| *** bcwaldon has joined #openstack-dev | 00:38 | |
| *** bcwaldon has quit IRC | 01:17 | |
| *** bcwaldon has joined #openstack-dev | 01:56 | |
| *** bcwaldon has quit IRC | 02:06 | |
| *** cloudgroups has joined #openstack-dev | 02:17 | |
| *** cloudgroups has left #openstack-dev | 02:17 | |
| *** cloudnod has quit IRC | 03:16 | |
| *** gholt_ is now known as gholt | 04:10 | |
| openstackjenkins | Project nova build #945: SUCCESS in 2 min 41 sec: http://jenkins.openstack.org/job/nova/945/ | 04:18 |
|---|---|---|
| openstackjenkins | Tarmac: nova/auth/novarc.template: Changed NOVA_KEY_DIR to allow symlink support | 04:18 |
| *** Arminder-Office has joined #openstack-dev | 05:08 | |
| *** Arminder has quit IRC | 05:09 | |
| *** Zangetsue has joined #openstack-dev | 05:19 | |
| *** cloudgroups has joined #openstack-dev | 05:29 | |
| *** cloudgroups has left #openstack-dev | 05:29 | |
| *** soren has quit IRC | 07:57 | |
| *** soren has joined #openstack-dev | 07:57 | |
| *** ChanServ sets mode: +v soren | 07:57 | |
| *** reidrac has joined #openstack-dev | 08:37 | |
| ttx | soren: yo | 08:53 |
| soren | ttx: o/ | 08:54 |
| ttx | soren: about the Swift proposed-milestone PPA, could we generate a package from current state of proposed-milestone branch ? | 08:54 |
| ttx | soren: should I just touch the ppa job ? | 08:55 |
| ttx | soren: I'll ping mtaylor to set the review team on the branch. | 08:55 |
| soren | ttx: That should do the trick, yes. | 08:55 |
| * ttx tries | 08:55 | |
| ttx | mtaylor: ping | 08:57 |
| ttx | soren: beh: http://jenkins.openstack.org/view/Swift/job/swift-milestone-ppa/series=oneiric/3/console | 08:58 |
| ttx | soren: the others worked ok. That's good enough for me | 08:58 |
| ttx | soren: looks like those two BMPs should be accepted: | 09:57 |
| ttx | https://code.launchpad.net/~morita-kazutaka/nova/snapshot-volume/+merge/61071 | 09:57 |
| ttx | https://code.launchpad.net/~morita-kazutaka/nova/clone-volume/+merge/62259 | 09:57 |
| ttx | (hmm, that second oine is missing a second nova-core review, since technically jaypipes is no longer nova-core) | 09:58 |
| soren | ttx: I should feed myself first. I'm starving. | 10:37 |
| * soren goes to lunch | 10:37 | |
| *** markvoelker has joined #openstack-dev | 10:58 | |
| *** adiantum has joined #openstack-dev | 11:10 | |
| openstackjenkins | Project nova build #946: SUCCESS in 2 min 40 sec: http://jenkins.openstack.org/job/nova/946/ | 11:58 |
| openstackjenkins | Tarmac: This adds a volume snapshot support with the EC2 api. | 11:58 |
| ttx | yay! | 11:59 |
| notmyname | ttx: soren: thanks for getting the packages ready | 11:59 |
| ttx | notmyname: they have not been "published" in PPA yet, should be done soon | 12:00 |
| notmyname | ttx: I'll look over them this morning and probably will be able to give you an answer around lunchtime (central) | 12:01 |
| ttx | notmyname: cool. Will let you know when the PPA is usable for testing | 12:01 |
| notmyname | thanks | 12:02 |
| *** dprince has joined #openstack-dev | 12:06 | |
| soren | PPA publishing is back. | 12:12 |
| ttx | notmyname: packages published at https://launchpad.net/~swift-core/+archive/milestone-proposed | 12:12 |
| *** ameade has joined #openstack-dev | 12:23 | |
| ttx | mtaylor: ping | 12:29 |
| *** bcwaldon has joined #openstack-dev | 12:32 | |
| *** jaypipes has joined #openstack-dev | 12:42 | |
| dprince | soren/ttx: If we have PPA package changes is it best to file tickets and attach branches? Or just create the branch and ping you guys? | 12:42 |
| ttx | dprince: you mean, Ubuntu packaging changes ? | 12:43 |
| dprince | seems like the code isn't in nova so the tickets don't belong there. | 12:43 |
| dprince | ttx: yes. | 12:43 |
| ttx | dprince: you should file them against the ubuntu packaging of nova... let me pull a link for you | 12:43 |
| ttx | dprince: https://bugs.launchpad.net/ubuntu/+source/nova/+filebug | 12:44 |
| dprince | ttx: thanks. Just found some missing deps with nova-network. It wasn't requiring that dnsmasq and bridge-utils get installed. | 12:47 |
| *** adiantum has quit IRC | 12:47 | |
| ttx | dprince: ok | 12:47 |
| soren | dprince: If those are the only two things, I can just go and fix them now. | 12:49 |
| soren | dprince: If you're doing the patch, make sure you add a dependency on dnsmasq-base, not dnsmasq. | 12:50 |
| dprince | soren: absolutely. | 12:51 |
| dprince | soren: I mean go for it. If not I can push a branch later today. | 12:51 |
| soren | dprince: Er... You're absolutely doing the patch or shold I absolutely just go and fix those two things? :) | 12:51 |
| soren | Ok :) | 12:52 |
| soren | I'm done anyway, just didn't want to spoil your fun :) | 12:52 |
| *** adiantum has joined #openstack-dev | 13:00 | |
| *** adiantum has quit IRC | 13:01 | |
| *** adiantum has joined #openstack-dev | 13:02 | |
| dprince | soren: Thank you sir. | 13:04 |
| *** deshantm_laptop has joined #openstack-dev | 13:05 | |
| *** mattray has joined #openstack-dev | 13:17 | |
| soren | dprince: Oh, thank *you*. | 13:18 |
| dprince | soren: actually. I have one more thing I wanted to talk about. So PPA packages currently require libvirt 0.8.8 (which you build in the PPA). | 13:19 |
| *** lorin1 has joined #openstack-dev | 13:19 | |
| dprince | soren: Libvirt 0.8.8 is a no-go on Cloud Servers (maverick). | 13:19 |
| dprince | soren: the Rackspace Cloud Servers kernel doesn't support it. (2.6.35.4-rscloud) | 13:20 |
| dprince | soren: anyway... I'm not asking you to dig into that mess. | 13:20 |
| dprince | soren: Libvirt 0.8.3 works great however. | 13:21 |
| dprince | soren: Is there an easy way to prevent Libvirt 0.8.8 from getting installed? Right now I'm maintaining a fork of the PPA installers branch. | 13:21 |
| dprince | soren: In my fork I simply remove the line which requires libvirt 0.8.8. I suppose I could patch it as well (to remove that line). | 13:22 |
| dprince | soren: Do you have any better suggestions? | 13:22 |
| soren | dprince: 0.8.3 is buggy. We really do need 0.8.8. | 13:25 |
| *** Zangetsue has quit IRC | 13:25 | |
| *** Zangetsue has joined #openstack-dev | 13:25 | |
| *** statik has joined #openstack-dev | 13:28 | |
| *** foxtrotgulf has joined #openstack-dev | 13:28 | |
| dprince | soren: Sure. I'm not using LXC or anything fancy so 0.8.3 works for my needs but I understand the requirement for 0.8.8 too. | 13:30 |
| dprince | soren: anyway. I've got my work around for now. Just wondering if there is a better way. | 13:30 |
| soren | Sure there is. | 13:30 |
| soren | Stop thinking about "workarounds". Let's *fix* things :) | 13:30 |
| soren | If the version of libvirt we have doesn't work, we shouldn't spend brain cycles trying to think of ways of avoiding installing it. We should fix the problem. | 13:32 |
| dprince | soren: Yeah. Well that fix may cost me time. But long term I'd really like my smoke tests to run on OpenStack... so presumably I could then have at using my own kernel or whatever. | 13:32 |
| dprince | soren: The issue here is that I don't have access to change some of these things (kernel version, etc.) so I'm doing the best I can for now. | 13:33 |
| soren | dprince: If there's somewhere libvirt doesn't work, that's a bug. | 13:34 |
| soren | in libvirt. | 13:34 |
| soren | ...that we should fix. | 13:36 |
| soren | We use libvirt, so if libvirt sucks, we suck. | 13:36 |
| dprince | soren: Right. But in this case I think our kernel just doesn't have the required dependencies. | 13:36 |
| soren | libvirt needs to work regardless. | 13:37 |
| soren | It even used to work, so it's a regression. | 13:37 |
| dprince | soren: Okay. Well let me get dig this error up... | 13:37 |
| jaypipes | dprince, soren: morning | 13:54 |
| mtaylor | ttx: morning. what's up? | 13:56 |
| ttx | mtaylor: hello sir. We recently created https://code.launchpad.net/~hudson-openstack/swift/milestone-proposed for Swift release branches | 13:57 |
| ttx | It needs a review team set. Because Hudson himself won't approve branches. | 13:57 |
| mtaylor | k | 13:57 |
| ttx | mtaylor: swift-core will do, for te moment | 13:57 |
| ttx | mtaylor: btw you should share the secret knock with someone, to double the bus factor. | 13:58 |
| mtaylor | ttx: agree. I'll write up a doc today - but this is done for now :) | 13:58 |
| ttx | mtaylor: cool, thanks | 13:59 |
| ttx | notmyname: https://code.launchpad.net/~ttx/swift/final-1.4.0/+merge/62891 is ready for your approval when you're done sanity-checking the current rc | 14:00 |
| notmyname | ttx: thanks | 14:00 |
| soren | mtaylor: I gues I could have stolen a cookie from Jenkins, but that would be rude. | 14:01 |
| jaypipes | mtaylor: mornin. | 14:02 |
| *** Zangetsue has quit IRC | 14:03 | |
| *** zykes- has quit IRC | 14:08 | |
| dprince | soren: I get these errors when trying to run libvirt 0.8.8 on Cloud Servers (maverick): | 14:14 |
| dprince | soren: http://paste.openstack.org/show/1487/ | 14:14 |
| dprince | jaypipes: good morning. | 14:15 |
| jaypipes | ttx: think it's OK to cut the glance milestone branch at around 3pm today? | 14:15 |
| jaypipes | dprince: hey, I figured out the body_file_seekable thing.. | 14:15 |
| dprince | jaypipes: word | 14:16 |
| jaypipes | dprince: it is a version issue. on my box, my local webob version is 0.9.8 which does not have weobo.Request.body_file_seekable. That was added in 1.0.3. | 14:16 |
| jaypipes | dprince: I'll make a workaround in the code for older webobs. | 14:16 |
| dprince | jaypipes: Cool. I was using Maverick (Cloud Servers). So not too old there. But sounds great. | 14:17 |
| jaypipes | dprince: I'm on Mav too. | 14:17 |
| jaypipes | dprince: funny thing is, 0.9.8 Request object has a make_body_seekable() method instead :) | 14:17 |
| blamar | jaypipes: was it you who had a celery blueprint out a while ago? did that end up being a really difficult transition? | 14:20 |
| *** rnirmal has joined #openstack-dev | 14:21 | |
| jaypipes | blamar: yes, and yes. mostly because of gflags support. | 14:21 |
| blamar | jaypipes: played around with it this weekend, would be awesome IMO but yeah, pain in the ass | 14:21 |
| notmyname | soren: ttx: what is the ppa that will be used for this swift release? do we need to update the docs anywhere? | 14:22 |
| jaypipes | blamar: celery uses a "celery config file" that was incompatible with gflags and we'd end up duplicating a lot of stuff. It's still an interesting blueprint for me, though. | 14:22 |
| soren | notmyname: "this" being 1.4.0? I'd ask ttx :) | 14:22 |
| jaypipes | blamar: would nicely separate the management and configuration of the messaging layer from everything else. | 14:22 |
| blamar | jaypipes: Yeah, I was working with doing a custom 'loader' which using config_from_object which would be populated using flags | 14:23 |
| blamar | jaypipes: What is killing me is the lack of separation, yeah | 14:23 |
| jaypipes | blamar: heh, yeah that might work. though I think having configuration for messaging *entirely* separate from nova flags would also be useful... | 14:23 |
| soren | jaypipes: Why? | 14:25 |
| bcwaldon | jaypipes: could you take a look at the api-pagination-limiting branch? Had to add one more commit | 14:25 |
| blamar | jaypipes: like separate nova-compute-manager.conf? which could be gflags/configparser? | 14:25 |
| blamar | jaypipes, jaypipes, jaypipes | 14:25 |
| bcwaldon | blamar: jaypipes | 14:25 |
| jaypipes | ahhhh | 14:25 |
| *** RobertLaptop has quit IRC | 14:25 | |
| jaypipes | soren: because management of messaging can be done in celery entirely (messaging is more than just Nova. other apps can use the message queue...) | 14:26 |
| jaypipes | soren: for instance, I don't put my my.cnf MySQL config in gflags, now do I? ;) | 14:30 |
| soren | jaypipes: Perhaps I'm missing something. | 14:32 |
| soren | jaypipes: celerey is a client, is it not? | 14:32 |
| soren | celery, even. | 14:32 |
| blamar | soren: ehhh, it's ... complicated? depends how you use it, but celery provides celeryd, which is a service | 14:34 |
| soren | Oh, I see. | 14:34 |
| blamar | it would replace all of the custom AMQP/queuing code we have in the project for the managers | 14:35 |
| jaypipes | soren: no, it's a little more than that. it's a way of abstracting the messages from the messaging layer. it allows you to have a worker daemon (celeryd) that listens for your particular messages and communicates with the MQ directly. Then, in your code, you write the code that should be executed as a remote task like normal, and simply decorate it with the @task decorator. | 14:35 |
| jaypipes | soren: course, it's a little more complicated than that, but you get the picture. it's kind of like ripping out nova/rpc.py, running that code as a daemon, and allowing remote RPC tasks in the code to be written in a more structured way with decorators... | 14:36 |
| soren | Ok. | 14:37 |
| jaypipes | soren: and the benefit is that we don't have to write all the custom RPC code :) | 14:37 |
| *** deshantm_laptop has quit IRC | 14:37 | |
| soren | Now *that* sounds lovely. | 14:37 |
| *** foxtrotdelta has joined #openstack-dev | 14:37 | |
| jaypipes | soren: ya. reinvented wheels and all that.. | 14:37 |
| soren | jaypipes: You just make it sound like a goal in itself to have the configuration of messaging outside of gflags, and I don't entirely get that. | 14:38 |
| *** foxtrotgulf has quit IRC | 14:39 | |
| soren | jaypipes: There's a good chance people aren't already using celery. A very good chance, even. So, to start using Nova, they need to not just deal with nova.conf, but also this other thing. This doesn't seem like a win. | 14:39 |
| soren | I mean... It makes sense! I just don't see why it can be a goal in itself. | 14:39 |
| ttx | notmyname: It will be in the "last milestone" PPA that we need to create. Not sure what the current docs point to | 14:40 |
| jaypipes | soren: understood. | 14:40 |
| jaypipes | soren: I push the idea of celery because it allows us to focus on the tasks, and not the implementation of consumers and publishers. the fact that it allows the messaging stuff to be configured separately is a side benefit. | 14:41 |
| *** zykes- has joined #openstack-dev | 14:41 | |
| ttx | notmyname: we might need to update Swift install docs with a PPA map like http://wiki.openstack.org/NewNovaInstall | 14:41 |
| ttx | (this one is work in progress) | 14:41 |
| soren | jaypipes: Right, and that's the part I don't get. Why is it a side "benefit"? | 14:42 |
| *** glenc has quit IRC | 14:43 | |
| notmyname | ttx: ah. ok. should we wait to release until that's ready or cut the release and backport or simply add it to the 1.4.1 tree and go on as normal? | 14:43 |
| ttx | notmyname: We can release now, it will be avilable in the proposed-milestone PPA. Then we'll setup the "last milestone" PPA and copy the package there as "1.4.0" rather than "1.4.0~20110531.r303" or whatever | 14:44 |
| *** glenc has joined #openstack-dev | 14:46 | |
| *** zykes- has quit IRC | 14:46 | |
| ttx | notmyname: I mean, we can accept the final versioning branch, which will result in a 1.4.0~20110531.r303 tarball, that I'll rename as swift-1.4.0.tar.gz and promote on the LP download page | 14:46 |
| creiht | jaypipes: just curious, have you looked at kombu? | 14:47 |
| ttx | notmyname: then we can setup the "last milestone" PPA and push that 1.4.0 tarball to it, though it will be te same as the 1.4.0~20110531.r303 build that will appear in the proposed-milestone PPA. If that makes more sense. | 14:47 |
| jaypipes | soren: I guess I just prefer it that way :) I like that http://ask.github.com/celery/configuration.html allows you to configure all your messaging stuff separately from any individual application config. Kind of like I can configure all my DB settings in my my.cnf and not in my application code itself. | 14:48 |
| blamar | creiht: celery uses kombu, and provides more on top | 14:48 |
| creiht | ahh | 14:48 |
| jaypipes | soren: but it may just be personal preference :) | 14:48 |
| notmyname | ttx: I was referring to the auto-generated docs. those should stay with trunk, although we should probably add something to the top of them to point to a wiki page or the release download location. I think docs are ok for now | 14:49 |
| jaypipes | creiht: yeah, I think you'd pinged me about that early on when you told me about celery :) | 14:49 |
| ttx | notmyname: ack | 14:49 |
| notmyname | ttx: also, the packaging looks fine to us. I'll approve the merge unless you have another reason to wait | 14:49 |
| creiht | blamar: ahh, I knew the same guy worked on both, didn't realize that celery was backended with it | 14:49 |
| * ttx makes mental note | 14:49 | |
| creiht | jaypipes: hah | 14:49 |
| blamar | creiht: Yeah, it's a pretty impressive project | 14:49 |
| * jaypipes thanks creiht for being the genesis man. | 14:50 | |
| creiht | I don't have a lot of experience with either, but one my coworkers has been playing with kombu | 14:50 |
| notmyname | ttx: soren: any outstanding issues from your end that should cause me to wait to approve swift 1.4.0 final? | 14:51 |
| *** glenc has quit IRC | 14:52 | |
| *** zykes- has joined #openstack-dev | 14:52 | |
| soren | notmyname: Nope. | 14:54 |
| ttx | notmyname: no remaining issue from my pov | 14:54 |
| notmyname | soren: ttx: 1.4.0 final approved | 14:54 |
| ttx | notmyname: I'll let the Jenkins magic operate and promote the rtesulting tarball. Will send an email when I'm done. | 14:55 |
| notmyname | yay. thanks :-) | 14:55 |
| *** glenc has joined #openstack-dev | 14:55 | |
| soren | dprince: Ok, found the problem in libvirt. | 14:55 |
| dprince | soren: cool. Simple fix? | 14:56 |
| soren | dprince: Possibly. It just looks intentional, so I'm trying to work out what's going on. | 14:58 |
| soren | dprince: If it's not intentional, then yes, very much a simple fix. | 14:58 |
| *** jkoelker has joined #openstack-dev | 14:58 | |
| dprince | soren: Is there a ticket I can follow? | 14:59 |
| soren | dprince: Don't know. Didn't check. | 14:59 |
| soren | dprince: I prefer spending half an hour looking at code to spending half an hour enduring bugzilla any day of the week. | 15:00 |
| dprince | soren: Okay. Anyway. Thanks for poking at it. I've seen a couple other people hit this issue so I know people are hitting this. | 15:00 |
| *** Zangetsue has joined #openstack-dev | 15:00 | |
| *** vikasg has quit IRC | 15:03 | |
| *** avati has quit IRC | 15:03 | |
| * soren runs off (late) | 15:05 | |
| jaypipes | dabo, gholt: nice thread you have going :) | 15:06 |
| *** bcwaldon has quit IRC | 15:08 | |
| *** avati has joined #openstack-dev | 15:11 | |
| *** vikasg has joined #openstack-dev | 15:11 | |
| *** jkoelker has quit IRC | 15:12 | |
| *** jkoelker has joined #openstack-dev | 15:12 | |
| *** dragondm has joined #openstack-dev | 15:14 | |
| gholt | jaypipes: Heh, yeah, that's my fault. I was being nosy and should just butt out, lol. | 15:15 |
| jaypipes | gholt: not at all. I'm learning a lot from you too. Nothing wrong with those kinds of discussions, IMHO. | 15:15 |
| jaypipes | you, two | 15:15 |
| jaypipes | where'd my grammar go today? | 15:15 |
| gholt | :) | 15:16 |
| *** reidrac has quit IRC | 15:16 | |
| *** RobertLaptop has joined #openstack-dev | 15:21 | |
| *** galstrom has joined #openstack-dev | 15:28 | |
| *** foxtrotdelta has quit IRC | 15:33 | |
| *** Zangetsue has quit IRC | 15:35 | |
| jaypipes | dprince, markwash: pls do review https://code.launchpad.net/~rackspace-titan/glance/api-pagination-limiting/+merge/61155 if you have some time. It's the priority for the D1 milestone release. | 15:41 |
| jaypipes | dprince, markwash: thx in advance! | 15:41 |
| *** foxtrotgulf has joined #openstack-dev | 15:42 | |
| ttx | notmyname: ok, it's out, just sent the email to the list. | 15:42 |
| notmyname | thanks :-) | 15:43 |
| gholt | Hopefully nothing important happened with bugs in the past ten minutes. | 15:45 |
| * ttx will be back in a couple hours. | 15:45 | |
| *** foxtrotdelta has joined #openstack-dev | 15:45 | |
| ttx | notmyname: congrats on your first release as Swift PTL :) | 15:46 |
| *** foxtrotgulf has quit IRC | 15:47 | |
| *** johnpur has joined #openstack-dev | 15:48 | |
| *** ChanServ sets mode: +v johnpur | 15:48 | |
| Vek | can I get some eyes on https://code.launchpad.net/~klmitch/nova/os_int_tests/+merge/61474 ? It's been languishing for lack of review... | 15:57 |
| *** vladimir3p has joined #openstack-dev | 16:06 | |
| vladimir3p | Hi Folks, I have a question related to multi-nic support - Is it allowed to place a VM into different networks? There was a blueprint for it in cactus and couple blueprints for diablo, but will it be allowed to have a user VM sitting in different networks - its private vlan and at the same time public network? | 16:12 |
| jaypipes | dang it's hot in Ohio today... | 16:26 |
| *** openstackjenkins has quit IRC | 16:35 | |
| *** openstackjenkins has joined #openstack-dev | 16:35 | |
| jaypipes | tr3buchet: see vladimir3p's ? above. | 16:36 |
| openstackjenkins | Project nova build #947: SUCCESS in 2 min 53 sec: http://jenkins.openstack.org/job/nova/947/ | 16:39 |
| openstackjenkins | Tarmac: Fixed the APIError typo | 16:39 |
| vladimir3p | jaypipes: will try to shed more light to reason of this Q. We are working on couple of blueprints (including admin VM) and networking part one of essential ones... we definitely would like to have admin VMs connected not only to user's VLANs, but to the entire nova network | 16:45 |
| vladimir3p | in general, it should be kind of similar to VPN... but with access not only from outside and within vlan, but to other projects/vlans as well... | 16:46 |
| jaypipes | vladimir3p: completely understand your use case. tr3buchet has handled the implementation so far, and I don't know much about the networking internals, sorry! :) | 16:51 |
| vladimir3p | np, will wait for his answer ... seems like this channel is pretty silent today :-) | 16:52 |
| *** jdurgin has joined #openstack-dev | 16:57 | |
| *** blamar has quit IRC | 16:59 | |
| *** mgius has joined #openstack-dev | 17:13 | |
| *** blamar has joined #openstack-dev | 17:15 | |
| *** adiantum has quit IRC | 17:37 | |
| *** agarwalla has joined #openstack-dev | 17:48 | |
| *** dysinger has joined #openstack-dev | 18:07 | |
| openstackjenkins | Project nova build #948: SUCCESS in 2 min 53 sec: http://jenkins.openstack.org/job/nova/948/ | 18:14 |
| openstackjenkins | Tarmac: Add vnc_keymap flag, enable setting keymap for vnc console and fix bug #782611. | 18:14 |
| uvirtbot | Launchpad bug 782611 in nova "vnc keymap should be configurable" [Wishlist,In progress] https://launchpad.net/bugs/782611 | 18:14 |
| *** vladimir3p has quit IRC | 18:32 | |
| *** vladimir3p has joined #openstack-dev | 18:32 | |
| *** vladimir3p has quit IRC | 18:33 | |
| *** vladimir3p has joined #openstack-dev | 18:33 | |
| *** vladimir3p has left #openstack-dev | 18:33 | |
| *** vladimir3p has joined #openstack-dev | 18:34 | |
| *** dragondm has quit IRC | 18:36 | |
| ttx | The following merge proposals could use one final core review for final approval: | 18:44 |
| ttx | https://code.launchpad.net/~nttdata/nova/kvm-pause-suspend/+merge/62255 | 18:44 |
| ttx | https://code.launchpad.net/~xtoddx/nova/provider-fw-rules/+merge/62517 | 18:44 |
| ttx | https://code.launchpad.net/~morita-kazutaka/nova/clone-volume/+merge/62259 | 18:44 |
| ttx | https://code.launchpad.net/~rackspace-titan/glance/api-pagination-limiting/+merge/61155 | 18:45 |
| soren | comstud: Do you think you could be pursuaded to spend 15 minutes adjusting the test suite for your eventlet patch? | 18:45 |
| soren | comstud: I spent some time staring at it today to no avail. | 18:45 |
| soren | comstud: I can't work out if your code is wrong or if the test suite is wrong. | 18:45 |
| soren | comstud: Or both, I suppose. | 18:45 |
| *** dragondm has joined #openstack-dev | 18:45 | |
| ttx | soren: looks like you could have a look at the provider-fw-rules one | 18:46 |
| ttx | since it ouches things you used to work on :) | 18:46 |
| ttx | touches* | 18:46 |
| ttx | nova-core (anotherj1sse, _0x44, devcamcar, dabo, jk0, markwash, _cerberus_, pvo, sandywalsh, termie, xtoddx, tr3buchet, vishy): ^^ | 18:49 |
| dprince | ttx: I'll run smoke tests on a few of them and then check them out. | 18:51 |
| ttx | dprince: oh, missed you, you were hiding... | 18:52 |
| *** bcwaldon has joined #openstack-dev | 18:54 | |
| ttx | Meeting in two hours in #openstack-meeting | 19:00 |
| ttx | vishy: Thanks for your review, any reason why you didn't flip the approved bit on kvm-pause-suspend ? | 19:01 |
| vishy | i noticed that dprince said he wanted to smokestack 'em | 19:02 |
| vishy | so i was going to give him a chance before i approved them all from under him :) | 19:02 |
| ttx | if it moves, smokestack it. | 19:02 |
| dprince | vishy: I expect that one to run fine. But might as well give em a go. | 19:03 |
| dprince | ttx: Yessir. That is the idea. Honestly I'd like more buy in to this sort of thing. | 19:03 |
| dprince | Honestly I sent an email to the list. Nobody expect vish has hit me up for an account.... | 19:03 |
| ttx | dprince: honestly I'd like to see it integrated in pre-merge tests, so that you don't have to require an account. | 19:04 |
| openstackjenkins | Project nova build #949: SUCCESS in 2 min 49 sec: http://jenkins.openstack.org/job/nova/949/ | 19:04 |
| openstackjenkins | Tarmac: This enables us to create a new volume from a snapshot with the EC2 api. | 19:04 |
| dprince | And well it is free. | 19:04 |
| vishy | ttx: I really think we need a functional testing team | 19:04 |
| vishy | we have too many efforts and it needs a whole team to manage it | 19:04 |
| ttx | vishy: I'm not involved in it but from where I stand it doesn't seem to be lacking manpower | 19:05 |
| dprince | ttx: We could probably arrange that. I see no problem going ahead and running anything that 'Needs review' via an automated system. | 19:06 |
| ttx | dprince: right, the thing could grab proposed branches and comment whether it passes smoketests or not. | 19:06 |
| jk0 | any of you getting this failure in trunk? http://paste.openstack.org/show/1493/ | 19:07 |
| vishy | ttx: the problem is that there are multiple efforts | 19:07 |
| dprince | ttx/vishy: This is something I really care about (large functional testing). So many issues can't be found in any other way. | 19:07 |
| ttx | dprince: that said, it probably duplicates somethign mtaylor's team is doing | 19:07 |
| dprince | ttx: I agree that we aren't all working together here. I needed a system I had access to. It wasn't clear to me that everyone would have access to run tests via that Hudson box. | 19:07 |
| vishy | jk0: haven't seen that one. Are you running tests on a box that is also running a nova install? | 19:08 |
| ttx | vishy: how do you think that can be solved ? A specific functional testing regular meeting for everyone to coordinate ? | 19:08 |
| jk0 | vishy: yeah | 19:08 |
| dprince | ttx: in any case I think I was really going for something a bit more custom than what Hudson can provide. | 19:08 |
| jk0 | in my compute node/dev node | 19:08 |
| vishy | dprince, ttx: main efforts seem to be dprince, kmitchell, mtaylor, and old smoketests which are probably starting to bit-rot | 19:09 |
| vishy | and stuff termie did | 19:09 |
| ttx | vishy: and stuff soren did | 19:09 |
| dprince | vishy: My game plan was to run it all. I want the ability to run anybody's functional testing suite via a system like Smokestack. | 19:10 |
| jk0 | vishy: ahh, the problem is when /etc/nova/nova.conf exists | 19:10 |
| vishy | dprince: that is awesome but we need coordination on this. It is clearly bigger than a one person project | 19:11 |
| dprince | vishy: What Smokestack really does is define a way to manage all the configurations, projects, test suites, etc that we need to make this process sane. | 19:11 |
| ttx | vishy: I'll raise the subject during the meeting tonight. But I agree the interested parties could form a team, or at least gather in a regular meeting to discuss progress | 19:11 |
| vishy | regular meeting would be a good start | 19:11 |
| ttx | vishy: obviously that doesn't require everyone to be present, only those that have strong feelings on it, or a horse already in the race | 19:12 |
| dprince | vishy: I'd like to coordinate. Like I said the other day I'm actually a bit frustrated with it at the moment. Most of I work on it so far has been after hours. I do see a lot of value in this sort of thing though. | 19:12 |
| vishy | dprince: ++ I think it is the most important thing in the project right now | 19:12 |
| dprince | Actually. I'd be interested in attending a meeting on this (in IRC). Can't make the meeting tonight though. | 19:12 |
| dprince | The tuesday meeting time is always bad for me. | 19:13 |
| vishy | dprince: ok we will try and schedule a weekly functional testing meeting | 19:13 |
| ttx | vishy: added to meetign agenda | 19:15 |
| soren | Oh, no. | 19:23 |
| ttx | no? | 19:25 |
| johan___ | is XML support in a state of flux recently? the results i'm getting vary from the docs quite a bit | 19:25 |
| johan___ | and in some cases seem buggy | 19:25 |
| soren | ttx: More late night meetings. | 19:25 |
| ttx | soren: not necessarily. We can mandate that no west coast or Japanese people are interested in QA. | 19:26 |
| ttx | Alternatively, we could say no European cares. | 19:26 |
| bcwaldon | johan___: very much so. XML suppor is not complete. Hopefully we can start beefing that up soon after the osapi-serialization branch goes in | 19:27 |
| johan___ | bcwaldon: ok, that's good to know | 19:27 |
| vishy | soren: :) | 19:27 |
| soren | ttx: Or everyone could be free to join the meeting at noon UTC. | 19:29 |
| ttx | soren: that could reduce duplication of effort quite effectively. | 19:29 |
| vishy | ttx, soren: :) | 19:31 |
| *** mgius has quit IRC | 19:33 | |
| *** mgius has joined #openstack-dev | 19:35 | |
| johnpur | vishy, ttx: mtaylor should be the hub of the testing discussions. i suggest that he be "volunteered" to run the coordination meeting. Additionally, we are looking to double the "monty" bandwidth on tooling automation, qa automation, and automated build workflow by bringing another senior person on-board. | 19:40 |
| ttx | johnpur: hopefully he will attend the meeting in 75 minutes. | 19:43 |
| openstackjenkins | Project nova build #950: SUCCESS in 2 min 45 sec: http://jenkins.openstack.org/job/nova/950/ | 19:44 |
| openstackjenkins | Tarmac: Added the filtering of image queries with image metadata. This is exposing the filtering functionality recently added to Glance. Attempting to filter using the local image service will be ignored. | 19:44 |
| ttx | yay! | 19:46 |
| vishy | johnpur: that would be great. | 19:47 |
| jaypipes | bcwaldon: or we can just forget about XML :P | 19:51 |
| jaypipes | bcwaldon: I found it funny that Mark Nottingham, one of the folks behind the XML spec, recommends not even bothering with XML as it produces more problems than it solves. | 19:52 |
| ttx | jaypipes: http://twitter.com/#!/colmmacuait/status/74107450332168193 | 19:55 |
| jaypipes | ttx: heh | 19:57 |
| *** mgius has quit IRC | 19:57 | |
| *** mgius has joined #openstack-dev | 19:57 | |
| *** adiantum has joined #openstack-dev | 20:03 | |
| bcwaldon | jaypipes: I'm right there with you | 20:05 |
| dprince | vish: So all the branches ttx mentioned look fine except Todd's. | 20:11 |
| dprince | vish: Smokestack caught an issue where the migration was off. Kind of cryptic but sometimes failures aren. | 20:12 |
| sandywalsh | annegentle, any chance of a peek at https://code.launchpad.net/~sandy-walsh/nova/zones-dev-docs/+merge/62870 to help people with the other dist-sched merges? | 20:12 |
| dprince | are. | 20:12 |
| dprince | vish: Anyway. Feel free to pull the trigger on the rest. | 20:12 |
| soren | dprince: https://bugs.launchpad.net/bugs/790837 by the way | 20:13 |
| uvirtbot | Launchpad bug 790837 in libvirt "Fails to start inside PV domU" [Medium,Triaged] | 20:13 |
| dprince | soren: thank you sir. | 20:14 |
| soren | I'll catch up with one of the libvirt guys tomorrow. | 20:14 |
| openstackjenkins | Project nova build #951: SUCCESS in 2 min 42 sec: http://jenkins.openstack.org/job/nova/951/ | 20:19 |
| openstackjenkins | Tarmac: Update the rebuild_instance function in the compute manager so that it accepts the arguments that our current compute API sends. | 20:19 |
| openstackjenkins | Fixes to the SQLAlchmeny API such that metadata is saved on an instance_update. Added integration test to ensure that instance metadata is updated on a rebuild. | 20:19 |
| *** ameade has quit IRC | 20:23 | |
| sandywalsh | s1rp, dabo "No proposals found for merge of lp:~rconradharris/nova/dist-sched-1 into lp:nova." ... any ideas? | 20:24 |
| dabo | did that get merged? | 20:24 |
| *** dprince has quit IRC | 20:24 | |
| sandywalsh | y | 20:24 |
| s1rp | sandywalsh: i thought it did | 20:24 |
| dabo | maybe since it's marked as a dependency? | 20:25 |
| sandywalsh | I'll reissue without the dependency | 20:25 |
| sandywalsh | k, re-approve please | 20:26 |
| ttx | Team meeting starts in 30 minutes in #openstack-meeting ! | 20:28 |
| mtaylor | ttx: sorry - had stepped away to grab coffee for a moment ... yes, I will be at the meeting | 20:28 |
| ttx | kewl | 20:29 |
| * ttx pauses for a bit | 20:29 | |
| bcwaldon | I've seen that error before, and the code got merged but hudson just handled that error reporting incorrectly | 20:29 |
| devcamcar | hey gang | 20:29 |
| devcamcar | saw that keystone is getting put into swift | 20:30 |
| devcamcar | thats awesome | 20:30 |
| devcamcar | i'm adding swift support to the dashboard | 20:30 |
| devcamcar | so once keystone is in there i can un-evil-hack the auth stuff | 20:30 |
| openstackjenkins | Project nova build #952: SUCCESS in 2 min 42 sec: http://jenkins.openstack.org/job/nova/952/ | 20:34 |
| openstackjenkins | Tarmac: added pause/suspend implementation to nova.virt.libvirt_conn | 20:34 |
| annegentle | sandywalsh: sure, I'll take a look thanks for the pointer to it | 20:36 |
| sandywalsh | s1rp, dabo ready for a re-approve https://code.launchpad.net/~sandy-walsh/nova/dist-sched-2a/+merge/63039 | 20:39 |
| sandywalsh | annegentle, thx! | 20:39 |
| sandywalsh | annegentle, where should images go for the .rst files? | 20:39 |
| sandywalsh | annegentle, same directory? | 20:39 |
| annegentle | sandywalsh: doc/source/images | 20:39 |
| sandywalsh | annegentle, and it's a relative path to it? ../images or ? | 20:40 |
| dabo | sandywalsh: sure, I'll get to that next | 20:40 |
| annegentle | sandywalsh: like so: .. image:: images/novashvirtually.png | 20:41 |
| annegentle | sandywalsh: in an RST file | 20:41 |
| sandywalsh | annegentle, ah, perfect. Thx | 20:41 |
| annegentle | sandywalsh: np | 20:41 |
| sandywalsh | annegentle, I'll add some illustrations at the start of next sprint | 20:41 |
| annegentle | sandywalsh: awesome. I am now going to see if the dating service metaphor holds up :) | 20:41 |
| sandywalsh | annegentle, heh | 20:42 |
| annegentle | sandywalsh: my fav. technical editor still refers to a metaphor of velvet ropes and red carpets and gawd I don't even remember what else. Classic. | 20:42 |
| sandywalsh | nice ... I'll have to shoot for that next time :) | 20:43 |
| *** lorin1 has quit IRC | 20:58 | |
| ttx | Meeting starting in one minute in #openstack-meeting, please join | 20:59 |
| ttx | jaypipes: join ? | 21:00 |
| *** rnirmal has quit IRC | 21:03 | |
| openstackjenkins | Project nova build #953: SUCCESS in 2 min 44 sec: http://jenkins.openstack.org/job/nova/953/ | 21:24 |
| openstackjenkins | Tarmac: Basic hook-up to HostFilter and fixed up the passing of InstanceType spec to the scheduler. | 21:24 |
| openstackjenkins | Project nova build #954: SUCCESS in 2 min 41 sec: http://jenkins.openstack.org/job/nova/954/ | 21:48 |
| openstackjenkins | Tarmac: Adds a few more status messages to error states on image register for the ec2 api. This will hopefully provide users of the ec2 api with a little more info if their registration fails. | 21:48 |
| *** galstrom has quit IRC | 21:50 | |
| vladimir3p | Hi! Sorry, but I will repeat my question from this morning... | 22:10 |
| vladimir3p | trying to understand state of multi-nic | 22:10 |
| vladimir3p | in particular ability to have VMs connected to different types of networks | 22:11 |
| vladimir3p | I mean to regular private VLANs to public/cross VLANs (entire infrastructure) | 22:11 |
| *** agarwalla has quit IRC | 22:12 | |
| vladimir3p | I saw some code entered back in March to cactus, but it is mostly related to integration with xs/etc layer | 22:13 |
| vishy | vladimir3p: proposal for multinic should be in soon | 22:14 |
| *** mattray has quit IRC | 22:15 | |
| vladimir3p | vishy: I guess the main question here - will it cover some case? seems like it addresses mostly redundant HW / multi networks ... | 22:15 |
| vladimir3p | vishy: I guess the main question here - will it cover this case? seems like it addresses mostly redundant HW / multi networks ... | 22:16 |
| vishy | i'm not sure it will cover all cases. | 22:17 |
| vishy | but we can add stuff too it if it doesn't | 22:17 |
| vladimir3p | ok ... just trying to see if we could create admin-vms cross projects/users ... | 22:18 |
| vladimir3p | with flat nw mode it works fine, but vlan is a bit tricky | 22:18 |
| vladimir3p | or probably we are missing something ... | 22:19 |
| vishy | vladimir3p: yes agreed. We may not be able to support admin vms in vlan mode until quantum | 22:19 |
| vishy | * at least ones that can support multiple projects at the same time | 22:19 |
| *** bcwaldon has quit IRC | 22:20 | |
| vladimir3p | vishy: thanks a lot ... as a side question - how popular vlan mode is? I thought it is a default/recommended one, correct? | 22:20 |
| *** foxtrotdelta has quit IRC | 22:22 | |
| vishy | vladimir3p: i think people are starting to use flatDHCP more | 22:23 |
| vladimir3p | ok, got it. at least it reduces one headache | 22:23 |
| *** heckj has joined #openstack-dev | 22:30 | |
| *** johnpur has quit IRC | 22:45 | |
| *** RobertLaptop has quit IRC | 22:46 | |
| *** jdurgin has quit IRC | 23:01 | |
| vladimir3p | vishy: Vish, sorry for too many questions ... would like to return to multi-nic discussion. | 23:02 |
| vladimir3p | there is a blueprint https://blueprints.launchpad.net/nova/+spec/multi-nic re accessing private and public networks | 23:02 |
| vladimir3p | it seems to be already implemented (based on code in xs portion only) | 23:03 |
| vladimir3p | tryng to understand we instance connected to particular vlan could also access public/general network ... for example for accessing LunR or another "shared" storage | 23:04 |
| vladimir3p | but we would like to have this access not through hv, but directly | 23:04 |
| vladimir3p | means ~ several networks ... (but not different projects) | 23:05 |
| vladimir3p | will it be enough to setup security group as in case of VPN instances? | 23:06 |
| mtaylor | annegentle: ping | 23:23 |
| vishy | vladimir3p: i think this should be creating different interfaces in the guest | 23:34 |
| vishy | which is what multinic does | 23:34 |
| *** vladimir3p has left #openstack-dev | 23:42 | |
| *** RobertLaptop has joined #openstack-dev | 23:48 | |
| *** adiantum has quit IRC | 23:54 | |
| *** jaypipes has quit IRC | 23:55 | |
| *** jaypipes has joined #openstack-dev | 23:56 | |
| *** zykes- has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!