 16:12 | |
| *** markvoelker has joined #openstack-dev | 16:13 | |
| *** sacharya has quit IRC | 16:17 | |
| *** matwood has joined #openstack-dev | 16:20 | |
| *** thovden has joined #openstack-dev | 16:23 | |
| *** spiffxp has joined #openstack-dev | 16:25 | |
| *** shang has quit IRC | 16:25 | |
| davidha | hey swifters, I wonder if people had noticed the drastic slowdown in ring loading from 2.6 to 2.7 (also takes 100% CPU while loading) | 16:26 |
| *** shang has joined #openstack-dev | 16:28 | |
| davidha | chmouel around? | 16:29 |
| davidha | Anyone from the swift core team? | 16:29 |
| *** derekh has quit IRC | 16:30 | |
| s34n | devstack should run just fine in a vm, right? | 16:31 |
| *** andrewsmedina has joined #openstack-dev | 16:31 | |
| *** bencherian has joined #openstack-dev | 16:32 | |
| *** mokas has quit IRC | 16:32 | |
| jaypipes | s34n: yup. | 16:34 |
| *** eglynn has quit IRC | 16:34 | |
| *** timello has joined #openstack-dev | 16:40 | |
| *** anniec has quit IRC | 16:47 | |
| *** shang_ has joined #openstack-dev | 16:52 | |
| *** mokas has joined #openstack-dev | 16:52 | |
| *** sacharya has joined #openstack-dev | 16:53 | |
| *** Mandell has joined #openstack-dev | 16:54 | |
| *** koolhead17 has quit IRC | 16:55 | |
| *** shang has quit IRC | 16:55 | |
| dolphm | heckj: are we still going forward with page / per_page, or sticking with marker / limit? | 16:56 |
| *** n0ano has joined #openstack-dev | 16:56 | |
| *** salgado-lunch is now known as salgado | 16:56 | |
| *** markmc is now known as mcaway | 16:57 | |
| *** hemna has joined #openstack-dev | 16:57 | |
| *** dprince has joined #openstack-dev | 16:59 | |
| heckj | dolphm: I'm still on the page/per-page plan - was there greater consensus to not do that I missed while being distracted? | 17:00 |
| *** danwent has joined #openstack-dev | 17:00 | |
| *** nati_ueno has joined #openstack-dev | 17:00 | |
| *** nati_uen_ has joined #openstack-dev | 17:01 | |
| *** nati_uen_ is now known as nati_ueno_ | 17:02 | |
| *** Aaton_off is now known as Aaton | 17:03 | |
| *** nati_ueno has quit IRC | 17:04 | |
| *** gargya has quit IRC | 17:04 | |
| *** nati_ueno_ is now known as nati_ueno | 17:05 | |
| *** davidha has quit IRC | 17:07 | |
| *** davidha has joined #openstack-dev | 17:07 | |
| *** matwood_ has joined #openstack-dev | 17:09 | |
| *** johnpur has joined #openstack-dev | 17:09 | |
| *** ChanServ sets mode: +v johnpur | 17:09 | |
| *** matwood has quit IRC | 17:09 | |
| *** matwood_ is now known as matwood | 17:09 | |
| *** cloudvirt has quit IRC | 17:10 | |
| *** davidha has quit IRC | 17:13 | |
| *** Ryan_Lane has joined #openstack-dev | 17:13 | |
| *** jog0 has joined #openstack-dev | 17:13 | |
| *** anniec has joined #openstack-dev | 17:16 | |
| *** anniec_ has joined #openstack-dev | 17:16 | |
| *** sstent has quit IRC | 17:17 | |
| *** epim has joined #openstack-dev | 17:17 | |
| *** sstent has joined #openstack-dev | 17:18 | |
| dolphm | heckj: not necessarily -- seems like a fairly even divide on opinion from what i've seen | 17:18 |
| dolphm | heckj: i implemented page / per page for policy crud -- was just wondering if i should revise before posting for review | 17:18 |
| *** anniec has quit IRC | 17:20 | |
| *** anniec_ is now known as anniec | 17:20 | |
| heckj | dolphm: I don't think you need to - with the V3 api, the desire from the folks who're using it (horizon) are page/per-page | 17:20 |
| heckj | (dolphm: welcome back from vaca!) | 17:20 |
| *** matwood has quit IRC | 17:21 | |
| *** matwood has joined #openstack-dev | 17:25 | |
| *** thovden has quit IRC | 17:27 | |
| *** cloudvirt has joined #openstack-dev | 17:29 | |
| ayoung | heckj, dolphm so which one of you guys are going to implement the LDAP paging support? | 17:29 |
| heckj | ayoung: ahh! there's the kicker | 17:32 |
| ayoung | heckj, yep | 17:32 |
| ayoung | and, I think, unsolvable. | 17:32 |
| ayoung | As different LDAP servers do paging different ways | 17:33 |
| ayoung | heckj, so, what I suggest is that we fake it | 17:33 |
| Ryan_Lane | well, there should be an rfc for a standard way | 17:33 |
| Ryan_Lane | and we could check for the standard support by oid | 17:33 |
| ayoung | Ryan_Lane the problem is that LDAP does not specify the order of results | 17:34 |
| ayoung | and so in order to do paging, you need to have the equivalent of a cursor | 17:34 |
| ayoung | which is different for different LDAP servers | 17:34 |
| ayoung | created differently | 17:34 |
| Ryan_Lane | http://tools.ietf.org/html/rfc2696 | 17:35 |
| *** Mandell_ has joined #openstack-dev | 17:35 | |
| ayoung | Ryan_Lane, and, since the writers are from Microsoft, I am guessing that is what Active Directory implements | 17:35 |
| Ryan_Lane | likely | 17:36 |
| Ryan_Lane | it's supported in openldap too | 17:36 |
| Ryan_Lane | and opendj | 17:36 |
| ayoung | Ryan_Lane, the thing is, we probably should not use it, if I understand correctly | 17:36 |
| ayoung | bascially, it puts a huge load on the DS | 17:37 |
| ayoung | unlike SQL, which has offset/limit | 17:37 |
| ayoung | you need to keep the cursor alive in order to get paging. Which means you have state maintained on both the DS and the Keystone side | 17:37 |
| ayoung | The only way to know when to free it up is session time out | 17:37 |
| * Ryan_Lane nods | 17:38 | |
| Ryan_Lane | hm. I had to head off. have a meeting | 17:38 |
| *** Mandell has quit IRC | 17:38 | |
| *** Ryan_Lane has quit IRC | 17:39 | |
| *** anniec has quit IRC | 17:40 | |
| *** japage has quit IRC | 17:47 | |
| zul | bcwaldon: ping | 17:49 |
| *** andrewbogott is now known as andrewbogott_afk | 17:50 | |
| zul | bcwaldon: looks like your tidying up of nova.images.glance caused a regression in s3 | 17:51 |
| *** reed has joined #openstack-dev | 17:53 | |
| *** blamar has joined #openstack-dev | 17:54 | |
| *** panpengjun has left #openstack-dev | 17:54 | |
| *** Mandell_ has quit IRC | 17:57 | |
| *** Mandell has joined #openstack-dev | 17:57 | |
| *** littleidea has quit IRC | 17:58 | |
| *** littleidea has joined #openstack-dev | 18:00 | |
| bcwaldon | zul: oh no! | 18:00 |
| bcwaldon | zul: looks like we're missing tests | 18:00 |
| *** AlanClark has quit IRC | 18:00 | |
| zul | bcwaldon: when features got removed then s3 stopped working...getting tracebacks http://pastebin.ubuntu.com/1123587/ | 18:01 |
| *** apevec has quit IRC | 18:02 | |
| bcwaldon | zul: ah, ok, I'll fix it | 18:04 |
| *** utlemming has quit IRC | 18:04 | |
| zul | bcwaldon: i got a commit here locally that re-adding works for me, if you want me to submit that im ok with | 18:05 |
| bcwaldon | zul: sure, push it up | 18:05 |
| bcwaldon | zul: we need to write a test that exercises that code, though | 18:05 |
| zul | bcwaldon: sure np | 18:05 |
| zul | bcwaldon: yeah | 18:05 |
| bcwaldon | since I shouldnt have been allowed to break that :) | 18:05 |
| *** Adri2000 has quit IRC | 18:07 | |
| *** utlemming has joined #openstack-dev | 18:07 | |
| *** Adri2000 has joined #openstack-dev | 18:07 | |
| *** zaitcev has joined #openstack-dev | 18:10 | |
| zul | bcwaldon: no worries s3 is kind of screwy | 18:10 |
| *** gabrielhurley has joined #openstack-dev | 18:16 | |
| *** gabrielhurley has quit IRC | 18:16 | |
| *** dhellmann has quit IRC | 18:17 | |
| *** negronjl has quit IRC | 18:19 | |
| *** darraghb has quit IRC | 18:21 | |
| *** maoy_ has joined #openstack-dev | 18:21 | |
| *** maoy_ has quit IRC | 18:22 | |
| *** maoy has quit IRC | 18:22 | |
| *** maoy_ has joined #openstack-dev | 18:22 | |
| *** heckj has quit IRC | 18:22 | |
| *** blamar has quit IRC | 18:23 | |
| *** blamar has joined #openstack-dev | 18:23 | |
| *** johnpostlethwait has joined #openstack-dev | 18:23 | |
| *** johnpostlethwait has quit IRC | 18:24 | |
| vishy | mtaylor: what is it going to take to get stable-diablo gating working again? | 18:25 |
| vishy | mtaylor: maybe we should just turn it off so we can merge the security fixes that are in the queue? | 18:25 |
| vishy | jeblair, LinuxJedi: ^^ | 18:25 |
| *** andrewbogott_afk is now known as andrewbogott | 18:26 | |
| *** negronjl has joined #openstack-dev | 18:26 | |
| *** eglynn has joined #openstack-dev | 18:27 | |
| jeblair | vishy: as far as i am aware, the issue is that the tests no longer pass. | 18:28 |
| jeblair | vishy: (I'm reading latest on change 9093 to get caught up) | 18:28 |
| vishy | jeblair: that is correct, there is some env / tox weirdness | 18:29 |
| vishy | russellb: what was berrange's irc nick again? | 18:30 |
| russellb | danpb | 18:30 |
| russellb | guess he's out for the day (UK) | 18:31 |
| vishy | russellb: he's not online. I'm trying to find someone who knows if there has been any work done on passing a dev through virtio so we can actually control where the guest puts it | 18:32 |
| russellb | could try the ML, dan will see it there | 18:32 |
| mtaylor | vishy: yeah - it's that the tests don't pass - and as best I can tell it's related to dependencies - I believe I tried running the tests with run_tests.sh as well and that also did not work | 18:33 |
| dansmith | vishy: you mean other than udev rules in the guest? | 18:33 |
| jeblair | jtran: ping | 18:33 |
| vishy | dansmith: yeah, xen manages it by passing the data somehow, I figure they are using a plug and play vendor id or something | 18:33 |
| jtran | jeblair, hi | 18:34 |
| jeblair | jtran: i was looking at your latest comment on https://review.openstack.org/#/c/9093/ | 18:34 |
| vishy | dansmith: perhaps someone needs to write a fancy udev rule that can put the volume at the right location based on the metadata server | 18:34 |
| dansmith | does xen use virtio now? my xen knowledge is really old, but they used to be able to do it because they passed xenbus info in and that driver handled registration | 18:34 |
| jeblair | jenkins doesn't run devstack when running unit tests | 18:34 |
| jtran | ok, so that's why me trying to reproduce it even tho passing , gate still fails? | 18:35 |
| jeblair | jtran: adapting the commands you pasted, i think it should be more like this: http://paste.openstack.org/show/19905/ | 18:35 |
| jeblair | mtaylor: ^ double check me | 18:35 |
| mtaylor | jeblair: yup | 18:36 |
| dansmith | vishy: I think virtio must pass the index in, otherwise I'm not sure how the current behavior of not being able to detach without acpiphp loaded would be possible | 18:36 |
| jtran | jeblair, would need to git checkout stable/diablo tho right | 18:36 |
| jeblair | jtran: wouldn't hurt, but i think the git-review cmd will do the right thing | 18:37 |
| jtran | jeblair, i'll spin up a clean oneiric vm and try that out | 18:37 |
| vishy | dansmith: it passes an index, but that means nothing to udev and is solely based on the order in the xml | 18:37 |
| vishy | dansmith: it happily ignores the device name that you specify | 18:37 |
| dansmith | I know it does now, but I thought the index would be the vd{a-z} index, since libvirt uses that to generate a unique id (i.e. virtio3 for vdc) | 18:38 |
| dansmith | vishy: but after poking around in /sys, I suppose not :( | 18:42 |
| *** davidha has joined #openstack-dev | 18:42 | |
| vishy | dansmith: yeah :( | 18:43 |
| vishy | dansmith: magic udev rule required. Who wants to write it? | 18:44 |
| dansmith | vishy: you mean a magic rule to query out to openstack infrastructure to make the determination? | 18:44 |
| vishy | aye | 18:44 |
| dansmith | I think that'd be highly unconventional :) | 18:44 |
| *** novas0x2a|laptop has quit IRC | 18:45 | |
| vishy | dansmith: but at least when we get complaints we can say, hey install this udev rule! | 18:45 |
| vishy | dansmith: it would use the metadata server so it could work on amazon as well (although it already does because they use xen) | 18:45 |
| zykes- | udev rule for what ? | 18:46 |
| dansmith | I'm surprised that virtio doesn't expose it, but if there's a reason, I wonder if we could provide some more local channel.. | 18:46 |
| jog0 | vishy: I am trying to reproduce the sqlalchemy problem you found with my migration script | 18:47 |
| jog0 | vishy: what version of sqlalchemy was it 0.7.4? | 18:47 |
| vishy | i believe 0.7.4 | 18:49 |
| vishy | may have been .3 | 18:50 |
| jog0 | vishy: thanks just realized pypi only has 7.8, so running 'tox -efull' on Ubuntu 12.04 | 18:50 |
| vishy | jog0: it was whatever version is packaged in precise | 18:50 |
| *** johnpostlethwait has joined #openstack-dev | 18:51 | |
| *** danwent_ has joined #openstack-dev | 18:52 | |
| andrewsmedina | danwent_: hi | 18:52 |
| *** danwent has quit IRC | 18:55 | |
| *** danwent_ is now known as danwent | 18:55 | |
| danwent | andrewsmedina: hey, about to run off to a meeting, but can chat quickly | 18:57 |
| andrewsmedina | danwent: we talk later | 18:58 |
| andrewsmedina | danwent: call me when you return | 18:59 |
| danwent | k | 18:59 |
| *** spiffxp has quit IRC | 19:06 | |
| *** matwood has quit IRC | 19:09 | |
| *** andrewbogott is now known as andrewbogott_afk | 19:22 | |
| *** dhellmann has joined #openstack-dev | 19:22 | |
| *** ewindisch has quit IRC | 19:22 | |
| *** dubsquared has quit IRC | 19:28 | |
| *** jtran has quit IRC | 19:33 | |
| *** matwood has joined #openstack-dev | 19:38 | |
| *** matwood has quit IRC | 19:40 | |
| *** andrewbogott_afk is now known as andrewbogott | 19:44 | |
| *** ewindisch has joined #openstack-dev | 19:47 | |
| *** mdomsch has quit IRC | 19:49 | |
| jog0 | vishy: when I install python-sqlalchemy from 12.04 and run the 'tox -efull' I cannot even get teh unit tests to run. | 19:56 |
| *** AlanClark has joined #openstack-dev | 19:57 | |
| *** jtran has joined #openstack-dev | 19:58 | |
| vishy | hehe | 19:59 |
| vishy | jog0: you're probably missing other dependencies | 19:59 |
| vishy | jog0: fyi i was using ./run_tests.sh -N | 19:59 |
| jog0 | vishy: tox -efull installs missing deps | 20:02 |
| *** steveb_ has joined #openstack-dev | 20:03 | |
| *** novas0x2a|laptop has joined #openstack-dev | 20:03 | |
| danwent | andrewsmedina: ping, i'm back | 20:08 |
| *** Ryan_Lane has joined #openstack-dev | 20:11 | |
| andrewsmedina | danwent: hi | 20:11 |
| *** eglynn has quit IRC | 20:12 | |
| andrewsmedina | danwent: I will try to end support for xml up Monday | 20:12 |
| andrewsmedina | danwent: and | 20:12 |
| andrewsmedina | danwent: the fake plugin are working with api v2? | 20:13 |
| danwent | andrewsmedia, fake plugin was only for v2 | 20:15 |
| danwent | sorry v1 | 20:15 |
| danwent | but there is something similar for v2 | 20:15 |
| davidha | Hi, Anyone from the swift core team? | 20:15 |
| notmyname | danwent: what's up? | 20:15 |
| davidha | notmyname: Hi. | 20:16 |
| danwent | look in quantum.db.db_base_plugin_v2 | 20:16 |
| danwent | andrewsmedina: test_db_plugin.py uses this extensively, and all other v2 plugins are subclasses of it | 20:16 |
| danwent | notmyname: hey | 20:16 |
| davidha | I found a problem in Ring Loading under 2.7 | 20:17 |
| *** spiffxp has joined #openstack-dev | 20:17 | |
| danwent | notmyname: sorry, were you looking to chat with me about something? my above comments were directed at andrewsmedina | 20:17 |
| danwent | ah, you probably meant davidha, not danwent | 20:18 |
| notmyname | danwent: sorry. danwent v. davidha. silly tab autocomplete ;-) | 20:18 |
| danwent | got it | 20:18 |
| davidha | notmyname: I got confused also :) anyhow see ^^ | 20:18 |
| notmyname | davidha: what details do you have? | 20:18 |
| davidha | notmyname: Unpickling a 18 power ring is two orders of magnitude slower in 2.7 compared to 2.6 | 20:19 |
| notmyname | that's not good. do you have any ideas as to why? | 20:19 |
| davidha | I also found out why with the help of the #python | 20:19 |
| davidha | Yep - a change made in 2.7 | 20:20 |
| notmyname | do tell :-) | 20:20 |
| davidha | The change is meant to align it to 3.x | 20:20 |
| davidha | __reduce__ of array.array was changed | 20:20 |
| davidha | now instea dof a single BINSTRING it is many many many (did I say many already) BININTs | 20:21 |
| davidha | I think it is time to move away from pickle | 20:22 |
| davidha | :) | 20:22 |
| notmyname | is that something you could work on? | 20:23 |
| andrewsmedina | danwent: but the fake plugin doesnt has fields attr in the get_network method | 20:24 |
| andrewsmedina | danwent: and it is expected by v2 api | 20:24 |
| danwent | andrewsmedina: fake plugin does not work with v2, and will be removed | 20:24 |
| danwent | so you can just ignore it | 20:24 |
| davidha | Not right away. For now I can: 1. describe the problem in some details. 2. suggest an intermediate resolution for people using the system in production and 3. Suggest a direction for resolution | 20:25 |
| notmyname | davidha: that sounds great | 20:25 |
| davidha | notmyname: I will not be able to work on in Q3 for sure. | 20:25 |
| danwent | if you need a simple plugin for testing, I would use either the quantum.db.db_base_plugin_v2.QuantumDbPluginV2 | 20:26 |
| andrewsmedina | danwent: ok, thanks | 20:26 |
| davidha | ok. | 20:26 |
| *** wiliam has quit IRC | 20:26 | |
| danwent | andrewsmedina: or quantum.plugins.openvswitch.ovs_quantum_plugin.OVSQuantumPluginV2 | 20:26 |
| andrewsmedina | danwent: would be nice to create a issue to fix the fake plugin | 20:27 |
| *** EmilienM has left #openstack-dev | 20:27 | |
| davidha | notmyname, there is a mechanism for reloading the ring every X hours right? | 20:27 |
| notmyname | davidha: not exactly like that. the ring is reloaded based on its mtime (but not more that every X seconds, where X is set in a config file) | 20:28 |
| *** bcwaldon has quit IRC | 20:28 | |
| *** bcwaldon has joined #openstack-dev | 20:28 | |
| danwent | andrewsmedina: the fake plugin is v1 only, and all v1 code will be removed next week. is there something that you need to do with the fake plugin that can't be done by using quantum.db.db_base_plugin_v2.QuantumDbPluginV2 with a in-memory sqlite database? | 20:28 |
| *** bcwaldon has quit IRC | 20:28 | |
| davidha | ok, great - I thought it is measured in hours - this is what I found in the documentation.... | 20:29 |
| *** bcwaldon has joined #openstack-dev | 20:29 | |
| danwent | andrewsmedina: so far, the latter has been sufficient for us. | 20:29 |
| notmyname | davidha: ah, you may be referring to min_part_hours which is in hours and is the fastest a ring can be updated | 20:29 |
| notmyname | davidha: thanks for doing the detective work. even if someone else gets to the code before you, any info you give will be quite helpful | 20:29 |
| andrewsmedina | danwent: ok, you're right | 20:31 |
| andrewsmedina | danwent: thanks | 20:31 |
| danwent | andrewsmedina: no problem. thanks for working on this bug! | 20:31 |
| davidha | After finding that the startup became very slow, I was intrigued to find out that my old ring file loaded 100 times faster than the new ones I was creating - although they all had the same size.... | 20:32 |
| andrewsmedina | danwent: I intend to contribute more to the project | 20:32 |
| notmyname | davidha: where are you planning on documenting this? a bug report in launchpad is one possibility | 20:32 |
| davidha | I will do that and send an email for pepole to be aware | 20:33 |
| notmyname | davidha: thanks | 20:33 |
| davidha | Where are you now? | 20:33 |
| danwent | andrewsmedina: great. | 20:33 |
| davidha | :-X | 20:33 |
| notmyname | davidha: like my physical location? | 20:34 |
| davidha | Last we spoke you were heading to a new post | 20:34 |
| notmyname | davidha: ah, yes. I'm in San Francisco now, working at SwiftStack | 20:35 |
| davidha | Ok, sorry for being out of touch :) | 20:35 |
| notmyname | not too much. it's still quite new | 20:37 |
| zykes- | notmyname: is georeplication getting into swift? | 20:39 |
| *** dubsquared has joined #openstack-dev | 20:40 | |
| *** lorin1 has quit IRC | 20:41 | |
| notmyname | zykes-: I think (hope) someone will probably try to tackle it in the grizzly timeframe. there are a lot of people asking about it, but so far nobody has stepped up to figure it out and code it | 20:41 |
| *** lorin1 has joined #openstack-dev | 20:51 | |
| *** lloydde has quit IRC | 20:52 | |
| *** Mandell has quit IRC | 20:52 | |
| *** novas0x2a|laptop has quit IRC | 20:53 | |
| dolphm | ayoung: this is technically a blocker for policy crud impl https://review.openstack.org/#/c/10682/ | 20:54 |
| *** ewindisch has quit IRC | 20:56 | |
| ayoung | dolphm, OK. I'll review it. | 20:56 |
| *** AlanClark has quit IRC | 20:56 | |
| ayoung | dolphm, looks reasonable. So long as it passes the tests, I should be able to merge | 20:58 |
| ayoung | dolphm, the only change in the code base is to raise an exception. What is the justification for that? | 20:58 |
| dolphm | ayoung: test_get_endpoint_404 would fail otherwise (the rest of the driver raises explicit exceptions) | 21:00 |
| dolphm | ayoung: however, get_endpoint isn't exposed through to the rest API, so there's no client tests to check for a 404 -- (but the driver still needs to raise the appropriate exception, so i can catch it in the policy driver =) ) | 21:01 |
| *** lloydde has joined #openstack-dev | 21:01 | |
| *** cloudvirt has quit IRC | 21:01 | |
| ayoung | dolphm, OK. Looks good to me. Shouldn't break anything.... | 21:02 |
| zykes- | vishy: ping | 21:02 |
| *** lts has quit IRC | 21:02 | |
| ayoung | dolphm, quid pro quo: https://review.openstack.org/#/c/10544/ | 21:04 |
| jtran | jeblair, so i installed a fresh oneiric vm, using the exact sequence here: http://paste.openstack.org/show/19905/ , did not use devstack at all. Results : | 21:05 |
| jtran | py26: commands succeeded | 21:05 |
| jtran | congratulations :) | 21:05 |
| jtran | - py26: commands succeeded | 21:05 |
| jtran | Ran 1780 tests in 743.175s | 21:05 |
| dolphm | ayoung: revise keystone.sample.conf as well? | 21:06 |
| *** dprince has quit IRC | 21:06 | |
| ayoung | dolphm, good catch | 21:08 |
| *** cloudvirt has joined #openstack-dev | 21:08 | |
| dolphm | ayoung: ping me when it's patch and i'll +2 | 21:08 |
| *** Mandell has joined #openstack-dev | 21:08 | |
| *** maurosr has quit IRC | 21:11 | |
| *** novas0x2a|laptop has joined #openstack-dev | 21:13 | |
| *** ewindisch has joined #openstack-dev | 21:14 | |
| *** dolphm has quit IRC | 21:14 | |
| *** cloudvirt has quit IRC | 21:22 | |
| *** cloudvirt has joined #openstack-dev | 21:22 | |
| notmyname | zykes-: watch out. don't feed the trolls :-) | 21:24 |
| *** matwood has joined #openstack-dev | 21:29 | |
| *** winston-d has quit IRC | 21:29 | |
| *** markvoelker has quit IRC | 21:30 | |
| zykes- | vishy: in cinder will it be able to set some kind of reasonable default / policy for where volumes go for diff things ? | 21:30 |
| zykes- | if you didnt answer it already | 21:30 |
| *** lorin1 has quit IRC | 21:32 | |
| *** salgado has quit IRC | 21:33 | |
| jgriffith | zykes-: What do you mean by that? | 21:34 |
| jgriffith | zykes-: You mean something to use a sort of tiering? | 21:35 |
| jgriffith | backend selection? | 21:35 |
| jgriffith | zykes-: geographical location.... | 21:35 |
| *** eglynn has joined #openstack-dev | 21:44 | |
| zykes- | jgriffith: tiering yes | 21:44 |
| zykes- | or basically: if I have a SAN and a ISCSI storage in my cluster | 21:44 |
| zykes- | and I wanna say that all vm disks go to SAN but all data goes to ISCSI ? | 21:44 |
| zykes- | by default but still be choosable | 21:44 |
| jgriffith | zykes-: I'm hoping to have multiple back-end support available in cinder which would let you do something like this | 21:45 |
| jgriffith | zykes-: Don't know if it'll make it though, we'll have to see | 21:45 |
| zykes- | :D | 21:45 |
| jgriffith | zykes-: Definitely something on the roadmap | 21:45 |
| zykes- | its a enterprise thign as well :p | 21:45 |
| zykes- | kinda | 21:45 |
| jgriffith | zykes-: although when you say SAN, if you're talking FC that's a whole seperate deal | 21:45 |
| jgriffith | zykes-: That's debatable, but I understand what you're saying | 21:46 |
| *** jgriffith has quit IRC | 21:49 | |
| *** jgriffith has joined #openstack-dev | 21:52 | |
| *** johnpur has left #openstack-dev | 21:52 | |
| *** cloudvirt has quit IRC | 21:54 | |
| *** dolphm has joined #openstack-dev | 21:56 | |
| *** avishay has quit IRC | 21:56 | |
| *** Shrews has quit IRC | 22:00 | |
| *** flaviamissi has quit IRC | 22:00 | |
| *** johnpostlethwait has quit IRC | 22:00 | |
| *** eglynn has quit IRC | 22:01 | |
| *** maoy_ has quit IRC | 22:01 | |
| *** lloydde has quit IRC | 22:02 | |
| *** lloydde has joined #openstack-dev | 22:03 | |
| *** johnpostlethwait has joined #openstack-dev | 22:03 | |
| *** andrewsmedina has quit IRC | 22:04 | |
| *** _mancdaz has quit IRC | 22:05 | |
| *** mancdaz has joined #openstack-dev | 22:05 | |
| jeblair | jtran: i have reproduced your resoults, thanks. | 22:07 |
| jeblair | mtaylor: ^ | 22:07 |
| jtran | jeblair, you're welcome ;) | 22:08 |
| jeblair | jtran: i'm kicking jenkins again on that change; when it's done, i'll take that done offline and see if i can reproduce the result manually on the host and dig further into it there. | 22:09 |
| notmyname | davidha: thanks for posting the bug. | 22:09 |
| jeblair | s/take that done/take that node/ | 22:09 |
| jtran | jeblair, sounds like an excellent plan | 22:10 |
| *** andrewbogott is now known as andrewbogott_afk | 22:10 | |
| *** markmcclain has quit IRC | 22:10 | |
| *** dspano has quit IRC | 22:13 | |
| *** datsun180b has quit IRC | 22:14 | |
| *** rnirmal has quit IRC | 22:24 | |
| *** kbringard has quit IRC | 22:27 | |
| jaypipes | wow, neverending foundation thread today... | 22:27 |
| *** dubsquared has quit IRC | 22:29 | |
| cloudfly | i am running some lengthy qa checks today | 22:31 |
| cloudfly | so i have a lot of free time to read and comment on that thread | 22:31 |
| *** cloudvirt has joined #openstack-dev | 22:31 | |
| cloudfly | i should probably try to cut myself off a bit | 22:31 |
| jaypipes | cloudfly: and you are? :) | 22:34 |
| *** danwent has quit IRC | 22:37 | |
| *** tgall_foo has quit IRC | 22:37 | |
| cloudfly | matt joyce | 22:37 |
| *** danwent has joined #openstack-dev | 22:38 | |
| jaypipes | cloudfly: oh, duh. I knew that :) | 22:38 |
| cloudfly | =D | 22:38 |
| *** danwent has quit IRC | 22:39 | |
| jaypipes | cloudfly: yeah, much like soren, I'm having a bit of chuckle from the conversation about people representing their company in the foundation board... since I've worked for 3 of them in the past year ;) | 22:39 |
| jaypipes | cloudfly: and frankly, my representation in the community or the PPB hasn't changed at all regardless of employer | 22:39 |
| bcwaldon | cloudfly: I dont remember seeing a recent response RE python-glanceclient | 22:40 |
| bcwaldon | cloudfly: how do you feel aboutt a temporary compatibility layer in the cli? | 22:41 |
| jaypipes | bcwaldon: what about just calling the CLI glance2? | 22:42 |
| bcwaldon | jaypipes: we already have people using it as 'glance' | 22:42 |
| bcwaldon | jaypipes: and I dont want to carry that baggage forever | 22:42 |
| bcwaldon | jaypipes: this is just glance client v2 | 22:42 |
| jaypipes | bcwaldon: well, yeah, but we already have a lot more using bin/glance as glance :) | 22:42 |
| *** dachary has quit IRC | 22:43 | |
| bcwaldon | jaypipes: I do understand that | 22:43 |
| jaypipes | bcwaldon: it's a tough situation, I know. | 22:43 |
| *** Gordonz has quit IRC | 22:43 | |
| bcwaldon | jaypipes: I do want to talk to someone who is really feeling this pain | 22:43 |
| bcwaldon | jaypipes: because this upgrade is going to do *nothing* to nodes that aren't running openstack services | 22:44 |
| jaypipes | bcwaldon: understood. | 22:44 |
| bcwaldon | jaypipes: And I understand all the concerns that were brought up, I just want to determine which ones are *real* concerns | 22:44 |
| jaypipes | yup, ++ | 22:45 |
| cloudfly | jaypipes sure but you guys are different you are core technical talent | 22:45 |
| *** edygarcia has quit IRC | 22:45 | |
| bcwaldon | cloudfly: that doesnt matter right now | 22:45 |
| cloudfly | they guy working third tier support for hp cloud isn't going to want to 'rock the boat' i doubt he even cares who gets elected | 22:45 |
| cloudfly | he votes to look good | 22:45 |
| jaypipes | bcwaldon: no, I think cloudfly was referring to a different conversation :) | 22:45 |
| cloudfly | and that introduces a bias | 22:45 |
| bcwaldon | OH | 22:45 |
| cloudfly | sorry yes i was idle for a bit | 22:45 |
| bcwaldon | whew | 22:45 |
| jaypipes | cloudfly: understood. | 22:45 |
| cloudfly | bcwaldon yeah i wish the user community would be more engaged but i get why they aren't. | 22:47 |
| cloudfly | it's a tough situation | 22:47 |
| *** mnewby_ has joined #openstack-dev | 22:47 | |
| bcwaldon | cloudfly: so do you have an ideal path forard for this? | 22:48 |
| bcwaldon | forward* | 22:49 |
| *** mnewby has quit IRC | 22:49 | |
| *** mnewby_ is now known as mnewby | 22:49 | |
| *** thingee is now known as thingee_zz | 22:49 | |
| *** sacharya has quit IRC | 22:50 | |
| *** maurosr has joined #openstack-dev | 22:55 | |
| *** heckj has joined #openstack-dev | 22:55 | |
| *** PotHix has quit IRC | 22:56 | |
| mtaylor | jaypipes: dude. I haven't had a email thread this fun since marten mickos wanted to make portions of MySQL closed-source :) | 23:01 |
| jaypipes | mtaylor: oooh. another one I was involved in. :) | 23:01 |
| mtaylor | jaypipes: it all comes down to you | 23:01 |
| Daviey | jaypipes and his freedom hating agenda. | 23:02 |
| mtaylor | jaypipes: I should dig up some of my emails on that thread and send them to this one ... and then ask who thinks my company's voice is mine :) | 23:02 |
| mtaylor | Daviey: jaypipes hates so many things, freedom doesn't even make the list | 23:02 |
| mikal | I'm missing a ranty thread on the internet? | 23:03 |
| mikal | Why was I not informed? | 23:03 |
| dolphm | ayoung: heckj: just put policy CRUD up for review -- it's still on /v2.0/ but otherwise ready to rock | 23:03 |
| jaypipes | Daviey: :) | 23:03 |
| Daviey | mikal: If you were not included on the thread, it's probably ABOUT you... | 23:04 |
| jaypipes | mikal: because freedom hates you, that's why. don'tcha know? ;P | 23:04 |
| mikal | :( | 23:04 |
| mikal | I can't help it that I'm a communist | 23:04 |
| Daviey | So, including you in the thread may of hurt your feelings. | 23:04 |
| mikal | Its just the way I was brought up | 23:04 |
| Daviey | </troll> | 23:04 |
| jgriffith | everybody quiet... mikal is here, act busy | 23:05 |
| heckj | dolphm: just saw it hit in email - need to change that to v3, but otherwise sounds good. I'd really like to get client tests written to force the driving of it, so that's first off. | 23:05 |
| dolphm | heckj: client-side tests: https://review.openstack.org/#/c/9237/8/tests/v2_0/test_policies.py | 23:06 |
| dolphm | heckj: client-dependent tests: https://review.openstack.org/#/c/9744/4/tests/test_keystoneclient_sql.py | 23:06 |
| *** e1mer has joined #openstack-dev | 23:07 | |
| *** mnewby has quit IRC | 23:07 | |
| *** lloydde has quit IRC | 23:07 | |
| heckj | dolphm: awesome dude!~ | 23:08 |
| cloudfly | bcwaldon i support the glance2 binary | 23:09 |
| cloudfly | =/ | 23:09 |
| cloudfly | i don't like it | 23:09 |
| cloudfly | but i see it as the lesser evil | 23:09 |
| bcwaldon | cloudfly: I think thats a little silly, no? | 23:10 |
| jaypipes | lol, at this rate the foundation thread might take the all-time thread count title. | 23:10 |
| *** roge has quit IRC | 23:10 | |
| cloudfly | i think its a little silly for this. but i think it sets a great precedent and it does say we tried to keep compatibility for users. | 23:10 |
| cloudfly | we can always deprecate it later when a year from now someone revamps the client again | 23:11 |
| cloudfly | =/ | 23:11 |
| bcwaldon | cloudfly: are you planning on doing that? | 23:11 |
| bcwaldon | cloudfly: what's wrong with thinking of this as a major version change? | 23:12 |
| cloudfly | it comes down to breaking compatibility of shell scripts on users between versions with no warning | 23:12 |
| cloudfly | as far as they are concerned it is no warning unless you feed them "THIS IS DEPRECATED STOP USING IT" in their STDERR they aren't going to know. | 23:12 |
| bcwaldon | cloudfly: how would you feel about making the same interface work in python-glanceclient? | 23:12 |
| dolphm | heckj: my only api concern after playing with the impl... {'policy': {'policy': '...'}} is a bit awkward imo; more explicitly: {'policy': {'blob': '...'}} | 23:13 |
| cloudfly | i think the only concern i have is shell scripts calling glance and glance breaking | 23:13 |
| bcwaldon | cloudfly: and I am suggesting we fix that | 23:13 |
| cloudfly | if it's easy enough to support all the old calls glance made. | 23:13 |
| cloudfly | sure | 23:13 |
| bcwaldon | cloudfly: anything is possible, and I'm going to be doing all the freaking work anywyas | 23:13 |
| cloudfly | with deprecation warnings etc. | 23:13 |
| bcwaldon | cloudfly: yep | 23:14 |
| cloudfly | i mean yeah that'd be bananas | 23:14 |
| bcwaldon | cloudfly: bananas == bad? | 23:14 |
| jaypipes | from ewan: "In general I hate customers, and the last thing I would ever be fanatical about is support.". Classic. | 23:14 |
| cloudfly | good | 23:14 |
| bcwaldon | cloudfly: good | 23:14 |
| cloudfly | more exceptional | 23:14 |
| cloudfly | jaypipes i was dying when i read that | 23:14 |
| bcwaldon | cloudfly: so how about I do that and take the issue back to the ML | 23:14 |
| cloudfly | i'll +` | 23:15 |
| cloudfly | err 1 | 23:15 |
| bcwaldon | ok | 23:15 |
| bcwaldon | I think that gets everybody what they want | 23:15 |
| bcwaldon | I want to move everyone to python-glanceclient | 23:15 |
| bcwaldon | and I'm sorry its taking a while but the resources we have to do things in glance is extremely small | 23:15 |
| bcwaldon | we should have had this done as soon as we released essex | 23:16 |
| *** andrewsmedina has joined #openstack-dev | 23:17 | |
| *** dolphm has quit IRC | 23:18 | |
| jeblair | jtran: curiouser and curiouser -- as a user on the jenkins slave, i can run the tests successfully. | 23:22 |
| jeblair | (still digging) | 23:22 |
| jeblair | mtaylor: ^ | 23:22 |
| heckj | dolphm: that's why I left the spec open, so we could tweak it to make it more sensible for implementation :-) | 23:24 |
| jeblair | jtran, mtaylor: here is the output at the end of run_tests.log from the (failed) jenkins run: http://paste.openstack.org/show/19913/ | 23:24 |
| *** heckj has quit IRC | 23:24 | |
| bcwaldon | jaypipes: https://review.openstack.org/#/c/10518/2 | 23:34 |
| *** markwash has joined #openstack-dev | 23:34 | |
| bcwaldon | markwash: https://review.openstack.org/#/c/10520/ | 23:34 |
| *** halfss has joined #openstack-dev | 23:36 | |
| *** matwood has quit IRC | 23:40 | |
| *** sacharya has joined #openstack-dev | 23:46 | |
| *** mnewby has joined #openstack-dev | 23:46 | |
| *** davidha has quit IRC | 23:58 | |
| *** davidha has joined #openstack-dev | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!