| openstackjenkins | Project burrow build #32: SUCCESS in 11 sec: https://jenkins.openstack.org/job/burrow/32/ | 00:02 |
|---|---|---|
| openstackjenkins | Tarmac: Cleaned up WSGI frontend. | 00:02 |
| *** bengrue has quit IRC | 00:03 | |
| *** zns has joined #openstack-dev | 00:06 | |
| *** zns1 has joined #openstack-dev | 00:11 | |
| *** zns has quit IRC | 00:13 | |
| mtaylor | vishy: ok- that was pretty easy | 00:13 |
| mtaylor | vishy: lemme get this uploaded somewhere and fire off a bug | 00:14 |
| vishy | coolness | 00:14 |
| mtaylor | vishy: they may bitch about lack of manpages ... | 00:14 |
| *** heckj has quit IRC | 00:23 | |
| mtaylor | vishy: k. bug filed - cc'd you on it | 00:23 |
| mtaylor | vishy: do you need it in archive right now? or can I do that bit tomorrow? | 00:24 |
| vishy | mtaylor: tomorrow is fine | 00:26 |
| mtaylor | cool. that allows for more beer | 00:26 |
| *** donald650 has joined #openstack-dev | 00:26 | |
| *** mwhooker has quit IRC | 00:35 | |
| *** mwhooker has joined #openstack-dev | 00:36 | |
| *** donald650 has quit IRC | 00:39 | |
| *** chipc has joined #openstack-dev | 00:46 | |
| *** Tushar has joined #openstack-dev | 00:47 | |
| *** jakedahn has quit IRC | 00:48 | |
| *** anotherjesse_ has joined #openstack-dev | 00:51 | |
| *** anotherjesse_ has quit IRC | 00:51 | |
| *** anotherjesse has quit IRC | 00:55 | |
| *** anotherjesse has joined #openstack-dev | 00:56 | |
| *** anotherjesse has quit IRC | 01:00 | |
| *** martine has joined #openstack-dev | 01:02 | |
| openstackgerrit | Verification of a change to openstack/keystone failed: More contract changes https://review.openstack.org/252 | 01:06 |
| *** chipc has quit IRC | 01:11 | |
| *** Tushar has quit IRC | 01:22 | |
| openstackgerrit | Ziad Sawalha proposed a change to openstack/keystone: More contract changes https://review.openstack.org/252 | 01:25 |
| *** vladimir3p has quit IRC | 01:25 | |
| *** mfer has joined #openstack-dev | 01:30 | |
| openstackgerrit | Ziad Sawalha proposed a change to openstack/keystone: Contract changes https://review.openstack.org/252 | 01:31 |
| openstackgerrit | A change was merged to openstack/keystone: Contract changes https://review.openstack.org/252 | 01:34 |
| *** jakedahn has joined #openstack-dev | 01:34 | |
| *** zns1 has quit IRC | 01:35 | |
| *** chipc has joined #openstack-dev | 02:28 | |
| *** hisaharu has quit IRC | 02:28 | |
| *** mfer has quit IRC | 02:45 | |
| *** jakedahn has quit IRC | 03:16 | |
| *** jakedahn has joined #openstack-dev | 03:18 | |
| openstackgerrit | Monty Taylor proposed a change to openstack/openstack-ci: Added a version of the rfc script from gluster. https://review.openstack.org/235 | 03:21 |
| openstackgerrit | Monty Taylor proposed a change to openstack/openstack-ci: Added a version of the rfc script from gluster. https://review.openstack.org/235 | 03:22 |
| *** dolphm_ has quit IRC | 03:30 | |
| *** jakedahn has quit IRC | 03:31 | |
| *** anotherjesse has joined #openstack-dev | 03:43 | |
| *** jakedahn has joined #openstack-dev | 03:51 | |
| *** martine_ has joined #openstack-dev | 03:59 | |
| *** zns has joined #openstack-dev | 04:00 | |
| *** martine_ has quit IRC | 04:00 | |
| *** martine_ has joined #openstack-dev | 04:00 | |
| *** nmistry has joined #openstack-dev | 04:01 | |
| *** martine_ has quit IRC | 04:01 | |
| *** chipc has quit IRC | 04:02 | |
| *** amccabe has joined #openstack-dev | 04:10 | |
| *** zul has quit IRC | 04:16 | |
| *** martine has quit IRC | 04:16 | |
| *** zaitcev has quit IRC | 04:22 | |
| *** nmistry has quit IRC | 04:28 | |
| *** anotherjesse has quit IRC | 04:35 | |
| *** anotherjesse has joined #openstack-dev | 04:50 | |
| *** anotherjesse has quit IRC | 04:58 | |
| *** reidrac has joined #openstack-dev | 06:55 | |
| *** toluene has joined #openstack-dev | 08:06 | |
| toluene | hi! | 08:06 |
| *** mnour has joined #openstack-dev | 08:30 | |
| *** amccabe has quit IRC | 08:38 | |
| *** rods has joined #openstack-dev | 09:56 | |
| *** progers has joined #openstack-dev | 10:01 | |
| openstackgerrit | Yuriy Taraday proposed a change to openstack/keystone: Made all sample data loading in one script. https://review.openstack.org/255 | 10:26 |
| *** progers has left #openstack-dev | 10:30 | |
| *** titaniumrain_ has joined #openstack-dev | 10:36 | |
| *** toluene has quit IRC | 11:11 | |
| *** markvoelker has joined #openstack-dev | 11:16 | |
| *** mfer has joined #openstack-dev | 11:38 | |
| *** bsza has joined #openstack-dev | 12:11 | |
| *** deshantm_laptop has joined #openstack-dev | 12:17 | |
| *** titaniumrain_ has quit IRC | 12:17 | |
| *** lorin1 has joined #openstack-dev | 12:19 | |
| *** lorin1 has left #openstack-dev | 12:21 | |
| *** zns has quit IRC | 12:38 | |
| *** lts has joined #openstack-dev | 12:43 | |
| sandywalsh | jaypipes, around? | 13:15 |
| jaypipes | sandywalsh: nope. | 13:15 |
| sandywalsh | phew | 13:15 |
| jaypipes | sandywalsh: :) course I am. | 13:15 |
| jaypipes | sandywalsh: what's new, Canadian? | 13:15 |
| sandywalsh | jaypipes, I added some notes on the glance zone etherpad http://etherpad.openstack.org/GlanceZones | 13:16 |
| sandywalsh | jaypipes, be interested on your $0.02 CDN | 13:16 |
| jaypipes | sandywalsh: looking. :) | 13:16 |
| *** zns has joined #openstack-dev | 13:18 | |
| jaypipes | sandywalsh: so, Glance now has support for shared image groups and keystone authentication (Vek's code was merged yesterday) | 13:18 |
| jaypipes | zns: ah, good morning fellow. | 13:18 |
| *** joesavak has joined #openstack-dev | 13:19 | |
| *** ameade has joined #openstack-dev | 13:20 | |
| sandywalsh | jaypipes, cool ... wasn't sure how deep he had gone with it (beyond basic auth) ... I guess my larger concern is the potential single point of failure. | 13:21 |
| *** dprince has joined #openstack-dev | 13:22 | |
| jaypipes | sandywalsh: I raised that concern as well.. | 13:22 |
| zns | jaypipes: monin :-) | 13:23 |
| sandywalsh | jaypipes, thoughts on the pubsubhubbub sync from the zone-specific registries back to the global registry? and using the keystone Service Catalog to find the set of endpoints? | 13:24 |
| jaypipes | sandywalsh: I think that the idea we settled on was to have a single global glance registry for Rackspace that would house "golden" images or public images shared by partners like RightScale. Then have regional Glance registries that are essentially read-only slaves of the global registry. In addition, to provide redundancy and locality of data, introduce a set of image replica locations that regional glance API nodes | 13:25 |
| jaypipes | would use to find the "closest" image copy... | 13:25 |
| sandywalsh | jaypipes, does that mean you can only restore a backup within the zone it was created in? | 13:26 |
| jaypipes | sandywalsh: no. it would backup the image to the local zone glance API server, register it with the global registry database, then replicate itself to separate locations. | 13:26 |
| jaypipes | sandywalsh: with a configurable setting for after X copies of the backup are stored, return success... | 13:27 |
| jaypipes | sandywalsh: of course, if you are storing the backups in Swift, Swift handles that redundancy for you.. | 13:28 |
| sandywalsh | jaypipes, replicate the image or the metadata/registry entry? | 13:28 |
| jaypipes | sandywalsh: just not currently across multiple zones... | 13:28 |
| jaypipes | sandywalsh: replicate the image file itself. the metadata would go to the global registry and be "replicated" via standard MySQL replication... | 13:28 |
| jaypipes | sandywalsh: in other words, there is no local read/write glance registries.. | 13:29 |
| sandywalsh | jaypipes, that was a real surprise for me in the design. I assume Swift did all that redundancy for us and that it would be deployed as a fabric underneath the nova zones (much like keystone is) | 13:29 |
| sandywalsh | *assumed | 13:29 |
| jaypipes | sandywalsh: are you surprised that Swift doesn't do cross-zone replication or surprised with the global glance registry design? | 13:30 |
| sandywalsh | jaypipes, well both, but that swift would only be deployed per-zone and not as a larger fabric | 13:30 |
| jaypipes | sandywalsh: well, you will have to discuss the swift question with s1rp and pvo (I'm not familiar enough with the discussion on that... I *think* it has to do with limitations in Swift itself, but I may be wrong). For the global glance registry, the design decision was made mostly because of a desire by product management to have a single URI to refer to an image... | 13:32 |
| sandywalsh | jaypipes, yeah, I can see the attraction of a single URI, but we could still take a DNS-like approach to resolving the final destination vs. assuming 1 large/stationary registry (I think). A set of registries that could resolve the URI. | 13:34 |
| jaypipes | sandywalsh: sure, that is true I suppose. I think this route was taken because it was easier and faster, with the thought that we would tackle the scale-out concerns in a phase 2. | 13:37 |
| jaypipes | sandywalsh: with "intermediate scale-out" resolved with the introduction of image replicas... | 13:37 |
| sandywalsh | jaypipes, yeah, fair enough. I guess the only "critical" design feature would be to settle on UUID/URI vs. integer ID, for now. | 13:37 |
| jaypipes | sandywalsh: I think that's a settled issue. It's just a matter of prioritizing that and making it happen... | 13:38 |
| sandywalsh | jaypipes, ok, cool. thanks for the feedback sir! | 13:39 |
| jaypipes | sandywalsh: any time! :) | 13:39 |
| jaypipes | sandywalsh: Erik Carlin is a good person to talk to about this, too. | 13:39 |
| sandywalsh | jaypipes, gotcha ... his notes in the etherpad do reflect more of the requirements that I'm perhaps not privy to | 13:40 |
| jaypipes | sandywalsh: ya, no worries. Erik's always up for a chat about this stuff. Very approachable about it... | 13:41 |
| sandywalsh | jaypipes, excellent, will do | 13:41 |
| jaypipes | sandywalsh: the only thing I've been pushing back on in the design stuff is to make sure the design decisions we make are optional.. we cannot assume that Rackspace is the only use-case for Glance. So, things like a global registry must not be assumptions in the core architecture. They must be options... | 13:42 |
| jaypipes | sandywalsh: and by "pushing back" I didn't mean to sound like I've been a negative Nancy. Just pointing out that making things configurable is the way to go.. | 13:43 |
| sandywalsh | jaypipes, understood. I'm really just focused on the issue where, let's say, ORD goes down and we have a black hole. The "optional" requirement is a valid one ... just as we're becoming more dependent on keystone. | 13:44 |
| jaypipes | sandywalsh: well, *technically* Keystone is optional in Glance :) However, without *some* auth solution (Keystone or otherwise), you don't get a bunch of additional functionality (like shared image groups, attribute ownership, etc). But, that said, a Glance registry for a few global, public images doesn't need authentication at all... | 13:46 |
| sandywalsh | jaypipes, right. So how would bursting (public/private clouds) work? Does the private cloud need to adopt the public clouds image registry? | 13:48 |
| jaypipes | sandywalsh: totally depends on the specific requirements of the relationship. If the private cloud is attempting to boot an instance in a public cloud, and supplies an image ID that refers to a publically accessible image on their own auth-less Glance registry, absolutely nothing needs to change. If they supply an image ID to an auth-enabled private Glance registry, the public cloud's user/tenant would be used to dete | 13:49 |
| jaypipes | rmine if the supplied image ID is available to them. | 13:49 |
| *** zul has joined #openstack-dev | 13:52 | |
| *** zul has quit IRC | 13:52 | |
| sandywalsh | jaypipes, yeah, that makes sense. They'd have to make their registry publicly available. However we could use federated auth so they don't have to depend on "auth-less" glance | 13:52 |
| sandywalsh | jaypipes, I wouldn't want my backups to be publicly available | 13:52 |
| jaypipes | sandywalsh: yes, once the oauth (and other) drivers are completed in Keystone. | 13:52 |
| *** zul has joined #openstack-dev | 13:53 | |
| *** chipc has joined #openstack-dev | 13:58 | |
| openstackjenkins | Project nova build #1,260: SUCCESS in 3 min 35 sec: https://jenkins.openstack.org/job/nova/1260/ | 14:00 |
| openstackjenkins | Tarmac: DB object was being casted to dict() in API code. This did not work as intended and logic has been updated to reflect a more accurate way of getting information out of DB objects. | 14:00 |
| *** amccabe has joined #openstack-dev | 14:14 | |
| openstackgerrit | Yuriy Taraday proposed a change to openstack/keystone: Made all sample data loading in one script. https://review.openstack.org/255 | 14:16 |
| dolphm | ^^ awesome! | 14:16 |
| uvirtbot | dolphm: Error: "^" is not a valid command. | 14:16 |
| dolphm | uvirtbot: You only ever bark at me, and I don't even know what you're for. | 14:18 |
| uvirtbot | dolphm: Error: "You" is not a valid command. | 14:18 |
| dolphm | uvirtbot: :( | 14:18 |
| uvirtbot | dolphm: Error: ":(" is not a valid command. | 14:18 |
| dolphm | uvirtbot: sadness should be a command | 14:18 |
| uvirtbot | dolphm: Error: "sadness" is not a valid command. | 14:18 |
| joesavak | uvirbot: uvirbot sucks | 14:19 |
| joesavak | uvirtbot: uvirtbot sucks | 14:19 |
| uvirtbot | joesavak: Error: "uvirtbot" is not a valid command. | 14:19 |
| joesavak | damn right. | 14:19 |
| dolphm | uvirtbot: help me understand why you complain so much | 14:19 |
| uvirtbot | dolphm: Error: There is no command "me understand why you complain so much". | 14:19 |
| dolphm | uvirtbot: disconnect yourself | 14:20 |
| uvirtbot | dolphm: Error: "disconnect" is not a valid command. | 14:20 |
| joesavak | uvirtbot: quit() | 14:20 |
| uvirtbot | joesavak: Error: "quit()" is not a valid command. | 14:20 |
| joesavak | dolph: quick q - "anp" is a new user to keystone and he's setting up the sample data in his install and using the dashboard to login. When loggin in he gets "class Roles has no attribute get_default_page" | 14:22 |
| joesavak | dolph: is this because keystone may be out of sync with dashboard? | 14:22 |
| dolphm | uvirtbot: eval('quit()') | 14:22 |
| uvirtbot | dolphm: Error: "eval('quit()')" is not a valid command. | 14:22 |
| openstackgerrit | Ziad Sawalha proposed a change to openstack/keystone: Made all sample data loading in one script. https://review.openstack.org/255 | 14:23 |
| dolphm | joesavak: i've never seen anything called "get_default_page" in keystone... let me look into it | 14:23 |
| joesavak | dolph: thanks | 14:23 |
| dolphm | joesavak: it's got to be a dashboard issue... keystone sort of has a concept of pages, but not concept of 'default' pages | 14:26 |
| joesavak | dolph: thanks | 14:27 |
| joesavak | dolph: who do we go to for that? | 14:29 |
| pvo | sandywalsh: re: multiple copies of an image in glance…. | 14:30 |
| pvo | the thinking was if you had an image in a dc in the US and you wanted to boot that image in the UK you wouldn't want to pay the bandwidth every time | 14:30 |
| pvo | but you would want to keep the same image id across all regions | 14:30 |
| dolphm | devcamcar: jakedahn: ping ^^ | 14:30 |
| dolphm | joesavak: those are two of the top dashboard contributors in this channel (from https://launchpad.net/openstack-dashboard ) | 14:31 |
| *** cp16net has joined #openstack-dev | 14:31 | |
| pvo | so you could have copies of that image nearby and the nova deployment would have a concept of regionality and try to pull the 'closest' copy of the image to save bandwidth. | 14:31 |
| joesavak | dolph: thanks | 14:31 |
| zykes- | &clea | 14:32 |
| sandywalsh | pvo, makes sense. But it would have to be an explicit action by the user/sysadmin to place it in the optimal location, yes? | 14:32 |
| pvo | yes, its optional to do it that way. | 14:33 |
| sandywalsh | pvo, otherwise the SP has to foot the bill of having a copy of the image in all locations | 14:33 |
| openstackgerrit | Ziad Sawalha proposed a change to openstack/keystone: Minor fix to run_tests https://review.openstack.org/256 | 14:33 |
| pvo | sandywalsh: yes, that is how I thought it would work. User moves it around and registers alternate locations. | 14:33 |
| *** vladimir3p has joined #openstack-dev | 14:34 | |
| pvo | I was wondering if we'd use something like mx preference or just let the zone try to match on dns. | 14:34 |
| sandywalsh | mx? | 14:34 |
| pvo | when you look up mail records in dns, there is a preferred order to try. | 14:35 |
| sandywalsh | ah, right ... was thinking some RS thing :) | 14:35 |
| pvo | na, normal internet funtimes | 14:35 |
| sandywalsh | I like the idea of the user setting their preference | 14:35 |
| sandywalsh | with possible overrides | 14:35 |
| sandywalsh | ... less magic | 14:35 |
| pvo | yea, that would be my first iteration since it feels simpler. | 14:36 |
| openstackgerrit | A change was merged to openstack/keystone: Minor fix to run_tests https://review.openstack.org/256 | 14:36 |
| *** amccabe has quit IRC | 14:46 | |
| sandywalsh | http://www.colinhowe.co.uk/2011/aug/17/visualising-eventlet/ | 14:46 |
| openstackgerrit | Ziad Sawalha proposed a change to openstack/keystone: Made all sample data loading in one script. https://review.openstack.org/255 | 14:53 |
| *** deshantm_laptop_ has joined #openstack-dev | 14:54 | |
| *** dragondm has joined #openstack-dev | 14:55 | |
| openstackgerrit | A change was merged to openstack/keystone: Made all sample data loading in one script. https://review.openstack.org/255 | 14:55 |
| *** deshantm_laptop has quit IRC | 14:56 | |
| *** deshantm_laptop_ is now known as deshantm_laptop | 14:56 | |
| *** deshantm_laptop has quit IRC | 14:58 | |
| *** zul has quit IRC | 14:58 | |
| *** deshantm_laptop has joined #openstack-dev | 14:58 | |
| openstackgerrit | Ziad Sawalha proposed a change to openstack/keystone: Additional changes to fix minor service support stuff and increase test coverage. Also making validate token call available using service admin tokens https://review.openstack.org/242 | 14:59 |
| openstackgerrit | Verification of a change to openstack/keystone failed: Additional changes to fix minor service support stuff and increase test coverage. Also making validate token call available using service admin tokens https://review.openstack.org/242 | 14:59 |
| *** amccabe has joined #openstack-dev | 15:03 | |
| *** reidrac has quit IRC | 15:07 | |
| sandywalsh | jaypipes, if we have an image URL, why do we need the registry at all? | 15:16 |
| *** martine has joined #openstack-dev | 15:16 | |
| pvo | sandywalsh: when you say image url, that is where the actual bits of the image are located? | 15:17 |
| sandywalsh | pvo, I assume the only reason for the registry is to map "image name" or "unique image identifier" to a URL? So why not make the "unique image identifier" the URL and be done with it? | 15:18 |
| *** mnour has quit IRC | 15:18 | |
| sandywalsh | pvo, the payload from fetching the URL could be the metadata + another URL to the actual image | 15:18 |
| sandywalsh | pvo, so the registry is nothing more than a static HTTP server | 15:19 |
| *** mattray has joined #openstack-dev | 15:19 | |
| jaypipes | sandywalsh: the registry stores other information that is not in the URI (such as security creds) :) | 15:21 |
| sandywalsh | jaypipes, what are the security creds? permissions on the image or something else? | 15:22 |
| *** kbringard has joined #openstack-dev | 15:23 | |
| jaypipes | sandywalsh: such as user/key in Swift, accessky/secretkey in S3, etc. There is a "store URI" and the "image URI". The image URI might be http://glance.rackspace.com/images/1, but the store URI might be http://account:user:pass@ord1.rackspace.com/auth/v1/glance/12345 | 15:23 |
| *** deshantm_laptop has quit IRC | 15:23 | |
| jaypipes | sandywalsh: and when I mentioned image replicas earlier, what I was referring to is a single image URI being tied to multiple store URIs... | 15:24 |
| *** dolphm_ has joined #openstack-dev | 15:24 | |
| sandywalsh | jaypipes, hmm, ok. I guess my confusion comes from assuming that swift was going to do all the heavy lifting and glance was just an API that made more sense to image management + metadata. | 15:25 |
| sandywalsh | jaypipes, but it sounds like glance is also doing double duty on some things that swift does | 15:26 |
| jaypipes | sandywalsh: no, it's not really... | 15:26 |
| jaypipes | sandywalsh: what "things that swift does" is Glance doing? | 15:27 |
| openstackgerrit | Monty Taylor proposed a change to openstack/openstack-ci-puppet: Prevent add-apt-repository from running every time https://review.openstack.org/257 | 15:27 |
| notmyname | jaypipes: sandywalsh: each storage URL is one logically autonomous cluster (with it's own replication). there are still use cases for storing soemthing in 2+ swift clusters. it's more costly to the user and/or deployer, but it can give you things like geographic redundancy | 15:27 |
| sandywalsh | jaypipes, (again, excuse my ignorance here) ... does it also deal with image replication/movement/storage? | 15:27 |
| openstackgerrit | Verification of a change to openstack/keystone failed: Additional changes to fix minor service support stuff and increase test coverage. Also making validate token call available using service admin tokens https://review.openstack.org/242 | 15:28 |
| notmyname | jaypipes: sandywalsh: at lest until we can solve wide geographic distribution in one logical swift cluster :-) | 15:28 |
| jaypipes | sandywalsh: geographic redundancy is the only thing that Glance image replicas add | 15:29 |
| sandywalsh | notmyname, ah, ok, so swift doesn't deal with geographic replication | 15:29 |
| jaypipes | sandywalsh: correct | 15:29 |
| jaypipes | sandywalsh: yet, at least :) | 15:29 |
| sandywalsh | gotcha ... I was assuming it did | 15:29 |
| jaypipes | sandywalsh: I'm sure notmyname can code it up in a couple hours ;) | 15:29 |
| sandywalsh | wouldn't that be the right place for it? | 15:30 |
| notmyname | jaypipes: np. I'll start by messing with changing c to lower latency so it doesn't cause any replication problems ;-) | 15:30 |
| jaypipes | excellent :) | 15:30 |
| *** zns1 has joined #openstack-dev | 15:31 | |
| notmyname | sandywalsh: not necessarily. swift needs it (eventually). but not all deployers will be able to take advantage of it | 15:31 |
| sandywalsh | heh, well glance would need to redefine C too wouldn't it? :) | 15:31 |
| jaypipes | sandywalsh: what notmyname said ;) | 15:31 |
| creiht | I thought you just created a blueprint, then someone would come along and magically do it :) | 15:31 |
| notmyname | sandywalsh: for example, even if swift supports it, it would require big pipes between DCs. | 15:31 |
| jaypipes | creiht: the magic coding unicorn. | 15:31 |
| sandywalsh | creiht, yeah ... not the case? | 15:32 |
| *** zns has quit IRC | 15:32 | |
| sandywalsh | notmyname, but glance doing the replication has the same issue, no? | 15:32 |
| sandywalsh | so, is the intention that the user explicitly dumps the images where desired so as to be fully aware of the added costs? | 15:33 |
| notmyname | sandywalsh: jaypipes: or, another cool implementation that would be totally in glance would be putting it in swift clusters provided by different deployers. (eg store in rackspace cloud files and internap and KT). each could be using swift's geographic features but still there is a use case for having it in glance | 15:33 |
| notmyname | in that example, rackspace could store it in DFW and ORD. but the use also stores in with KT for extra sooper dooper redundancy | 15:34 |
| jaypipes | notmyname: that's definitely a cool feature, and something that the image replica functionality would address. | 15:34 |
| sandywalsh | gotta step out for a bit ... tbc | 15:35 |
| notmyname | it's the big-scale version of glance storing it in 2 separate RAID volumes locally mounted. | 15:35 |
| sandywalsh | notmyname, any docs you would recommend on swift architecture? | 15:36 |
| sandywalsh | or presentations? | 15:36 |
| notmyname | sandywalsh: what part? | 15:36 |
| sandywalsh | redundancy, etc | 15:36 |
| sandywalsh | notmyname, no so much the api's, etc, but overall scalability, etc | 15:37 |
| sandywalsh | *not | 15:37 |
| notmyname | sandywalsh: swift.openstack.org is the best place to start. we may have a few presentations floating around, but I'm not sure if they would be detailed enough | 15:37 |
| *** zul has joined #openstack-dev | 15:37 | |
| sandywalsh | notmyname, cool ... I'll give them a look ... thanks | 15:37 |
| notmyname | sandywalsh: most of the knowledge sharing there involved a whiteboard and a few hours of one-on-one time :-) | 15:37 |
| sandywalsh | notmyname, summit is coming up! | 15:38 |
| notmyname | sandywalsh: indeed | 15:38 |
| notmyname | and it looks like I'll be talking at the bay area meetup in a couple of weeks too | 15:38 |
| *** ohnoimdead has joined #openstack-dev | 15:40 | |
| *** rnirmal has joined #openstack-dev | 15:41 | |
| *** troytoman-away is now known as troytoman | 15:47 | |
| *** zul has quit IRC | 15:50 | |
| *** hisaharu has joined #openstack-dev | 15:58 | |
| *** dprince has quit IRC | 16:01 | |
| *** jsavak has joined #openstack-dev | 16:04 | |
| *** joesavak has quit IRC | 16:05 | |
| *** joesavak has joined #openstack-dev | 16:06 | |
| *** jsavak has quit IRC | 16:09 | |
| *** rnirmal has quit IRC | 16:12 | |
| *** rnirmal has joined #openstack-dev | 16:12 | |
| openstackgerrit | A change was merged to openstack/openstack-ci: Added a version of the rfc script from gluster. https://review.openstack.org/235 | 16:14 |
| openstackgerrit | Ziad Sawalha proposed a change to openstack/keystone: Additional changes to fix minor service support stuff and increase test coverage. Also making validate token call available using service admin tokens https://review.openstack.org/242 | 16:16 |
| openstackgerrit | A change was merged to openstack/keystone: Additional changes to fix minor service support stuff and increase test coverage. Also making validate token call available using service admin tokens https://review.openstack.org/242 | 16:18 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Tests running on in-memory sqlite db https://review.openstack.org/258 | 16:32 |
| *** zns1 has quit IRC | 16:33 | |
| openstackgerrit | Monty Taylor proposed a change to openstack/openstack-ci-puppet: Added debian package building slave definition. https://review.openstack.org/259 | 16:41 |
| openstackgerrit | Monty Taylor proposed a change to openstack/openstack-ci-puppet: Added debian package building slave definition. https://review.openstack.org/259 | 16:41 |
| *** kbringard has quit IRC | 16:45 | |
| *** bengrue has joined #openstack-dev | 16:50 | |
| *** dprince has joined #openstack-dev | 16:50 | |
| openstackgerrit | Monty Taylor proposed a change to openstack/openstack-ci-puppet: Add rfc.sh. https://review.openstack.org/260 | 16:51 |
| openstackgerrit | A change was merged to openstack/openstack-ci-puppet: Add rfc.sh. https://review.openstack.org/260 | 16:52 |
| *** ohnoimdead_ has joined #openstack-dev | 16:52 | |
| *** ohnoimdead has quit IRC | 16:56 | |
| *** ohnoimdead_ is now known as ohnoimdead | 16:56 | |
| *** rnirmal has quit IRC | 16:56 | |
| openstackgerrit | A change was merged to openstack/openstack-ci-puppet: Added debian package building slave definition. https://review.openstack.org/259 | 16:58 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Keystone manage.py cleanup https://review.openstack.org/261 | 17:01 |
| *** mnaser has joined #openstack-dev | 17:08 | |
| *** zns has joined #openstack-dev | 17:10 | |
| *** kbringard has joined #openstack-dev | 17:10 | |
| *** kbringard has left #openstack-dev | 17:10 | |
| *** Tushar has joined #openstack-dev | 17:11 | |
| *** anotherjesse has joined #openstack-dev | 17:12 | |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-ci: Update blueprint script to add topic links. https://review.openstack.org/262 | 17:15 |
| *** jakedahn has quit IRC | 17:16 | |
| *** jsavak has joined #openstack-dev | 17:18 | |
| *** joesavak has quit IRC | 17:21 | |
| *** ohnoimdead has quit IRC | 17:23 | |
| *** ohnoimdead has joined #openstack-dev | 17:23 | |
| *** glenc has joined #openstack-dev | 17:27 | |
| openstackgerrit | A change was merged to openstack/keystone: Tests running on in-memory sqlite db https://review.openstack.org/258 | 17:33 |
| *** mgius has joined #openstack-dev | 17:36 | |
| openstackgerrit | A change was merged to openstack/keystone: Keystone manage.py cleanup https://review.openstack.org/261 | 17:37 |
| *** dragondm has quit IRC | 17:39 | |
| *** anotherjesse_ has joined #openstack-dev | 17:43 | |
| *** anotherjesse has quit IRC | 17:46 | |
| *** anotherjesse_ is now known as anotherjesse | 17:46 | |
| *** kbringard has joined #openstack-dev | 17:52 | |
| *** kbringard has left #openstack-dev | 17:52 | |
| *** BK_man has joined #openstack-dev | 17:53 | |
| *** dprince has quit IRC | 17:58 | |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-ci-puppet: Rename test repo on gerrit dev server. https://review.openstack.org/263 | 17:59 |
| openstackgerrit | A change was merged to openstack/openstack-ci-puppet: Rename test repo on gerrit dev server. https://review.openstack.org/263 | 17:59 |
| *** zns has quit IRC | 18:02 | |
| *** zns has joined #openstack-dev | 18:02 | |
| *** zns1 has joined #openstack-dev | 18:07 | |
| *** zns has quit IRC | 18:09 | |
| *** dfg has quit IRC | 18:19 | |
| *** sandywalsh has quit IRC | 18:20 | |
| *** rnirmal has joined #openstack-dev | 18:21 | |
| *** dfg has joined #openstack-dev | 18:21 | |
| *** jakedahn has joined #openstack-dev | 18:24 | |
| *** jakedahn has joined #openstack-dev | 18:24 | |
| *** sandywalsh has joined #openstack-dev | 18:24 | |
| *** rnirmal has quit IRC | 18:25 | |
| *** rnirmal has joined #openstack-dev | 18:26 | |
| *** _cerberus_ has quit IRC | 18:33 | |
| *** dragondm has joined #openstack-dev | 18:37 | |
| *** cp16net has quit IRC | 18:38 | |
| *** dprince has joined #openstack-dev | 18:43 | |
| *** bengrue has quit IRC | 18:43 | |
| *** mwhooker has quit IRC | 18:54 | |
| *** Gordonz has quit IRC | 18:55 | |
| *** Gordonz has joined #openstack-dev | 19:04 | |
| *** anotherjesse has quit IRC | 19:07 | |
| *** anotherjesse has joined #openstack-dev | 19:07 | |
| *** zns1 has quit IRC | 19:07 | |
| *** zns has joined #openstack-dev | 19:10 | |
| *** _cerberus_ has joined #openstack-dev | 19:21 | |
| *** ChanServ sets mode: +v _cerberus_ | 19:21 | |
| *** yogirackspace has joined #openstack-dev | 19:23 | |
| *** mwhooker has joined #openstack-dev | 19:23 | |
| mwhooker | Hi all. If anyone has some time I'd appreciate a review of https://code.launchpad.net/~u-matt-h/nova/aws-api-validation I still need to do some functional testing, but it's a pretty good first draft. It resolves https://blueprints.launchpad.net/nova/+spec/aws-api-validation | 19:24 |
| zykes- | aws api ? | 19:27 |
| *** lts has quit IRC | 19:31 | |
| mwhooker | zykes-: how do you mean? | 19:32 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Admin WADL Revisions https://review.openstack.org/264 | 19:33 |
| *** mszilagyi has joined #openstack-dev | 19:47 | |
| smoser | if an instance goes to 'shutdown' is there a way to start it ? | 19:57 |
| *** Guest77784 has quit IRC | 19:59 | |
| *** rnirmal has quit IRC | 20:00 | |
| *** dprince has quit IRC | 20:01 | |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-ci-puppet: Add exim config to all servers. https://review.openstack.org/265 | 20:03 |
| openstackgerrit | A change was merged to openstack/openstack-ci-puppet: Add exim config to all servers. https://review.openstack.org/265 | 20:05 |
| *** BK_man has quit IRC | 20:05 | |
| openstackgerrit | A change was merged to openstack/openstack-ci: Update blueprint script to add topic links. https://review.openstack.org/262 | 20:06 |
| vishy | dprince: ping | 20:07 |
| openstackgerrit | Monty Taylor proposed a change to openstack/openstack-ci-puppet: Prevent add-apt-repository from running every time https://review.openstack.org/257 | 20:08 |
| vishy | blamar_: is dprince around? | 20:11 |
| blamar_ | vishy: not for the moment, can I help out? I think he might be online later tonight, but out of the office | 20:12 |
| *** jc_smith has left #openstack-dev | 20:12 | |
| vishy | blamar_: i had some questions about smokestack. Mostly, I'd like to know how one debugs a failure | 20:13 |
| *** zul has joined #openstack-dev | 20:14 | |
| openstackgerrit | A change was merged to openstack/openstack-ci-puppet: Prevent add-apt-repository from running every time https://review.openstack.org/257 | 20:14 |
| jaypipes | vishy: look at the smokestacktrace? ;) | 20:14 |
| vishy | jaypipes: yeah the outuput is not incredibly useful... :) | 20:15 |
| blamar_ | vishy: depends where/how it fails. you'll get some that say "failed to build packages", "OSAPI tests failed", "smoketests failed", and other messages which give you some information, otherwise you'll need to resort to stdout/stderr | 20:17 |
| blamar_ | Finished in 71.498338 seconds. | 20:17 |
| blamar_ | 1) Error: | 20:17 |
| blamar_ | test_create_server(TestServers): | 20:17 |
| blamar_ | RuntimeError: Timeout trying to ssh to server: 192.168.0.2 | 20:17 |
| * blamar_ apologizes for non-paste | 20:18 | |
| *** kbringard has joined #openstack-dev | 20:18 | |
| jk0 | jaypipes: the def integer() comment made me LOL | 20:30 |
| jk0 | good catch | 20:30 |
| jaypipes | jk0: :) | 20:30 |
| kbringard | hey jaypipes, have you seen this? | 20:31 |
| kbringard | http://paste.openstack.org/show/2191/ | 20:31 |
| kbringard | my migrate version is 7 | 20:32 |
| *** zul has quit IRC | 20:37 | |
| *** Guest77784 has joined #openstack-dev | 20:37 | |
| jaypipes | kbringard: yeah, nothing to do with migrate... add this to your glance-registry.conf, in the filter:context section: | 20:37 |
| jaypipes | context_class = glance.registry.context.RequestContext | 20:38 |
| jaypipes | kbringard: should look like this: | 20:38 |
| jaypipes | [filter:context] | 20:38 |
| jaypipes | context_class = glance.registry.context.RequestContext | 20:38 |
| jaypipes | paste.filter_factory = glance.common.context:filter_factory | 20:38 |
| kbringard | awesome, looks to be working again, thanks homie | 20:39 |
| jaypipes | kbringard: any time :) | 20:40 |
| mtaylor | jaypipes: the glance ppa has packages for python-gflags and python-greenlet - these are not in the nova ppa | 20:43 |
| mtaylor | jaypipes: does glance still need these (or special versions of them?) | 20:44 |
| mtaylor | jaypipes: oh - nevermind- those are lucid packages | 20:44 |
| jaypipes | mtaylor: not gflags. not sure about greenlet. whatever is in Swift for greenlet, Glance would need. | 20:44 |
| jaypipes | mtaylor: ah, l | 20:44 |
| jaypipes | k, even | 20:44 |
| mtaylor | jaypipes: let me re-phrase that question then - do we need to support lucid? | 20:44 |
| jaypipes | mtaylor: yes. | 20:45 |
| mtaylor | jaypipes: ok. great | 20:45 |
| jaypipes | mtaylor: :) | 20:45 |
| dolphm | if anyone is interested, i think the core keystone API has been pretty much nailed down: https://blueprints.launchpad.net/keystone/+spec/identity-api | 20:46 |
| vladimir3p | vishy: ping | 20:46 |
| tr3buchet | has anyone run into issue creating venv? import error no module named nose? | 20:47 |
| vishy | vladimir3p: pong | 20:47 |
| vladimir3p | vishy: are you aware of any plans to support PCI passthrough in nova? | 20:47 |
| vladimir3p | vishy: I will explain: | 20:47 |
| vishy | tr3buchet: ubuntu, osx? | 20:47 |
| vishy | vladimir3p: no plans yet, but it would be cool to support it | 20:47 |
| vladimir3p | vishy: would like to create a new instance_type wth extra spec that will specify that particular instance will need to receive a PCI device of type X | 20:48 |
| vladimir3p | vishy: if this device supports SR-IOV we could find an non-used Virtual Function (VF) and assign it to it | 20:48 |
| vishy | adding pci passthrough would be awesome, it will be a little tricky at the driver layer | 20:48 |
| vladimir3p | vishy: it might be a NIC or SAS HBA, etc... | 20:48 |
| vishy | sounds like a great addition | 20:49 |
| vladimir3p | vishy: ok, great. we will take a look | 20:49 |
| vladimir3p | vishy: as part of it we would like to create "plug-ins" for Cheetah templates ... | 20:49 |
| vladimir3p | or if PCI passthrough will be part of nova we could just add some special code there | 20:50 |
| vladimir3p | (and avoid tricky plugins) | 20:50 |
| Daviey | What bug num was the reporting of all instances, even terminated ones since genisis? | 20:50 |
| vishy | blamar_: mostly i'm trying to figure out why my keystone-migration-ec2 branch is breaking | 20:50 |
| Daviey | it was reported recently | 20:51 |
| vladimir3p | Daviey: it was 824008 | 20:51 |
| Daviey | vladimir3p: thanks! | 20:52 |
| vishy | bug 824008 ...damn | 20:52 |
| uvirtbot | Launchpad bug 824008 in nova "euca-describe-instances shows all instances in all states (even terminated)" [Undecided,In progress] https://launchpad.net/bugs/824008 | 20:52 |
| vladimir3p | Daviey: pls recheck your code against latest nova ver. It might be already fixed by unrelated merge | 20:53 |
| tr3buchet | vishy: ubuntu | 20:54 |
| *** Guest77784 has quit IRC | 20:55 | |
| vishy | tr3buchet: darn | 20:56 |
| tr3buchet | vishy: You were all set with a good osx answer? | 20:56 |
| *** Guest77784 has joined #openstack-dev | 20:56 | |
| vishy | tr3buchet: i know pip-requires fails on lion because it can't install m2crypto | 20:56 |
| *** ameade has quit IRC | 20:57 | |
| Daviey | vladimir3p: Yeah, i just read the bug - thanks for tackling that so quickly | 20:57 |
| vladimir3p | Daviey: u r welcome. it affected us as well. | 20:58 |
| tr3buchet | ah ok | 20:58 |
| tr3buchet | installed a couple of dev libraries i'll see if works now | 20:58 |
| vishy | tr3buchet: has anyone else noticed that nova-api main process exits immediately after forking subprocesses. This seems to be new... | 21:00 |
| vishy | tr3buchet: that wasn't directed specifically to you, but since you are here... | 21:01 |
| *** zul has joined #openstack-dev | 21:01 | |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-ci-puppet: Add gitweb to gerrit servers. https://review.openstack.org/266 | 21:04 |
| *** BK_man has joined #openstack-dev | 21:04 | |
| vishy | currently we are blaming python-kombu | 21:04 |
| redbo | probably something really complicated like photoshop | 21:04 |
| redbo | ack | 21:04 |
| *** ohnoimdead_ has joined #openstack-dev | 21:05 | |
| vishy | redbo: are you using the new openstack-photoshop plugin? | 21:05 |
| *** BK_man has quit IRC | 21:05 | |
| Daviey | vishy: kombu? for nova? | 21:07 |
| Daviey | vishy: I have an initial patch to try to support that, has someone else taken it further? | 21:07 |
| *** zul has quit IRC | 21:07 | |
| *** ohnoimdead has quit IRC | 21:08 | |
| *** ohnoimdead_ is now known as ohnoimdead | 21:08 | |
| redbo | vishy: it's very nice, but I'm not sure I agree with the architecture. It put my image on a queue to a be handled by a gaussian blur worker. | 21:08 |
| *** martine has quit IRC | 21:10 | |
| tr3buchet | vishy: i hadn't noticed ;) | 21:12 |
| vishy | it is apparently a glance dep now | 21:13 |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-ci-puppet: Cleanup. https://review.openstack.org/267 | 21:14 |
| jaypipes | vishy: doesn't nova require kombu, too? | 21:14 |
| vishy | not yet | 21:15 |
| vishy | uses carrot | 21:15 |
| Daviey | jaypipes: no, there is a bug | 21:15 |
| openstackgerrit | Monty Taylor proposed a change to openstack/keystone: Add run_tests.* to the MANIFEST.in. https://review.openstack.org/268 | 21:15 |
| jaypipes | Daviey: I see | 21:16 |
| Daviey | vishy / jaypipes: I did a first pass using the kombu compat with carrot module.. http://pb.daviey.com/Dx2d/ | 21:16 |
| Daviey | not finished yet. | 21:16 |
| openstackgerrit | A change was merged to openstack/openstack-ci-puppet: Add gitweb to gerrit servers. https://review.openstack.org/266 | 21:16 |
| jaypipes | k | 21:16 |
| Daviey | requires a new version of kombu than is in Debian or Ubuntu, which in turn requires a newer ampqlib.. so started to get painful | 21:17 |
| openstackgerrit | A change was merged to openstack/keystone: Add run_tests.* to the MANIFEST.in. https://review.openstack.org/268 | 21:17 |
| mtaylor | dolphm: you are a god | 21:17 |
| dolphm | mtaylor: /salute | 21:17 |
| zykes- | why use gitweb to gerrit ? | 21:18 |
| jeblair | zykes-: if gitweb is installed, gerrit will provide gitweb links to its internal repos | 21:19 |
| ironcamel | i started to look at doing the carrot -> kombu switch. it got to be very hairy. | 21:19 |
| ironcamel | there is a lot of seemingly overcomplicated things going on in amqp.py | 21:19 |
| jeblair | zykes-: can be useful for examining changes, looking at internal states, etc. | 21:19 |
| mtaylor | Daviey: I had started down the road of doing amqplib backports, but then stopped when glance didn't need _that_ new of a version of kombu | 21:19 |
| ironcamel | such as extending the carrot connection class, etc. | 21:19 |
| Daviey | well, i'd like to know if glance does work with the newer upstream version of kombu and ampq. | 21:20 |
| Daviey | I don't want to fix nova to break glance :) | 21:20 |
| zykes- | jeblair: meaning ? | 21:20 |
| mtaylor | Daviey: agree. I'd also spun up a node on which to install newer versions of things to test that ... but then did something else | 21:20 |
| Daviey | i did notice that kombu seems to have a richer fakerabbit.py.. so we might be able to drop the nova one | 21:21 |
| Daviey | ironcamel: If you want to compare patches and keep working on it, i'd most appreciate it :) | 21:21 |
| mtaylor | vishy: we're hanging on jenkins/tarmac in nova test cases again | 21:23 |
| mtaylor | vishy: which is probably something that should be sorted out - is there anything I can do to help track down why/where we're hanging? | 21:23 |
| jeblair | zykes-: it's a feature gerrit provides with no effort on our part that allows you to view changes and tree history in context through the web (before or after being merged). it's not a huge change, just a few gitweb links will show up next to changes. | 21:24 |
| jeblair | zykes-: you can see an example here: https://review-dev.openstack.org/#change,25 | 21:24 |
| mtaylor | dolphm: when I run run_tests.sh from a source tarball, I get: Starting test 1 of 2 with config: sql.conf.template ... [Errno 2] No such file or directory: keystone/test/etc/sql.conf.template | 21:30 |
| anotherjesse | soren: in a talk about training - will ping in an hour about the test stuff - sorry for delay | 21:31 |
| mtaylor | dolphm: nevermind - typing it makes me see the issue | 21:31 |
| dolphm | mtaylor: that file should definitely be there? lol | 21:31 |
| vishy | mtaylor: not sure | 21:32 |
| vishy | mtaylor: tracking down another odd bug at the moment | 21:32 |
| mtaylor | vishy: ok. I'll try to poke at it myself in a bit | 21:32 |
| soren | anotherjesse: I'm on my way to bed, so e-mail me instead :) | 21:33 |
| *** mnaser has quit IRC | 21:34 | |
| anotherjesse | soren: k | 21:34 |
| vishy | why does everyone try to add multiprocessing in with eventlet :( | 21:37 |
| *** jsavak has quit IRC | 21:38 | |
| *** chipc has quit IRC | 21:38 | |
| mtaylor | vishy: because it'll make it just magically work without having to learn about anything? :) | 21:38 |
| notmyname | concurrency is hard! let's add eventlet! | 21:39 |
| zykes- | vishy: what's bad with eventlet ? | 21:39 |
| kbringard | we need to start a quote wall on the wiki | 21:40 |
| *** yogirackspace has left #openstack-dev | 21:41 | |
| *** amccabe has quit IRC | 21:43 | |
| creiht | zykes-: nothing, just the multiprocessing libs and eventlet don't play well together | 21:44 |
| zykes- | ah | 21:44 |
| *** bengrue has joined #openstack-dev | 21:46 | |
| vishy | eventlet is fine, just greenthreads + multiprocessing has bad issues | 21:47 |
| vishy | i spent weeks debugging it in twisted | 21:47 |
| ironcamel | Daviey: do you have a public branch for the kombu stuff? | 21:48 |
| zykes- | twisted vishy and eventlet ? | 21:50 |
| vishy | twisted + multiprocessing | 21:50 |
| zykes- | isn't twisted 1 thread ? | 21:50 |
| Daviey | ironcamel: just the diff pasted above | 21:51 |
| Daviey | ironcamel: I made some other changes to try and prod the test suite through, but nothing worth committing | 21:52 |
| *** mfer has quit IRC | 21:54 | |
| openstackgerrit | Monty Taylor proposed a change to openstack/keystone: Add the files in keystone/test/etc. https://review.openstack.org/269 | 21:58 |
| ironcamel | Daviey: i don't see it. can you link to your patch again? | 21:58 |
| *** mfer has joined #openstack-dev | 22:05 | |
| *** zns has quit IRC | 22:07 | |
| openstackgerrit | A change was merged to openstack/keystone: Add the files in keystone/test/etc. https://review.openstack.org/269 | 22:08 |
| Daviey | ironcamel: it was a first pass, using the kombu compat with carrot module. http://pb.daviey.com/Dx2d/ | 22:11 |
| Daviey | Note, it requires a newer upstream version of kombu than in debian or ubuntu. Which in turn requires ampqlib >= 1.0 IIRC | 22:12 |
| Daviey | newer upstream version, anyway | 22:12 |
| *** kbringard has quit IRC | 22:16 | |
| *** mfer has quit IRC | 22:22 | |
| *** zns has joined #openstack-dev | 22:27 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Admin WADL Revisions https://review.openstack.org/264 | 22:34 |
| *** dolphm has quit IRC | 22:37 | |
| ironcamel | Daviey: i had assumed that the kombu carrot combat approach wouldn't work because of the crazy way we are using carrot | 22:37 |
| ironcamel | s/combat/compat/ | 22:37 |
| *** ohnoimdead has quit IRC | 22:37 | |
| ironcamel | did you have any luck with this? | 22:37 |
| *** kbringard has joined #openstack-dev | 22:38 | |
| *** ohnoimdead has joined #openstack-dev | 22:40 | |
| *** mnaser has joined #openstack-dev | 22:40 | |
| *** zns has quit IRC | 22:41 | |
| openstackgerrit | A change was merged to openstack/keystone: Admin WADL Revisions https://review.openstack.org/264 | 22:41 |
| vladimir3p | vishy: ping | 22:41 |
| vishy | sexified | 22:41 |
| vladimir3p | vishy: :-) any update on volume types/extra-data BP? Should we take it? | 22:42 |
| *** mattray has quit IRC | 22:43 | |
| vishy | vladimir3p: I am devoting all of tomorrow to it | 22:43 |
| vishy | vladimir3p: if you want to break off a chunk and do it it would be very helpful | 22:43 |
| *** zns has joined #openstack-dev | 22:44 | |
| vishy | the api layer stuff is basically just a straight copy from the instances stuff | 22:44 |
| vladimir3p | vishy: cool, thanks. I will check if we have any available resource or I will do it by myself. Just want to know if we should expect anything or just do it | 22:44 |
| vladimir3p | it seems like pretty straight-forward copy/paste, yeah | 22:44 |
| vishy | vladimir3p: I'm pairing with it on a guy from piston, but we have other cleanup related to volumes, so if you want to take some of it I'm all for it | 22:45 |
| vladimir3p | vishy: I've seen their branch for volume-cleanup ... currently it is mainly removal of AOE | 22:46 |
| vladimir3p | not sure what eaxctly their plains WRT cleanup, so I will leave it up to them | 22:46 |
| vladimir3p | but w ecan take types & extra-data | 22:46 |
| vishy | there is a blueprint showing the extra specs | 22:46 |
| vladimir3p | yep | 22:46 |
| vishy | i and there is one for changes as well | 22:47 |
| vishy | if you're curious what we are planning | 22:47 |
| vishy | clean up discovery / ips in the database / chap | 22:47 |
| vishy | mosly | 22:47 |
| vladimir3p | vishy: ok | 22:47 |
| vladimir3p | vishy: pls let me know when you will have any update from Piston. If they've already started we can wait :-) | 22:48 |
| vladimir3p | vishy: also, when you will have a chance, can you pls send a summary of what we've discussed last week? | 22:49 |
| *** ohnoimdead has quit IRC | 22:54 | |
| *** _cerberus_ has quit IRC | 22:55 | |
| mwhooker | is there a quick & easy way to test the ec2 api? (perhaps without installing all of nova) | 22:56 |
| *** kbringard has quit IRC | 22:59 | |
| *** mnaser has quit IRC | 23:01 | |
| *** mnaser has joined #openstack-dev | 23:01 | |
| vishy | vladimir3p: they haven't started on that part yet | 23:02 |
| vishy | tomorrow we are pairing on it | 23:02 |
| vishy | mwhooker: test_cloud in unit tests? | 23:02 |
| *** bsza has quit IRC | 23:02 | |
| mwhooker | vishy: I'm trying to go through the whole msgi middleware stack. doesn't look like this file does that | 23:05 |
| mwhooker | thinking functional tests | 23:05 |
| *** troytoman is now known as troytoman-away | 23:07 | |
| *** ohnoimdead has joined #openstack-dev | 23:08 | |
| *** tushar_ has joined #openstack-dev | 23:10 | |
| *** Tushar has quit IRC | 23:11 | |
| *** mfer has joined #openstack-dev | 23:11 | |
| vishy | mwhooker: it goes through most of it | 23:12 |
| vishy | smoketests will use the api from the outside | 23:12 |
| vishy | but you need a full running install of nova to use them | 23:13 |
| mwhooker | yeah | 23:13 |
| mwhooker | not that far yet | 23:13 |
| *** zns has quit IRC | 23:13 | |
| tushar_ | Hi All, Can some one please review my merge proposal at #link https://code.launchpad.net/~tpatil/nova/add-options-network-create-os-apis/+merge/68292. I would appreciate if you can give feedback ASAP because the merge proposal freeze dead line is approaching closer. | 23:14 |
| mwhooker | I think I'll just throw my patch over the wall with the caveat that it hasn't been functionally tested ;) | 23:15 |
| *** mszilagyi has quit IRC | 23:19 | |
| vishy | blamar_: https://code.launchpad.net/~vishvananda/nova/united-launch-strategy/+merge/71964 | 23:44 |
| vishy | blamar_: I hacked up some of your code there, see my rationale in the merge prop | 23:45 |
| Daviey | ironcamel: It seemed to work in principle, but there are bugs. | 23:45 |
| *** mgius has quit IRC | 23:56 | |
| mwhooker | if anyone has some suggested bugs for me to tackle, let me know | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!