| *** jtdowney has joined #openstack | 00:09 | |
| *** sirp1 has joined #openstack | 00:09 | |
| *** stewart has quit IRC | 00:12 | |
| *** zheng_li has quit IRC | 00:13 | |
| *** metoikos has quit IRC | 00:14 | |
| *** jdmaturen has joined #openstack | 00:15 | |
| *** stewart has joined #openstack | 00:17 | |
| *** littleidea has joined #openstack | 00:31 | |
| *** jsgotangco has joined #openstack | 00:55 | |
| *** RobertLJ has joined #openstack | 00:56 | |
| *** avsm has quit IRC | 00:59 | |
| *** e1mer has joined #openstack | 01:17 | |
| *** miclorb_ has quit IRC | 01:33 | |
| *** miclorb_ has joined #openstack | 01:33 | |
| *** miclor___ has joined #openstack | 01:34 | |
| *** miclorb_ has quit IRC | 01:38 | |
| *** littleidea_ has joined #openstack | 01:48 | |
| *** littleidea has quit IRC | 01:52 | |
| *** e1mer has quit IRC | 01:55 | |
| *** justinsheehy has quit IRC | 02:06 | |
| *** justinsheehy has joined #openstack | 02:12 | |
| *** littleidea_ has quit IRC | 02:16 | |
| *** jtdowney has quit IRC | 02:37 | |
| *** miclor___ has quit IRC | 02:42 | |
| *** jtimberman has left #openstack | 02:45 | |
| *** silassewell has quit IRC | 02:53 | |
| *** RobertLJ has quit IRC | 03:05 | |
| *** ali``` has joined #openstack | 03:09 | |
| *** e1mer has joined #openstack | 03:24 | |
| *** miclorb_ has joined #openstack | 03:43 | |
| *** dendro-afk is now known as dendrobates | 03:46 | |
| *** sirp1 has quit IRC | 03:52 | |
| *** sirp1 has joined #openstack | 04:09 | |
| *** miclorb_ has quit IRC | 04:11 | |
| *** ali``` has quit IRC | 04:31 | |
| *** kashyapc has joined #openstack | 04:33 | |
| *** sophiap has quit IRC | 04:38 | |
| *** sophiap has joined #openstack | 04:39 | |
| *** miclorb has joined #openstack | 04:40 | |
| *** miclorb has quit IRC | 04:50 | |
| *** ThisDB has joined #openstack | 04:54 | |
| *** rsampaio_ has joined #openstack | 04:56 | |
| *** dendrobates is now known as dendro-afk | 04:56 | |
| *** alekibango has quit IRC | 05:02 | |
| *** alekibango has joined #openstack | 05:03 | |
| *** miclorb_ has joined #openstack | 05:13 | |
| *** jsgotangco has quit IRC | 05:16 | |
| *** Mr_T has quit IRC | 05:41 | |
| *** Mr_T has joined #openstack | 05:42 | |
| *** zooko has joined #openstack | 05:42 | |
| *** ibarrera has joined #openstack | 05:52 | |
| *** jdmaturen has quit IRC | 05:56 | |
| *** allsystemsarego has joined #openstack | 06:16 | |
| *** pkhamre has joined #openstack | 06:27 | |
| *** adjohn_ has joined #openstack | 06:38 | |
| *** adjohn has quit IRC | 06:41 | |
| *** adjohn_ is now known as adjohn | 06:41 | |
| *** Glace has quit IRC | 07:00 | |
| *** miclorb_ has quit IRC | 07:00 | |
| *** Glace has joined #openstack | 07:05 | |
| *** mtaylor has joined #openstack | 07:07 | |
| *** ChanServ sets mode: +v mtaylor | 07:07 | |
| *** calavera has joined #openstack | 07:22 | |
| *** almaisan-away is now known as al-maisan | 07:41 | |
| *** rajijoom has joined #openstack | 07:43 | |
| *** zheng_li has joined #openstack | 07:48 | |
| *** rnewson has joined #openstack | 08:05 | |
| *** metoikos has joined #openstack | 08:23 | |
| *** larissa` has joined #openstack | 08:29 | |
| *** tobym has joined #openstack | 08:29 | |
| *** chromakode has quit IRC | 08:30 | |
| *** chromakode has joined #openstack | 08:30 | |
| *** loglaunch has quit IRC | 08:31 | |
| *** loglaunch has joined #openstack | 08:32 | |
| *** larissa` is now known as larissa | 08:32 | |
| *** zheng_li has quit IRC | 08:34 | |
| *** zheng_li has joined #openstack | 08:37 | |
| *** zheng_li has joined #openstack | 08:38 | |
| *** jdmaturen has joined #openstack | 08:41 | |
| *** ganteng has joined #openstack | 08:41 | |
| *** ganteng has left #openstack | 08:42 | |
| *** ganteng has joined #openstack | 08:42 | |
| *** rnewson has quit IRC | 08:44 | |
| *** ganteng has left #openstack | 08:48 | |
| *** ptremblett has quit IRC | 08:53 | |
| *** jdmaturen has quit IRC | 08:58 | |
| *** ibarrera has quit IRC | 09:00 | |
| *** rajijoom has quit IRC | 09:01 | |
| *** rnewson has joined #openstack | 09:03 | |
| *** rnewson has joined #openstack | 09:03 | |
| *** ibarrera has joined #openstack | 09:07 | |
| *** guigui has joined #openstack | 09:08 | |
| *** miclorb has joined #openstack | 09:27 | |
| *** perestrelka has joined #openstack | 09:40 | |
| *** lmcdowell has quit IRC | 09:45 | |
| *** pandemicsyn has quit IRC | 09:47 | |
| *** pandemicsyn has joined #openstack | 09:47 | |
| *** ptremblett has joined #openstack | 09:58 | |
| *** rnewson has quit IRC | 10:19 | |
| *** rnewson has joined #openstack | 10:20 | |
| *** rnewson has joined #openstack | 10:20 | |
| *** sophiap has quit IRC | 10:29 | |
| *** justinsheehy has quit IRC | 10:31 | |
| *** justinsheehy has joined #openstack | 10:37 | |
| *** justinsheehy has quit IRC | 10:55 | |
| *** al-maisan is now known as almaisan-away | 11:11 | |
| *** justinsheehy has joined #openstack | 11:12 | |
| *** ctennis has quit IRC | 11:13 | |
| *** miclorb has quit IRC | 11:16 | |
| *** justinsheehy has quit IRC | 11:21 | |
| *** rajijoom has joined #openstack | 11:25 | |
| *** ctennis has joined #openstack | 11:34 | |
| *** adjohn has quit IRC | 11:41 | |
| *** sophiap has joined #openstack | 11:45 | |
| *** sophiap has quit IRC | 11:48 | |
| *** justinsheehy has joined #openstack | 11:50 | |
| *** guigui has quit IRC | 12:02 | |
| *** guigui has joined #openstack | 12:03 | |
| *** zheng_li has quit IRC | 12:31 | |
| *** lmcdowell has joined #openstack | 12:31 | |
| *** sandywalsh has joined #openstack | 12:52 | |
| *** kashyapc has quit IRC | 12:58 | |
| *** metoikos has quit IRC | 13:01 | |
| *** metoikos has joined #openstack | 13:12 | |
| *** codejunkie has joined #openstack | 13:18 | |
| *** e1mer has quit IRC | 13:23 | |
| *** zheng_li has joined #openstack | 13:27 | |
| *** tonywolf has joined #openstack | 13:30 | |
| *** zheng_li has quit IRC | 13:31 | |
| *** zheng_li has joined #openstack | 13:36 | |
| *** ali``` has joined #openstack | 13:37 | |
| *** tobym has quit IRC | 13:37 | |
| *** zooko` has joined #openstack | 13:40 | |
| *** pvo has joined #openstack | 13:42 | |
| *** ChanServ sets mode: +v pvo | 13:42 | |
| *** zooko has quit IRC | 13:45 | |
| *** gondoi has joined #openstack | 13:47 | |
| *** pcrews has joined #openstack | 13:47 | |
| *** guigui has joined #openstack | 13:50 | |
| *** zooko` is now known as zooko | 13:50 | |
| *** aliguori has joined #openstack | 13:51 | |
| *** rajijoom has quit IRC | 13:51 | |
| *** aliguori has quit IRC | 13:51 | |
| *** aliguori has joined #openstack | 13:51 | |
| *** tonywolf has quit IRC | 13:54 | |
| *** tobym has joined #openstack | 13:58 | |
| *** jdarcy has joined #openstack | 14:04 | |
| *** gundlach has joined #openstack | 14:05 | |
| *** abecc has joined #openstack | 14:19 | |
| *** e1mer has joined #openstack | 14:20 | |
| *** coredump777 has joined #openstack | 14:21 | |
| coredump777 | I am starting a pilot project here on my company, wonder if openstack is already 'usable' for a proof of concept installation. | 14:26 |
|---|---|---|
| gundlach | proof of concept, yes, though the real deal will not be officially released until October. | 14:31 |
| gundlach | another person in here a week or so ago mentioned getting it up and running on some hardware at his company as a proof of concept. | 14:32 |
| *** sirp1 has joined #openstack | 14:32 | |
| *** almaisan-away is now known as al-maisan | 14:32 | |
| gholt | gundlach: You mean just nova? Or? The swift part is already used in a production environment. | 14:32 |
| gundlach | gholt: yep, my mistake, i'm focused on nova. swift is good to go. | 14:32 |
| *** jero has joined #openstack | 14:32 | |
| gundlach | coredump777: (swift == object storage, nova == compute cluster.) | 14:33 |
| jero | hi | 14:33 |
| coredump777 | nova being a version or a component? | 14:33 |
| coredump777 | Ah ok | 14:33 |
| gundlach | :) | 14:33 |
| gholt | Hehe, it'll only get "worse" as we add more stuff to the stack. :) | 14:33 |
| coredump777 | how will this object storage work? Is it based on any known DFS? | 14:33 |
| gholt | It's much like S3 in concept, though the API is a little different here and there. | 14:33 |
| gholt | http://swift.openstack.org/ might help | 14:36 |
| *** johnpur has joined #openstack | 14:37 | |
| *** ChanServ sets mode: +v johnpur | 14:37 | |
| jero | do you miss developer resources on specific parts ? | 14:37 |
| coredump777 | it's the first "cloud" we are trying, so I still studying and seeing how that works | 14:38 |
| jero | is openstack usable yet ? | 14:40 |
| coredump777 | gundlach said "proof of concept, yes, though the real deal will not be officially released until October." | 14:40 |
| jero | is the development code available on the repository anyway ? | 14:41 |
| gholt | jero: I'm not sure we have any specific needs, just general needs. :) Feel free to browse around the bug lists and blueprints or come with your own stuff you'd like to implement. | 14:41 |
| gundlach | jero: we can always use more help in both Object Storage and Compute (swift and nova) | 14:41 |
| gundlach | jero: see wiki.openstack.org to get started. the code is available on launchpad to download and start hacking away upon :) | 14:41 |
| *** mrayzenoss has joined #openstack | 14:43 | |
| *** mrayzenoss is now known as mray | 14:43 | |
| *** veszig has quit IRC | 14:45 | |
| jero | who is the project admin at the moment ? | 14:45 |
| gundlach | jero: it's a shared group -- see launchpad.net/openstack for the admin list | 14:45 |
| *** veszig has joined #openstack | 14:45 | |
| jero | thanks gundlach | 14:47 |
| gundlach | np | 14:47 |
| coredump777 | general question: can I use pg or mysql on applications that run in openstack or amazon ec2/s3? or is it better to run a nosql | 14:52 |
| jero | you can, as long as your app is able to properly deal with scaling issues if needed | 14:52 |
| gundlach | coredump777: the application running in openstack will not know that it's in a virtual environment. It will be running on a full computer, as far as it knows. So that shouldn't change your architectural decisions -- just pretend you were on dedicated hardware. | 14:53 |
| jero | who is the "project manager" being granted all copyright and patent licensing rights in the contributor license agreement ? | 14:56 |
| jero | rackspace ? | 14:56 |
| *** rsampaio_ has quit IRC | 14:57 | |
| coredump777 | jero, this is my concern, like, using cassandra I can add nodes and it will most of the time scale well | 14:57 |
| *** rsampaio has joined #openstack | 14:58 | |
| _0x44 | jero: OpenStack doesn't require a copyright assignment | 14:58 |
| gundlach | jero: i do believe that's Rackspace, but i'm not positive. | 14:58 |
| gundlach | _0x44: well bite my tongue | 14:58 |
| _0x44 | gundlach: Do I have to? People might talk. | 14:58 |
| gundlach | _0x44: pass | 14:59 |
| *** dendro-afk is now known as dendrobates | 14:59 | |
| jero | according to http://wiki.openstack.org/HowToContribute | 14:59 |
| _0x44 | jero: dendrobates will be able to say more (he's the project whozawotzit for OpenStack Compute) | 14:59 |
| gholt | jero: Does this help answer your question? http://bazaar.launchpad.net/%7Ehudson-openstack/swift/trunk/annotate/head%3A/LICENSE | 14:59 |
| _0x44 | jero: Yes, it requires a Contributor License Agreement, but that isn't a Copyright assignment. You maintain your copyright on all code you've written. | 15:00 |
| *** mrayzenoss has joined #openstack | 15:01 | |
| *** mrayzenoss has joined #openstack | 15:01 | |
| jero | okay | 15:01 |
| dendrobates | the CLA basically restates the apache license. I am told by the lawyers, that this is necessary for some countries. | 15:01 |
| jero | i see, thanks for the clarification dendrobates | 15:01 |
| coredump777 | gundlach, I mean, I want a infrastructure that easily can grow and shrink based on the load I am receiving. But using a single instance of pg or mysql can prevent that, and running multiple copies of those DBs is not trivial | 15:02 |
| *** mray has quit IRC | 15:03 | |
| jero | coredump777: using cassandra you can do that at the db level, using a sql you will need to have application level scaling | 15:03 |
| gundlach | coredump777: be aware that OpenStack doesn't buy you that right now. It manages spinning up and tearing down virtual machines images -- the infrastructure below your application. You are correct that if you want your application to scale, you may wish to use nosql. | 15:03 |
| dendrobates | BTW, we wil be changing the CLA to a participant agreement that includes a code of conduct, and will need to be signed by everyone, as recommended by Simon Phipps of the OSI | 15:03 |
| coredump777 | I WILL SIGN NOTHING BWAHAHAHAHAHAHAHAHA | 15:05 |
| coredump777 | sorry | 15:05 |
| *** rsampaio has quit IRC | 15:05 | |
| *** rsampaio has joined #openstack | 15:05 | |
| *** theron has joined #openstack | 15:05 | |
| coredump777 | I want to start this proof of concept and see the development area crying about using a different DB/paradigm | 15:06 |
| coredump777 | <- BOFH moment | 15:06 |
| *** hybridpollo has left #openstack | 15:06 | |
| *** lorin has joined #openstack | 15:08 | |
| *** lorin has left #openstack | 15:08 | |
| *** lorin has joined #openstack | 15:09 | |
| *** e1mer has quit IRC | 15:09 | |
| *** lorin has quit IRC | 15:09 | |
| *** babble has joined #openstack | 15:09 | |
| *** zooko has quit IRC | 15:12 | |
| *** kashyapc has joined #openstack | 15:14 | |
| creiht | mtaylor: is there a way in launchpad to get emails when some submits a LP Question? | 15:14 |
| *** rsampaio has quit IRC | 15:14 | |
| soren | creiht: You can set an answer contact for a project. | 15:15 |
| *** rsampaio has joined #openstack | 15:15 | |
| creiht | soren: ahh cool thanks! | 15:16 |
| *** guigui has quit IRC | 15:17 | |
| mtaylor | creiht: what soren said | 15:17 |
| zul | wheee lawyers | 15:20 |
| coredump777 | you know what look good on a lawyer? | 15:20 |
| *** e1mer has joined #openstack | 15:22 | |
| coredump777 | dobermans | 15:23 |
| dendrobates | actually our lawyers are great, the best I have ever worked with. | 15:24 |
| dendrobates | I usually don't like them too much | 15:24 |
| creiht | I've been fairly happy with how little they have gotten in the way | 15:25 |
| babble | I'd like to start playing with nova. Is it better to use the trunk, or the austin branch, or does it matter? | 15:26 |
| coredump777 | Lawyers are necessary for open source... GPL attacks, patent attacks... without lawyers we would be doomed. | 15:27 |
| eday | babble: you should always use lp:nova (trunk) for the lastest. Right now austin/trunk are the same | 15:29 |
| babble | eday: thanks | 15:29 |
| eday | babble: but at some point austin will freeze, and we'll be working on the next named series | 15:30 |
| *** rsampaio_ has joined #openstack | 15:31 | |
| *** Podilarius has joined #openstack | 15:31 | |
| *** calavera has quit IRC | 15:32 | |
| babble | I'm trying to follow the directions here: http://nova.openstack.org/getting.started.html, but I don't get what "Vendored python libraries (don’t require any installation)" means. I can't tell from the instructions how those libraries (Tornado, Twisted, boto, IPy) end up being installed. | 15:33 |
| *** rsampai__ has joined #openstack | 15:33 | |
| *** rsampaio_ has quit IRC | 15:33 | |
| *** rsampaio has quit IRC | 15:33 | |
| eday | babble: those are old currently, you need to install them manually | 15:33 |
| babble | eday: Gotcha. I also noticed a typo in the NginX config file (server-name -> server_name). | 15:34 |
| *** ibarrera has quit IRC | 15:39 | |
| *** pvo_ has joined #openstack | 15:42 | |
| *** rsampaio has joined #openstack | 15:43 | |
| *** rsampai__ has quit IRC | 15:43 | |
| babble | Ran nova-api, and get an ImportError: No module named gflags. Is python-gflags a dependency? | 15:43 |
| gundlach | babble: try these instructions instead, which i've found to work better (sorry, we move fast and the documentation is lagging behind): http://wiki.openstack.org/InstallationNova20100729 | 15:44 |
| babble | gundlach: thanks | 15:45 |
| *** pvo_ has quit IRC | 15:49 | |
| *** mrayzenoss has quit IRC | 15:52 | |
| *** dendrobates is now known as dendro-afk | 15:53 | |
| *** mrayzenoss has joined #openstack | 15:54 | |
| *** mrayzenoss is now known as mray | 15:54 | |
| *** mray has joined #openstack | 15:54 | |
| *** dendro-afk is now known as dendrobates | 15:55 | |
| *** hazmat has joined #openstack | 15:57 | |
| *** littleidea has joined #openstack | 16:02 | |
| *** jdmaturen has joined #openstack | 16:10 | |
| *** jdmaturen has joined #openstack | 16:10 | |
| *** heckj has joined #openstack | 16:10 | |
| eday | gundlach: so, re: the rsapi comments, I see your arguments (cleaner and what folks expect from routes). The one problem is that you can't use self.request, since with eventlet (or threaded/green threaded/...) application, all threads are sharing the same object, which means self.request won't always be the correct request. We either need to pass it as an arg or create a new controller object for each request somehow (yuck) | 16:16 |
| gundlach | yep, that's exactly what i was going to ask you about -- after i sent the email i realized it wasn't thread safe :) | 16:16 |
| gundlach | one alternate would be for the Router to be able to specify a target=MyControllerClass -- the class, not the instance -- and the router would instantiate it. | 16:17 |
| gundlach | which means the router has to say 'if inspect.isclass(target) target = target()' | 16:17 |
| eday | gundlach: so, I have an example using controller=object(), and clals fine. doesn't convert to a unicode string | 16:18 |
| gundlach | and i don't really like that -- now target= can have either a class or a wsgi app? | 16:18 |
| gundlach | ... | 16:18 |
| gundlach | let me go reverify. i kept getting unicode and dug down to the place where routes converts to a string... | 16:18 |
| eday | gundlach: I think the object just needs to be callable, so if __call__ is defined, you're set | 16:19 |
| eday | not sure why it converts to a string in that case, but I saw that behavior without __call__ | 16:19 |
| babble | How do I get the nova certificates set up so authentication happens properly? I tried to euca-bundle-image and got an "Invalid cert" error. Looking at novarc, I don't seem to have pk.pem or cacert.pem in my nova directory hierarchy. | 16:20 |
| gundlach | eday: yep, you're right, my mistake re needing to instantiate the class before calling the WSGI app. | 16:20 |
| *** sophiap has joined #openstack | 16:21 | |
| *** maplebed has joined #openstack | 16:21 | |
| eday | gundlach: I think i'd still prefer passing in an instance rather than a class, this allows us to keep pooled resources in the object | 16:22 |
| eday | gundlach: of course those could be static class members... hmm | 16:23 |
| gundlach | eh, i'd argue that instantiating a wsgi.Controller class is cheap (__init__ is a no-op) | 16:23 |
| jero | how many active contributors are there in openstack ? is the list of 20 people on launchpad relevant ? | 16:23 |
| eday | jero: that sounds about right, and yes, the LP dev list is relevant (not perfect, but a good starting place) | 16:24 |
| jero | cool! thanks | 16:24 |
| * gundlach can no longer get controller to coerce to str... going slightly crazy maybe | 16:25 | |
| *** rsampaio_ has joined #openstack | 16:26 | |
| eday | gundlach: so, we could still pass in (req, **args) rather than (req, args) if we wanted to keep the wsgi args in there and expand named kv's... gets around the thread-safe issue too. or do you just not want the wsgi/wsgify args in there at all? | 16:27 |
| gundlach | eday: that would work, except then every method would need to accept kwargs. so maybe we would inspect the target method and see whether it accepted kwargs before passing them, which seems a little nasty. or am i misunderstanding you? | 16:27 |
| *** rsampaio has quit IRC | 16:28 | |
| gundlach | ah, pardon, i did misunderstand | 16:28 |
| eday | gundlach: well, thats what you have in there now, **routing_args... I'm talking about just adding in wsgi/wsgify args before them | 16:28 |
| gundlach | right. yeah, that could work. | 16:29 |
| gundlach | what would you think of only passing request as a parameter if 'request' in inspect.getargspec(action)['args'] | 16:29 |
| gundlach | ? | 16:29 |
| gundlach | so actions could just def show(self, id): if they didn't care about the request, or def show(self, id, request) if they wanted it | 16:30 |
| gundlach | [i'm not sure myself that i like this magic, wanted to get your opinion) | 16:30 |
| eday | gundlach: i'm not a fan of that kind of magic :) | 16:30 |
| gundlach | ok :) | 16:30 |
| eday | so, it's really new object per request vs wsgi args for each method | 16:31 |
| gundlach | yes, i think so. where 'require that every method accept a request parameter' is really WSGI wrapped in a decorator that passes in **kwargs. right? | 16:32 |
| gundlach | FWIW, Pylons does the former -- by requiring that your controller by a type, not an instance. | 16:33 |
| eday | gundlach: well, I was thinking every method is a normal wsgi app (take envinron, start_response by default), and then provide a decorator to do the wsgify and expanded kwargs | 16:33 |
| gundlach | (which it then instantiates and then calls as a WSGI app) | 16:33 |
| gundlach | eday: right | 16:33 |
| gundlach | eday: i think that's fine. it still feels really weird to make every method a wsgi app (when the only person calling it is the class itself), but the fact that we can support kwargs makes me like it a lot more than i did :) | 16:34 |
| *** dendrobates is now known as dendro-afk | 16:34 | |
| eday | gundlach: that seems clean to me, although I'll admint I have less python/ruby web framework experience | 16:37 |
| eday | gundlach: do you want to make those changes to your branch? | 16:37 |
| gundlach | eday: yep, doing so now | 16:37 |
| eday | awesome | 16:37 |
| *** sophiap has quit IRC | 16:42 | |
| *** kashyapc has quit IRC | 16:45 | |
| *** sophiap has joined #openstack | 16:45 | |
| gundlach | eday: remind me -- what does the decorator buy us? | 16:47 |
| gundlach | if i'm going to have def show(self, id, request) | 16:48 |
| gundlach | why does it help us to put @wsgify_with_kwargs in front of it or whatever? | 16:48 |
| *** sophiap has quit IRC | 16:48 | |
| gundlach | is the benefit that wsgi.Controller will be able to call my_action_method() instead of my_action_method(**routing_dict_and_request_object)? | 16:49 |
| *** zheng_li has quit IRC | 16:49 | |
| eday | gundlach: the default method def would be def show(self, environ, start_response) (normal wsgi), where the decorator can either wsgify it or wsgify with args | 16:50 |
| eday | gundlach: of course we could just require webob and not need the decorator too | 16:51 |
| gundlach | eday: but since wsgi.Router is calling the wsgi.Controller which is a WSGI app, what does it buy us that the wsg.Controller.__call__ method can call self.action_method as a WSGI app? | 16:52 |
| gundlach | i like WSGI, it's great for breaking things apart and for standardizing communication, but i don't want to blindly adhere to it -- and i don't see how it helps in this situation. | 16:53 |
| gundlach | it means that every action method will have to decorate itself, or else peel its kwargs out of environ manually -- versus doing the WSGI work in one place (wsgi.Controller.__call__) and having all subclass action methods benefit | 16:54 |
| *** RobertLJ_ has joined #openstack | 16:55 | |
| eday | gundlach: I just want a way to pass in 'request' to the methods, and rather than inventing another API, i figured wsgi would be fine. We can just conform to webob req and expanded kwargs though (so no decorator needed) | 16:57 |
| *** kashyapc has joined #openstack | 16:57 | |
| gundlach | k. | 16:58 |
| *** sophiap has joined #openstack | 17:02 | |
| babble | Sorry to ask again... I couldn't determine from http://wiki.openstack.org/InstallationNova20100729 how to generate the X509 certificate for a user. I see the CA's cert and private key in the nova/CA directory, but I don't know what command to use to generate a user's private key and certificate. | 17:04 |
| gundlach | babble: sorry, haven't tried that myself yet | 17:05 |
| babble | gundlach: I thought this was necessary to get started. How do you get a new image onto a fresh install of nova without having the x509 cert? | 17:07 |
| gundlach | babble: i've been doing some pretty specific coding work that hasn't required using nova in full | 17:08 |
| babble | gundlach: If nobody answers on IRC, is this the sort of question somebody would respond to on https://answers.launchpad.net/openstack? | 17:11 |
| gundlach | babble: i haven't used LP much, i'm afraid; when *i* have questions, I ask in here :) | 17:12 |
| *** fsoyer has joined #openstack | 17:12 | |
| eday | babble: yeah, here is probably a better place than on LP. | 17:14 |
| eday | vishy: any idea about babble's q above? | 17:14 |
| babble | I think I figured it out. I think this does it: tools/with_venv.sh bin/nova-manage project zip admin admin creds.zip | 17:15 |
| babble | Based on looking at http://wiki.openstack.org/InstallFromSource | 17:16 |
| *** zheng_li has joined #openstack | 17:17 | |
| *** jamiew has joined #openstack | 17:18 | |
| *** RobertLJ_ has quit IRC | 17:18 | |
| *** metoikos has quit IRC | 17:21 | |
| *** e1mer has quit IRC | 17:26 | |
| *** fsoyer has left #openstack | 17:26 | |
| pvo | sandywalsh: helpdesk confirmed token shipped | 17:30 |
| pvo | you'll probably need to contact them when it is received to reset the pin | 17:30 |
| *** thobbs has joined #openstack | 17:31 | |
| thobbs | I couldn't find the best place to do this, but I just need to point out that there is an error in the Nova "Getting Started" documentation. | 17:32 |
| thobbs | http://nova.openstack.org/getting.started.html#configuration | 17:32 |
| thobbs | In the nginx configuration, the directive should be "server_name", not "server-name" | 17:33 |
| pvo | thobbs: yea, I noticed the same, just hadn't fixed yet either. | 17:34 |
| pvo | shoudl be a quick fix. | 17:34 |
| thobbs | thanks | 17:36 |
| *** ali``` has quit IRC | 17:42 | |
| *** dendro-afk is now known as dendrobates | 17:42 | |
| *** ali``` has joined #openstack | 17:42 | |
| *** kashyapc has quit IRC | 18:06 | |
| *** greenisus has joined #openstack | 18:13 | |
| *** sophiap has quit IRC | 18:17 | |
| *** zooko has joined #openstack | 18:20 | |
| gundlach | eday: so, while i'm waiting for ~gundlach/nova/rsapi to be merged, i'd like to start working on the Images and Flavors controllers. Would the right LP flow be to branch rsapi, wait for rsapi to merge, pull from trunk into my new branch, and then propose a merge into trunk? | 18:26 |
| gundlach | [with the actual writing of code going on somewhere on the new branch somewhere in that time period] | 18:26 |
| hazmat | gundlach, that would work | 18:27 |
| hazmat | gundlach, at least speaking from lp experience, that would minimize your workload and conflicts | 18:27 |
| hazmat | gundlach, an alternative would be to branch trunk, merge rsapi | 18:28 |
| hazmat | gundlach, bzr is pretty flexible | 18:28 |
| gundlach | hazmat: tyvm | 18:29 |
| gundlach | i think i'll go with the former, because my branch has diverged from trunk somewhat and i don't have the bzr chops to know how to merge it without screwing up my merge proposal :) | 18:29 |
| *** sirp1 has quit IRC | 18:30 | |
| hazmat | gundlach, typically if you've got another branch that's a dependency to your merge, you can note that in your merge proposal.. else lp will show the whole diff between your branch and trunk | 18:31 |
| eday | gundlach: yeah, either way works well. you can also use bzr piplines for stuff like this if you want to invest in learning that module | 18:32 |
| eday | gundlach: so rsapi ready to be looked at again? | 18:33 |
| gundlach | eday: yep, rsapi is ready | 18:33 |
| eday | cool | 18:33 |
| eday | gundlach: hmm, I was thining in Controller.__call__, it would be: return method(req, **arg_dict), so method's would take def show(self, req, id)... I'm ok with the kwargs how you have it. Do you think one or the other is simpler? | 18:39 |
| *** ptremblett has quit IRC | 18:40 | |
| gundlach | eday: your approach allows the callee to name the request parameter whatever they wish ('req' or 'request' are probably the only two they'd want to use); my approach lets them put it in any position in the arguments. Either way it'll have to be part of the interface. What do you think? | 18:41 |
| gundlach | We can choose to force them to put it first, or to force them to name it 'req', but we have to choose one | 18:42 |
| *** jero has quit IRC | 18:42 | |
| *** jero has joined #openstack | 18:42 | |
| eday | gundlach: ahh, hmm | 18:44 |
| eday | gundlach: yeah, I guess named params how you have it is the way to go | 18:44 |
| gundlach | k | 18:44 |
| *** sirp1 has joined #openstack | 18:46 | |
| gundlach | [and the third option is for Controller.__call__ to replace "method = getattr(self, action)" with "instance = type(self)(); instance.request = req; method = getattr(instance, action)" | 18:46 |
| gundlach | ] | 18:46 |
| gundlach | and then methods don't include 'req' in their signature. but i think you weren't excited about that approach :) | 18:47 |
| *** ptremblett has joined #openstack | 18:47 | |
| eday | I prefer how you have it now, but I'm just one vote :) | 18:49 |
| gundlach | eh, we should vote in the direction of "getting on with it", so I'll keep it as is :) | 18:50 |
| *** al-maisan is now known as almaisan-away | 18:52 | |
| *** dendrobates is now known as dendro-afk | 19:01 | |
| *** dendro-afk is now known as dendrobates | 19:01 | |
| *** Podilarius has quit IRC | 19:06 | |
| *** pcrews is now known as pcrews_lunch | 19:08 | |
| *** ptremblett has quit IRC | 19:24 | |
| coredump777 | <3 openstack | 19:27 |
| creiht | :) | 19:27 |
| *** avsm has joined #openstack | 19:28 | |
| *** greenisus has quit IRC | 19:35 | |
| *** thobbs has quit IRC | 19:41 | |
| *** pvo has quit IRC | 19:54 | |
| *** babble has quit IRC | 19:57 | |
| *** lew has joined #openstack | 20:06 | |
| *** pvo has joined #openstack | 20:09 | |
| *** pvo has joined #openstack | 20:09 | |
| *** ChanServ sets mode: +v pvo | 20:09 | |
| eday | lew: howdy :) | 20:09 |
| lew | hey stackers | 20:10 |
| gholt | Heh. o/ | 20:11 |
| notmyname | howdy | 20:12 |
| letterj_ | greetings | 20:12 |
| pandemicsyn | hola | 20:12 |
| * creiht bows | 20:12 | |
| *** Woodson has joined #openstack | 20:12 | |
| *** Woodson is now known as yeti | 20:12 | |
| *** ChanServ sets mode: +v pandemicsyn | 20:12 | |
| creiht | eveyone quick... act like you are busy :) | 20:13 |
| pandemicsyn | uh oh yeti is here too | 20:14 |
| * pandemicsyn goes and upgrades something | 20:14 | |
| gholt | Maybe we should make them read all the irc logs before being allowed to post in here. | 20:14 |
| *** letterj_ is now known as letterj | 20:18 | |
| pandemicsyn | lew: cheatsheet of who's who - http://wiki.openstack.org/People | 20:18 |
| *** justinsheehy has quit IRC | 20:18 | |
| *** rnewson has quit IRC | 20:18 | |
| pvo | gundlach is my favorite pic in that list | 20:20 |
| *** sparky_ has joined #openstack | 20:20 | |
| *** Podilarius has joined #openstack | 20:20 | |
| *** almaisan-away is now known as al-maisan | 20:21 | |
| *** rnewson has joined #openstack | 20:23 | |
| *** pcrews_lunch is now known as pcrews | 20:27 | |
| *** miclorb_ has joined #openstack | 20:30 | |
| *** ptremblett has joined #openstack | 20:37 | |
| *** metoikos has joined #openstack | 20:40 | |
| *** gundlach has quit IRC | 20:40 | |
| *** sophiap has joined #openstack | 20:41 | |
| *** allsystemsarego has quit IRC | 20:43 | |
| *** ctennis has quit IRC | 20:56 | |
| *** metoikos has quit IRC | 20:56 | |
| *** gustavomzw has joined #openstack | 21:00 | |
| *** metoikos has joined #openstack | 21:00 | |
| *** justinsheehy has joined #openstack | 21:01 | |
| *** gundlach has joined #openstack | 21:02 | |
| *** aliguori has quit IRC | 21:04 | |
| *** littleidea has quit IRC | 21:04 | |
| *** sparky_ has quit IRC | 21:04 | |
| *** ctennis has joined #openstack | 21:15 | |
| *** ctennis has joined #openstack | 21:15 | |
| *** justinsheehy has quit IRC | 21:15 | |
| *** aliguori has joined #openstack | 21:18 | |
| *** gundlach has quit IRC | 21:25 | |
| *** theron has quit IRC | 21:29 | |
| *** anm_ has joined #openstack | 21:33 | |
| *** justinsheehy has joined #openstack | 21:35 | |
| *** metoikos has quit IRC | 21:37 | |
| *** metoikos has joined #openstack | 21:37 | |
| *** dendrobates is now known as dendro-afk | 21:39 | |
| *** miclorb_ has quit IRC | 21:42 | |
| *** jdarcy has quit IRC | 21:46 | |
| *** littleidea has joined #openstack | 21:47 | |
| polvi | woot: http://cloud.gigaom.com/2010/08/16/why-openstack-is-ready-for-success/ | 21:49 |
| gholt | mtaylor: I can't figure out why hudson is trying to run our functional tests with my usermetadata branch. Any ideas? | 21:56 |
| *** lmcdowell has quit IRC | 21:59 | |
| *** perestrelka_ has joined #openstack | 22:00 | |
| mtaylor | gholt: will look | 22:01 |
| *** perestrelka has quit IRC | 22:02 | |
| *** perestrelka_ is now known as perestrelka | 22:02 | |
| mtaylor | gholt: tarmac runs this: python setup.py test | 22:05 |
| *** al-maisan is now known as almaisan-away | 22:05 | |
| mtaylor | gholt: so whatever python setup.py test runs is what's getting run - is it possible something got changed in the functional tests to make nosetest pick them up by default? | 22:06 |
| gholt | Yeah, I'd guess so, but not sure how since I supposedly just put it all back. :) But thanks for the confirmation; I'll try to remove all my functional testing changes and see how that fares [and try them again in another branch]. | 22:07 |
| mtaylor | gholt: I really do need to get hudson set up to also run those somewhere - but to do so in a sensible manner | 22:07 |
| *** littleidea_ has joined #openstack | 22:11 | |
| *** gondoi has left #openstack | 22:15 | |
| *** littleidea has quit IRC | 22:16 | |
| *** sophiap has quit IRC | 22:18 | |
| gholt | Well, I've now completely removed the new functional tests. I don't see how that'd have an effect since it was blowing up on the old func tests, of which I hadn't changed, hehe. | 22:22 |
| * gholt waits for results... | 22:22 | |
| creiht | heh | 22:22 |
| creiht | failed again | 22:22 |
| gholt | Bah, still failed. I guess I'll make a whole new branch and start adding things one by one until it fails again. | 22:22 |
| gholt | BTW, running 'sudo python setup.py test' does not run those tests on my box that hudson is. Weird. | 22:23 |
| creiht | od | 22:23 |
| creiht | erm odd | 22:23 |
| *** Podilarius has left #openstack | 22:29 | |
| *** sophiap has joined #openstack | 22:31 | |
| *** littleidea_ has quit IRC | 22:31 | |
| mtaylor | werid | 22:31 |
| mtaylor | creiht, gholt don't know if you guys know about http://hudson.openstack.org/job/swift-param/ | 22:32 |
| mtaylor | lets your run branches on the hudson host without submitting them through merge request process | 22:32 |
| gholt | No, had no clue, hehe. | 22:32 |
| creiht | cool | 22:32 |
| *** sophiap_ has joined #openstack | 22:33 | |
| gholt | Do I need a login or something? Or just launchpad creds? | 22:33 |
| mtaylor | gholt: make yourself a user on there and then let me know what the username is | 22:34 |
| mtaylor | (hooking in to launchpad creds is on my todo list) | 22:34 |
| *** sophiap has quit IRC | 22:35 | |
| *** sophiap_ is now known as sophiap | 22:35 | |
| *** justinsheehy has quit IRC | 22:35 | |
| gholt | mtaylor: Ok, gholt it is | 22:36 |
| mtaylor | gholt: k. you're all set :) | 22:37 |
| mtaylor | gholt: you should see a "build now" link on that page | 22:37 |
| *** ali``` has quit IRC | 22:37 | |
| mtaylor | which will let you enter a lp: url and it'll run it for ya | 22:38 |
| gholt | Ah, okay, cool! | 22:38 |
| gholt | Apparently 'something weird' was happening. I downloaded the merge diff from the first branch, made a new branch, patch -p0 <diff, commit, push, and it passes through hudson fine. ? | 22:40 |
| mtaylor | gholt: hehe | 22:41 |
| *** sophiap has quit IRC | 22:41 | |
| gholt | I'm going to try adding my new tests back in. | 22:41 |
| *** justinsheehy has joined #openstack | 22:42 | |
| *** sandywalsh has quit IRC | 22:42 | |
| *** littleidea has joined #openstack | 22:44 | |
| *** mray has quit IRC | 22:48 | |
| gholt | Yeah... no idea. Now pushing it to trunk. | 22:48 |
| *** rsampaio_ has quit IRC | 22:49 | |
| gholt | mtaylor: Okay, I'm confused again. The branch failed merge-proposal-avenue when it passed the direct hudson run. | 22:52 |
| creiht | something there is wacky | 22:52 |
| *** sophiap has joined #openstack | 22:52 | |
| *** justinsheehy has quit IRC | 22:52 | |
| gholt | Other than just me. | 22:52 |
| creiht | mtaylor: Does hudson completly remove the source between tests? | 22:53 |
| creiht | I'm thinking that maybe there is a leftover .pyc from gholt's earlier try that is causing it to try to run the functional tests? | 22:54 |
| *** rnewson has quit IRC | 22:54 | |
| gholt | [might be a <...>/test/functional/__init__.pyc floating around] | 22:54 |
| *** justinsheehy has joined #openstack | 22:58 | |
| *** tobym has quit IRC | 22:59 | |
| *** avsm has quit IRC | 22:59 | |
| *** zheng_li has quit IRC | 23:05 | |
| *** miclorb has joined #openstack | 23:05 | |
| *** pvo_ has joined #openstack | 23:10 | |
| *** pvo_ has joined #openstack | 23:10 | |
| *** ChanServ sets mode: +v pvo_ | 23:10 | |
| *** pvo has quit IRC | 23:13 | |
| *** pvo_ has quit IRC | 23:15 | |
| *** sophiap has quit IRC | 23:19 | |
| *** aliguori has quit IRC | 23:19 | |
| *** sophiap has joined #openstack | 23:20 | |
| mtaylor | creiht: HRM | 23:48 |
| mtaylor | creiht: well, hudson removes the files, but at the moment tarmac doesn't ... lemme see if I can cook something up real quick | 23:48 |
| *** lew has quit IRC | 23:51 | |
| *** johnpur has quit IRC | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!