| pvo | Lantizia: cloudstack and openstack are 2 different projects | 00:01 |
|---|---|---|
| Lantizia | both with a lot of glitzy words and little substance looking at the web sites | 00:01 |
| blamar | Lantizia: suggestions? | 00:02 |
| pvo | http://docs.openstack.org/ | 00:02 |
| Lantizia | that's what I'm reading | 00:02 |
| *** dprince has joined #openstack | 00:04 | |
| *** azneita has joined #openstack | 00:05 | |
| Lantizia | i hate the word cloud, it's thrown about promiscuously always with a different meaning | 00:05 |
| Lantizia | a page demonstrating useful scenarios and case studies without all the marketting bs would be handy | 00:05 |
| Lantizia | at the moment I've no idea why I'd want this over just libvirt | 00:05 |
| MotoMilind | Lantizia: Dashboard, Role based access control, potential to manage private, public and hybrid clouds, these may be some of the reasons to check out OpenStack | 00:07 |
| alekibango | Lantizia: if you have less than 5 servers, libvirt might just work | 00:08 |
| Lantizia | MotoMilind, so basically similar to Proxmox but not just KVM/OpenVZ | 00:08 |
| *** reldan has quit IRC | 00:08 | |
| Lantizia | and with management of actual customers not just the machines themselves | 00:08 |
| MotoMilind | Unfortunately, I haven't looked at Proxmax. Thanks for the tip, I will check them out. | 00:09 |
| Lantizia | np | 00:10 |
| MotoMilind | My driver was Python, based on Euca code, supported by Rackspace (and 50+ vendors) | 00:10 |
| Lantizia | it's Proxmox VE you want, not the Mail Gateway | 00:10 |
| MotoMilind | Aah, ok | 00:10 |
| Lantizia | pve.proxmox.com | 00:10 |
| MotoMilind | oh nice | 00:10 |
| *** joearnold has quit IRC | 00:11 | |
| MotoMilind | Time to read :) | 00:11 |
| RobertLaptop | Proxmox is very VSpear like | 00:11 |
| *** joearnold has joined #openstack | 00:11 | |
| alekibango | for me : 1) open (totally) 2) people (really smart) 3) companies (really great) 4) scaleable, adaptable | 00:11 |
| uvirtbot | New bug: #737240 in nova "sysadmin smoketests failing for ipv6" [Undecided,In progress] https://launchpad.net/bugs/737240 | 00:11 |
| *** allsystemsarego has quit IRC | 00:12 | |
| MotoMilind | Alright guys, see you tomorrow | 00:12 |
| *** reldan has joined #openstack | 00:12 | |
| alekibango | Proxmox makes people think that producing CO2 is bad... | 00:17 |
| *** jano has joined #openstack | 00:18 | |
| alekibango | the software looks easier to manage, but prolly not scalable as much as nova is | 00:18 |
| Lantizia | alekibango, not until v2.0 that is... see the roadmap | 00:19 |
| RobertLaptop | alekibango, Agreed on Proxmox | 00:20 |
| *** MotoMilind has quit IRC | 00:21 | |
| Lantizia | so are these CloudStack and OpenStack projects essentially trying to accomplish the same thing? | 00:22 |
| alekibango | ouch they have czech translation... i should work on translating nova lol | 00:22 |
| alekibango | Lantizia: no | 00:22 |
| Lantizia | and cloud.com are slowly giving up CloudStack for OpenStack? | 00:22 |
| Lantizia | ok so does CloudStack use OpenStack which uses libvirt... lol | 00:22 |
| *** jano has quit IRC | 00:22 | |
| Lantizia | it's just abstraction on top of abstraction on top of abstraction | 00:23 |
| *** sebastianstadil has quit IRC | 00:23 | |
| alekibango | Lantizia: openstack is not just product | 00:23 |
| alekibango | nto just way to manage servers | 00:23 |
| alekibango | it aims to be industry standard | 00:24 |
| Lantizia | in doing... _what_ ? | 00:24 |
| Lantizia | being an industry standard?! | 00:24 |
| alekibango | in doing clouds :) | 00:24 |
| Lantizia | oh ffs | 00:24 |
| alekibango | so people can migrate from one cloud to another | 00:24 |
| justinsb | Lantizia: Virtualization lets you run multiple virtual machines on one physical machine | 00:24 |
| alekibango | public/private clouds | 00:24 |
| Lantizia | this thing is either so amazingly good it's so hard to describe or it's just fluff | 00:24 |
| justinsb | Lantizia: Clouds are about running virtual machines on multiple physical machines | 00:25 |
| alekibango | Lantizia: hehe its hard... really and its changing so fast that if you describe it, it will not be true in a week :) | 00:25 |
| Lantizia | justinsb, that's one definition yes - and I know that's the one being used in this context | 00:25 |
| Lantizia | but is that to say a virtual machine runs in a cluster of hypervisors | 00:25 |
| *** sebastianstadil has joined #openstack | 00:26 | |
| Lantizia | or that a virtual machine is partly being ran on one physical machine whilst another part runs elsewhere? | 00:26 |
| justinsb | Lantizia: The former. | 00:26 |
| Lantizia | because that is true cloud in my mind | 00:26 |
| uvirtbot | New bug: #737249 in openstack-dashboard "image detail view incorrectly displays option to edit an image" [Critical,In progress] https://launchpad.net/bugs/737249 | 00:26 |
| Lantizia | right so it's not cloud then! | 00:26 |
| justinsb | Lantizia: Splitting one machine across multiple physical machines has its own problems | 00:26 |
| alekibango | Lantizia: you can run CPU and disk on different computers | 00:26 |
| Lantizia | it's just bloody managing clusters of hypervisors! | 00:26 |
| alekibango | or you might run them on one | 00:26 |
| justinsb | Lantizia: Which are probably very fundamental | 00:26 |
| alekibango | but its hard to separate ram and cpu :) | 00:26 |
| justinsb | Lantizia: _but_ the idea is that you can also rethink your unit of abstraction | 00:26 |
| justinsb | Lantizia: Away from a virtual machine | 00:27 |
| Lantizia | bah! it's just marketting nonsense | 00:27 |
| Lantizia | i hate the word cloud | 00:27 |
| justinsb | Lantizia: To thinking in terms of clusters of machines working together | 00:27 |
| alekibango | Lantizia: i do not like the word too.. still its better than intelectual property | 00:27 |
| alekibango | :) | 00:27 |
| alekibango | cloud is everything which is cloudy to your mind | 00:27 |
| alekibango | when you understand the problem, its crystal clear | 00:27 |
| *** maplebed has quit IRC | 00:28 | |
| Lantizia | because to get the _real_ definition of cloud (where you no VM runs on any 1 host as it runs on them all, like partition over a raided disk)... you'd have to have the hypervisor to support it | 00:28 |
| Lantizia | and to my knowledge only EC2 can do that? | 00:28 |
| bcwaldon | that's not necessarily the definition of cloud | 00:29 |
| alekibango | Lantizia: you can run disk from network (cloud/SAN) and you can continue teh virtual machine on another server immediatelly when you have problem | 00:29 |
| Lantizia | that's not a cloud! | 00:29 |
| justinsb | Lantizia: Are you trolling us? | 00:30 |
| bcwaldon | yeah | 00:30 |
| alekibango | Lantizia: btw swift is another type of cloud service - scalable object storage | 00:30 |
| Lantizia | justinsb, nah I'm not, just seriously annoyed at the use of the word cloud these days, means something different everytime I read it | 00:30 |
| justinsb | Lantizia: Cloud doesn't mean running one machine over multiple physical machines at the same time | 00:30 |
| alekibango | Lantizia: if you want solid, reliable cloud service, try http://www.supersimplestorageservice.com/ | 00:30 |
| justinsb | Lantizia: Nobody can do that | 00:30 |
| justinsb | Lantizia: What we can do is move one machine across a cluster of machines | 00:31 |
| justinsb | Lantizia: e.g. if one fails | 00:31 |
| Lantizia | essentially here the word cloud is being used to mean... virtualisation clusters that support many hypervisors... oh and storage can be elsewhere :P ... just like vSphere and Proxmox can already do | 00:31 |
| alekibango | Lantizia: forget cloud. openstack has object storage and compute part... | 00:31 |
| *** rnirmal has quit IRC | 00:32 | |
| alekibango | the word cloud essentially has no meaning | 00:32 |
| alekibango | its just like spam | 00:32 |
| *** sebastianstadil has quit IRC | 00:32 | |
| Lantizia | yup I agree | 00:32 |
| alekibango | its just something you do not understand | 00:32 |
| alekibango | (cloud) | 00:32 |
| alekibango | http://www.supersimplestorageservice.com/ is very cool | 00:32 |
| bcwaldon | here, I have the perfect link | 00:32 |
| bcwaldon | http://opensack.org/wiki/WhatIsCloud.html | 00:32 |
| Lantizia | bcwaldon, already read it - made me vomit | 00:33 |
| alekibango | opensack? | 00:33 |
| bcwaldon | Lantizia: you did not | 00:33 |
| Lantizia | lol | 00:33 |
| bcwaldon | lol | 00:33 |
| alekibango | there are no clouds on the page | 00:33 |
| Lantizia | sorry I read this one... http://docs.openstack.org/openstack-compute/admin/content/ch01s03.html | 00:33 |
| Lantizia | the first line made me vomit | 00:33 |
| alekibango | Lantizia: i can feel your pain | 00:34 |
| *** jano has joined #openstack | 00:34 | |
| *** dragondm has quit IRC | 00:34 | |
| justinsb | Lantizia: A lot of people are "cloud washing"... calling anything on the Internet "Cloud-X" | 00:34 |
| *** jano has quit IRC | 00:35 | |
| justinsb | Lantizia: because it attracts press | 00:35 |
| markwash | bcwaldon: your link is the best explanation I've seen so far | 00:35 |
| alekibango | reading nist makes me sometimes sick too | 00:35 |
| Lantizia | markwash, lol :P | 00:35 |
| justinsb | Lantizia: However, the original meaning of cloud is basically what OpenStack is building | 00:35 |
| Lantizia | which is basically just virtualisation clusters | 00:35 |
| Lantizia | which already had a perfectly good name | 00:35 |
| justinsb | Lantizia: I suggest you read "The Big Switch" if you want to understand the deeper implications | 00:35 |
| bcwaldon | get incepted | 00:35 |
| alekibango | the pdf document linked on the page you gave Lantizia is not that bad | 00:36 |
| alekibango | http://csrc.nist.gov/groups/SNS/cloud-computing/cloud-def-v15.doc | 00:36 |
| justinsb | Lantizia: True, it's not a million miles away from virtualization clusters | 00:36 |
| alekibango | Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. | 00:36 |
| alekibango | funny def, but not that evil | 00:36 |
| Lantizia | "Note 1: Cloud computing is still an evolving paradigm. " | 00:37 |
| Lantizia | no shit lol | 00:37 |
| alekibango | lol | 00:37 |
| bcwaldon | lol | 00:37 |
| alekibango | Lantizia: so just read the code... or wiki.openstack.org :) | 00:38 |
| Lantizia | were launching this line bonding service soon where you can use any ISP... and bond them over this vpn-like technology... but it's proper bonding... gonna stick the customers firewall/router as a VM on the other side of the connection at the datacentre plus all their desktops/servers etc... a neat idea | 00:39 |
| Lantizia | until i did something stupid in trying to explain the concept to my boss | 00:39 |
| Lantizia | i said... basically kinda like... cloud broadband | 00:39 |
| alekibango | lmao | 00:39 |
| Lantizia | oh ffs, what did he do? register the domain before anyone else | 00:39 |
| alekibango | i today heard 'cloud banking' | 00:39 |
| alekibango | :) | 00:39 |
| Lantizia | even I have contributed to the mess | 00:39 |
| alekibango | now i imagine river banks full of clouds | 00:40 |
| bcwaldon | clouds everywhere | 00:40 |
| justinsb | Lantizia: I think that's why we're called "OpenStack" and not "OpenCloud" | 00:40 |
| Lantizia | lol | 00:40 |
| bcwaldon | lol | 00:40 |
| alekibango | +1 | 00:40 |
| Lantizia | someone will make SkyNet next to manage the clouds :P | 00:40 |
| Lantizia | since clouds are in the sky :P | 00:40 |
| alekibango | skynet should be rady soon | 00:40 |
| alekibango | should be connected to internet soon | 00:41 |
| *** joearnold has quit IRC | 00:41 | |
| alekibango | singularity is near! | 00:41 |
| alekibango | :) | 00:41 |
| Lantizia | judgement day :) | 00:41 |
| alekibango | its closer and closer, every minute its 60 seconds closer! | 00:42 |
| *** adjohn has joined #openstack | 00:42 | |
| Lantizia | ok here is what my boss really wants to know... can we/is it possible to manage multiple ESXi boxes in a cluster without using stupid fucking vSphere Centre which needs windows | 00:43 |
| Lantizia | he was reading the register and noticed this CloudStack has an agent that goes on the hypervisor | 00:43 |
| Lantizia | which then lead me to you thinking "damn this sounds just like libvirt lol" | 00:43 |
| alekibango | ESXi? imho not yet | 00:44 |
| alekibango | in openstack | 00:44 |
| Lantizia | i'm still not entirely clear on what the differences of OpenStack vs CloudStack is | 00:44 |
| Lantizia | alekibango, oh I've seen - not until possibly the next version... but CloudStack (however that differs) claims it can now | 00:44 |
| alekibango | thats cloud.com | 00:45 |
| alekibango | right? | 00:45 |
| Lantizia | yup which makes reference to openstack | 00:45 |
| Lantizia | http://cloud.com/products/cloud-computing-software | 00:45 |
| justinsb | Lantizia: Sounds like they're "OpenStack" washing | 00:45 |
| Lantizia | http://cloud.com/products/driving-open-source-cloud-computing-open-stack | 00:46 |
| RobertLaptop | ESXi control without vSphere is really hard to do and almost impossible to get "right" | 00:46 |
| justinsb | Lantizia: Just like "Cloud" washing | 00:46 |
| justinsb | Lantizia: When something is good, everyone tries to associate themselves with it | 00:46 |
| Lantizia | justinsb, yeah - that gets damned annoying | 00:46 |
| alekibango | i love tomatoes! | 00:46 |
| alekibango | :) | 00:46 |
| *** vernhart has quit IRC | 00:46 | |
| Lantizia | TOMATOE WASHER | 00:46 |
| Lantizia | -e :S | 00:47 |
| alekibango | dont take it all so seriously. almost everythink you know is | 00:47 |
| justinsb | Lantizia: I think that cloud.com have announced that they intend to work with Openstack in the future | 00:47 |
| alekibango | emh ... wrong? | 00:47 |
| Lantizia | so they have like totally nothing to do with each other? | 00:47 |
| *** sebastianstadil has joined #openstack | 00:47 | |
| alekibango | should say 'prepared for your mind by some industry' | 00:47 |
| winston-d | hi, guys. Is there any CloudFiles PHP bindings example available ? | 00:47 |
| justinsb | Lantizia: I don't know how they intend to do that. You could ask them! | 00:47 |
| alekibango | Lantizia: cloud.com has its own solutions... they did have something even before openstack | 00:48 |
| Lantizia | justinsb, well their cloudstack is a free download for ubuntu too.. but from what I can see (tell me if you think differently) they do the same thing as openstack :S | 00:48 |
| justinsb | Lantizia: Thanks, I'll stick to OpenStack | 00:48 |
| Lantizia | oh I haven't tried either lol - i'm not taking sides | 00:49 |
| alekibango | justinsb: let him check it all out before he will understand waht N.O.V.A. really means | 00:49 |
| Lantizia | just trying to understand the relationship (if any) or if they serve the same purpose | 00:49 |
| justinsb | Lantizia: OpenStack is open. I think that's the big differentiator. | 00:49 |
| alekibango | Lantizia: they prolly want to be compatible with openstack in longer run somehow | 00:50 |
| alekibango | somehow | 00:50 |
| justinsb | Lantizia: Open source, open development model, open everything | 00:50 |
| justinsb | Lantizia: Not "open core" or another commercial model | 00:50 |
| Lantizia | haven't seen any tale tell signs of opencore | 00:51 |
| Lantizia | a shitty tactic i agree if it is | 00:51 |
| RobertLaptop | Getting out from under ESXi is really hard. | 00:51 |
| Lantizia | were not really under it yet... it's not like we have to use it | 00:52 |
| justinsb | Lantizia: I think that OpenStack and CloudStack both serve the same purpose | 00:52 |
| Lantizia | I personally use KVM under proxmox | 00:52 |
| justinsb | Lantizia: In the same sense that Linux and Windows do | 00:52 |
| RobertLaptop | You either have to pay lots of $ or you have to use tricks to work around the limits VMware has put in to try and force people to pay the $ | 00:52 |
| justinsb | Lantizia: To be fair to CloudStack, their product is a lot prettier at the moment - a bit like Windows | 00:53 |
| *** clauden has quit IRC | 00:53 | |
| Lantizia | justinsb, does openstack offer commercial support? | 00:53 |
| alekibango | still, openstack is one of largests and most active open projects | 00:54 |
| justinsb | Lantizia: Well, OpenStack itself won't. But companies can | 00:54 |
| alekibango | Lantizia: now yes, rackspace does | 00:54 |
| justinsb | Lantizia: OpenStack = the Linux project | 00:54 |
| alekibango | i do too | 00:54 |
| Lantizia | ok | 00:54 |
| alekibango | Linux of the cloud | 00:54 |
| justinsb | Lantizia: Other companies can be the "RedHat" of Openstack | 00:54 |
| alekibango | or apache of the cloud | 00:54 |
| Lantizia | redhat care about openstack ? thought they had their own thing based on libvirt? | 00:54 |
| RobertLaptop | ovirt is dead. | 00:55 |
| Lantizia | alekibango, stop washing :P | 00:55 |
| justinsb | Lantizia: You'll have to ask them | 00:55 |
| alekibango | lol | 00:55 |
| alekibango | Lantizia: they do have theirs own solution | 00:55 |
| Lantizia | cool ok | 00:55 |
| alekibango | they have even nice spice protocol | 00:55 |
| alekibango | which is open now | 00:55 |
| justinsb | alekibango: What is redhat's cloud product? | 00:56 |
| Lantizia | oVirt I think | 00:56 |
| *** adjohn_ has joined #openstack | 00:57 | |
| alekibango | not 100% sure | 00:57 |
| alekibango | www.redhat.com/solutions/cloud/products/ | 00:57 |
| dprince | justinsb: Thanks for knocking out the metadata delete ticket. | 00:58 |
| alekibango | they have also some included provisioning systems | 00:58 |
| RobertLaptop | Lantizia, oVirt is pretty much dead. http://www.redhat.com/virtualization/rhev/server/ | 00:58 |
| RobertLaptop | Redhat bought some tech and they are calling it rhev | 00:59 |
| justinsb | dprince: No worries. I'm a little concerned that I wasn't able to write a unit test for it | 00:59 |
| justinsb | dprince: Did it "work for you"? | 00:59 |
| *** adjohn has quit IRC | 00:59 | |
| dprince | justinsb: just saw your fix. Will try it out in a minute. | 00:59 |
| alekibango | btw really nice example of cloud computing is sheepdog screensaver... and cool one... | 00:59 |
| justinsb | dprince: Thanks! How did you find the bug, by the way | 00:59 |
| dprince | justinsb: Unit tests through the whole stack (accross a cast) are hard to do. | 00:59 |
| justinsb | dprince: I was wondering if I could try to test it however you found it | 01:00 |
| justinsb | dprince: Or were you poking around in the DB ? | 01:00 |
| dprince | justinsb: send me your ssh key? I'm using Cloud Servers to test things functionally. | 01:02 |
| dprince | justinsb: dan.prince@rackspace.com | 01:02 |
| alekibango | dprince: you are doing fuctional tests of nova? | 01:03 |
| markwash | justinsb: don't fall for dan's 'send me your ssh private key' tricks | 01:03 |
| *** littleidea has quit IRC | 01:04 | |
| justinsb | markwash: I'll watch out for his tricks :-) | 01:04 |
| justinsb | dprince: Key on its way, thanks! | 01:04 |
| dprince | alekibango: I'm mostly working on API feature at the moment. I do functionally test things using this before I merge prop however: | 01:04 |
| dprince | https://github.com/dprince/openstack_vpc | 01:04 |
| dprince | Very similar setup to Vish's Vagrant setups. Just uses cloud servers instead. | 01:05 |
| *** Oneiropolo has quit IRC | 01:06 | |
| alekibango | dprince: i am using fai-project to install openstack on real hardware... which might be interesting for you | 01:06 |
| *** adjohn has joined #openstack | 01:06 | |
| dprince | alekibango: Sure. I'll check it out. | 01:06 |
| alekibango | it is possible to install nova on debian stable/sid/testing and few ubuntu distros too -- and run tests | 01:07 |
| alekibango | if you will have some real servers, i can help you to set it up | 01:07 |
| alekibango | i will release my own config when i will be ok with it | 01:07 |
| alekibango | still not good enough | 01:07 |
| alekibango | last night i tried i have seen such weird errors | 01:07 |
| alekibango | mysql complaining about key lenght or something like that | 01:08 |
| * alekibango is going to check todays version | 01:08 | |
| alekibango | dprince: installing nova is pretty complex task | 01:08 |
| dprince | alekibango: I use Chef. ;) | 01:09 |
| alekibango | having fai setup it and doing tests would give: | 01:09 |
| *** adjohn_ has quit IRC | 01:09 | |
| alekibango | possible automated tests on different oses on real hw | 01:09 |
| alekibango | public configuration sets -- as examples for nova deployment | 01:09 |
| *** jano has joined #openstack | 01:10 | |
| alekibango | when all runs well, its really fast, 4-10 minutes for installing all cloud | 01:10 |
| *** sebastianstadil has quit IRC | 01:10 | |
| alekibango | we really need good functional tests of nova to make sure it is stable | 01:13 |
| alekibango | unit testing is not enough | 01:13 |
| justinsb | alekibango: +1 | 01:16 |
| justinsb | alekibango: I think there are a number of people working on it | 01:17 |
| justinsb | alekibango: Like dprince, it seems! | 01:17 |
| alekibango | :) | 01:17 |
| Lantizia | Since you've all been a lovely lot - what are peoples thoughts on AoE... dead duck or still worth exploring? | 01:18 |
| dprince | justinsb: see my email | 01:19 |
| *** mahadev has quit IRC | 01:20 | |
| *** adjohn has quit IRC | 01:22 | |
| *** littleidea has joined #openstack | 01:22 | |
| *** nerens_ has joined #openstack | 01:24 | |
| *** nerens has quit IRC | 01:26 | |
| *** nerens_ is now known as nerens | 01:26 | |
| *** littleidea has quit IRC | 01:29 | |
| *** jbryce_ has joined #openstack | 01:32 | |
| *** jbryce has quit IRC | 01:34 | |
| *** jbryce_ is now known as jbryce | 01:34 | |
| *** mahadev has joined #openstack | 01:44 | |
| *** mahadev has quit IRC | 01:48 | |
| *** nelson has quit IRC | 01:52 | |
| *** jwilcox has joined #openstack | 01:54 | |
| *** jwilcox is now known as jeffjapan | 01:55 | |
| *** mahadev has joined #openstack | 01:55 | |
| alekibango | Lantizia: i would rather use sheepdog | 01:56 |
| alekibango | but no storage solution is 100%... still needs to improve | 01:56 |
| Lantizia | alekibango, can sheepdog be uses for storage area networks as well as rounding up sheep? | 01:57 |
| *** pvo has quit IRC | 01:57 | |
| alekibango | kind of. see http://www.osrg.net/sheepdog/ | 01:59 |
| alekibango | luckily they are again online | 01:59 |
| alekibango | (japan) | 01:59 |
| alekibango | its block level storage abstraction | 02:00 |
| *** reldan has quit IRC | 02:14 | |
| *** mahadev has quit IRC | 02:17 | |
| *** mahadev has joined #openstack | 02:38 | |
| *** mahadev has quit IRC | 02:42 | |
| *** jeffjapan has quit IRC | 02:42 | |
| *** mahadev has joined #openstack | 02:49 | |
| *** j05h has quit IRC | 03:06 | |
| *** jwilcox has joined #openstack | 03:09 | |
| *** rpedde has joined #openstack | 03:09 | |
| *** jwilcox is now known as jeffjapan | 03:10 | |
| *** mahadev has quit IRC | 03:20 | |
| *** adjohn has joined #openstack | 03:21 | |
| *** ryker has quit IRC | 03:25 | |
| HugoKuo_ | I try to deploy nova in a single machine | 03:32 |
| HugoKuo_ | two question.... | 03:32 |
| HugoKuo_ | what if I want my instance's network same with the host , can I ony use one NIC ? | 03:33 |
| *** mahadev has joined #openstack | 03:34 | |
| rpedde | It's easiest probably to create a bridge with your eth0, assign your host ip to the bridge, and specify that bridge as the bridge to add your veth devices to | 03:36 |
| rpedde | http://wiki.debian.org/BridgeNetworkConnections | 03:36 |
| rpedde | But you could as easily use a second ethernet card, make a bridge with that ethernet device in it, and use *that* bridge for your guest machines | 03:38 |
| *** mahadev has quit IRC | 03:42 | |
| *** kashyap has joined #openstack | 03:44 | |
| HugoKuo_ | I success to use second ethernet device accomplish my goal... | 03:48 |
| rpedde | good deal | 03:48 |
| *** kashyap has quit IRC | 03:53 | |
| HugoKuo_ | thanks rpedde , I'll try first way to use one NIC | 03:53 |
| rpedde | Just double-check your configs before you restart networking if you don't have physical access to the machine. I've messed up configs turning an eth into a bridge and lost connectivity with the machine. Not bad if you are right next to the machine, but really painful when you are running remote over ssh. :) | 03:55 |
| HugoKuo_ | haha , I know :> | 04:01 |
| HugoKuo_ | well , just finish the testing | 04:07 |
| HugoKuo_ | but there's some issue | 04:07 |
| rpedde | what's the issue? | 04:08 |
| HugoKuo_ | instance can not run up , due to metadata get fail | 04:08 |
| HugoKuo_ | thinking | 04:09 |
| HugoKuo_ | let me draw a diagram first~ | 04:09 |
| *** pvo has joined #openstack | 04:12 | |
| *** mahadev has joined #openstack | 04:17 | |
| *** mahadev has quit IRC | 04:18 | |
| *** adjohn has quit IRC | 04:19 | |
| *** adjohn has joined #openstack | 04:22 | |
| *** mahadev has joined #openstack | 04:27 | |
| HugoKuo_ | http://i217.photobucket.com/albums/cc280/tonytkdk/succe.jpg | 04:27 |
| HugoKuo_ | I don't know why that my instance can not get matadata from api server | 04:28 |
| rpedde | what image are you using? | 04:28 |
| *** dprince has quit IRC | 04:28 | |
| HugoKuo_ | oh sorry , It's not network unreachable , the response is url error | 04:29 |
| HugoKuo_ | I try two images | 04:29 |
| HugoKuo_ | UEC lucid 32 | 04:30 |
| HugoKuo_ | both tty image | 04:30 |
| HugoKuo_ | flatDHCP will start bridge br100 with eth0 , and assign the IP address which is writen in mysql database . | 04:32 |
| HugoKuo_ | automatically ~! | 04:32 |
| HugoKuo_ | so confusing .... | 04:33 |
| rpedde | Yes... okay | 04:33 |
| HugoKuo_ | waiting for metadata service at http://169.254.169.254/2009-04-04/meta-data/instance-id | 04:33 |
| HugoKuo_ | 02:49:17 [ 1/100]: url error [timed out] | 04:33 |
| HugoKuo_ | how do you think about this error , cause by nova-network or nova-api ?? | 04:34 |
| rpedde | so I'm not a super nova brainiac, but this is how I understand it. UEC images want to get provisioning metadata from the UEC metadata server | 04:34 |
| rpedde | nova api will emulate that, but you have to make traffic bound to 169.254.169.254 go to the nova api host | 04:35 |
| rpedde | When using flat/dhcp, that's easy... you are dnsmasqing from the host and routing through it, so you can use iptables to redirect traffic from the guest to the nova api | 04:36 |
| rpedde | like so: iptables -t nat -A PREROUTING -d 169.254.169.254/32 -p tcp -m tcp --dport 80 -j DNAT --to-destination $NOVA_API_IP:8773 | 04:36 |
| rpedde | But with regular flat, you aren't forcing that traffic through the compute node, so it doesn't get a chance to redirect that traffic to the metadata server | 04:37 |
| HugoKuo_ | this rule is already exist , that's writen by linux_net.py | 04:37 |
| HugoKuo_ | btw , I use FlatDHCP mode | 04:37 |
| *** nerens has quit IRC | 04:38 | |
| rpedde | Right, but if you are just flat with no dhcp, then you aren't routing through the compute node, you are bridging | 04:38 |
| rpedde | oh. meh. | 04:38 |
| rpedde | should go then, afaik. :/ | 04:38 |
| HugoKuo_ | see ya | 04:38 |
| HugoKuo_ | thanks | 04:38 |
| rpedde | sorry. during the day here GMT -6 there are way smarter people here than me | 04:39 |
| rpedde | Might have more luck dropping by then | 04:39 |
| HugoKuo_ | :> | 04:39 |
| HugoKuo_ | where r u ? | 04:39 |
| rpedde | Texas! | 04:39 |
| HugoKuo_ | I'm in TW.... nice to meet you! | 04:39 |
| HugoKuo_ | oh TW= Taiwan :> | 04:40 |
| rpedde | Good to meet you! Do try later though, I know the folks here can help | 04:40 |
| HugoKuo_ | :-) | 04:40 |
| rpedde | Sorry so useless. heh. | 04:40 |
| HugoKuo_ | dude, don't say so ..... | 04:41 |
| HugoKuo_ | I'm sorry about my poor english XD | 04:41 |
| rpedde | HA! It's better than a lot of the texans I work with! | 04:42 |
| *** pvo has quit IRC | 04:47 | |
| *** omidhdl has joined #openstack | 04:49 | |
| openstackjenkins | Project nova build #662: SUCCESS in 2 min 3 sec: http://hudson.openstack.org/job/nova/662/ | 04:53 |
| openstackjenkins | Tarmac: Make the rpc cast/call debug calls show what topic they are sending to. This aides in debuugging. | 04:53 |
| *** rpedde has quit IRC | 04:53 | |
| *** rpedde has joined #openstack | 04:55 | |
| rpedde | Hrm. I don't really have an answer. The obvious things to check would be to check brctl br100 and make sure the vm interface joined the bridge and that eth0 is a part of it. And check to make sure that dnsmasq is running. It should show the config file in the command line... you can check that and make sure it's handing out a correct ip address via dhcp. Beyond that, I don't have any good answers. | 04:57 |
| *** mahadev has quit IRC | 04:58 | |
| *** pvo has joined #openstack | 05:05 | |
| openstackjenkins | Project nova build #663: SUCCESS in 2 min 6 sec: http://hudson.openstack.org/job/nova/663/ | 05:08 |
| openstackjenkins | Tarmac: When changing the project manager, if the new manager is not yet a project member, be sure to make them be a project member. | 05:08 |
| *** kashyap has joined #openstack | 05:09 | |
| *** bcwaldon has quit IRC | 05:10 | |
| *** hadrian has quit IRC | 05:10 | |
| *** adjohn has quit IRC | 05:15 | |
| *** adjohn_ has joined #openstack | 05:15 | |
| openstackjenkins | Project nova build #664: SUCCESS in 2 min 8 sec: http://hudson.openstack.org/job/nova/664/ | 05:15 |
| openstackjenkins | Tarmac: Make "ApiError" the default error code for ApiError instances, rather than "Unknown." | 05:15 |
| openstackjenkins | Note Dashoard's error handling code: | 05:15 |
| openstackjenkins | except boto.exception.BotoServerError, e: | 05:15 |
| openstackjenkins | if e.status == 400 and e.error_code == 'ApiError': | 05:15 |
| openstackjenkins | raise NovaApiError(e) | 05:15 |
| openstackjenkins | elif e.status == 401: | 05:15 |
| openstackjenkins | raise NovaUnauthorizedError() | 05:15 |
| openstackjenkins | elif e.status == 503: | 05:15 |
| openstackjenkins | raise NovaUnavailableError(e) | 05:15 |
| openstackjenkins | raise NovaServerError(e) | 05:15 |
| openstackjenkins | So an error_code of 'Unknown' will raise an ugly exception. | 05:15 |
| openstackjenkins | This change also makes the euca errors a bit prettier: | 05:15 |
| openstackjenkins | > euca-attach-volume -i i-1 vol-2 -d /dev/vdb | 05:15 |
| openstackjenkins | > ApiError: ApiError: Volume status must be available | 05:15 |
| openstackjenkins | Rather than: | 05:15 |
| openstackjenkins | > euca-attach-volume -i i-1 vol-2 -d /dev/vdb | 05:15 |
| openstackjenkins | > Unknown: Unknown: Volume status must be available | 05:15 |
| openstackjenkins | Though the reiteration of the error code is still a bit ugly. | 05:15 |
| *** koolhead17|afk is now known as koolhead17 | 05:16 | |
| koolhead17 | hi all | 05:17 |
| *** pvo has quit IRC | 05:17 | |
| *** localhost has joined #openstack | 05:17 | |
| *** sebastianstadil has joined #openstack | 05:22 | |
| *** rpedde has quit IRC | 05:27 | |
| HugoKuo_ | fine , | 05:27 |
| HugoKuo_ | success to use one NIC | 05:27 |
| HugoKuo_ | but there's something wired @@ | 05:28 |
| HugoKuo_ | whatever , instnace is run well | 05:28 |
| *** adjohn_ has quit IRC | 05:28 | |
| HugoKuo_ | I test with all kinds of image | 05:30 |
| HugoKuo_ | include UEC and tty | 05:30 |
| HugoKuo_ | eventhough that tty can not get metadata , but it still working fine..... | 05:31 |
| HugoKuo_ | funny XD | 05:31 |
| *** littleidea has joined #openstack | 05:37 | |
| *** mahadev has joined #openstack | 05:40 | |
| *** f4m8_ is now known as f4m8 | 05:40 | |
| *** mahadev has quit IRC | 05:41 | |
| uvirtbot | New bug: #737345 in nova "API returns UnknownError when executing euca-add-keypair in redhat6" [Undecided,New] https://launchpad.net/bugs/737345 | 05:41 |
| uvirtbot | New bug: #737346 in nova "RunInstance failed due to network error in redhat6" [Undecided,New] https://launchpad.net/bugs/737346 | 05:52 |
| *** zenmatt_ has quit IRC | 06:07 | |
| *** zenmatt has joined #openstack | 06:10 | |
| *** mahadev has joined #openstack | 06:10 | |
| *** romain_lenglet has joined #openstack | 06:20 | |
| *** nerens has joined #openstack | 06:22 | |
| *** miclorb_ has quit IRC | 06:25 | |
| *** mahadev has quit IRC | 06:30 | |
| *** mahadev has joined #openstack | 06:33 | |
| *** sebastianstadil has quit IRC | 06:34 | |
| *** littleidea has quit IRC | 06:35 | |
| *** romain_lenglet has left #openstack | 06:49 | |
| *** ggy has joined #openstack | 06:51 | |
| *** sebastianstadil has joined #openstack | 06:54 | |
| *** guigui has joined #openstack | 06:56 | |
| *** ggy has quit IRC | 06:58 | |
| uvirtbot | New bug: #737360 in openstack-dashboard "project managers should be able to modify image attributes of an image belonging to the project" [High,Confirmed] https://launchpad.net/bugs/737360 | 07:06 |
| *** shrewduh has joined #openstack | 07:06 | |
| *** mahadev has quit IRC | 07:08 | |
| *** vernhart has joined #openstack | 07:14 | |
| *** sebastianstadil has quit IRC | 07:14 | |
| *** shrewduh_ has joined #openstack | 07:19 | |
| *** shrewduh has quit IRC | 07:20 | |
| *** shrewduh has joined #openstack | 07:21 | |
| *** shrewduh_ has quit IRC | 07:23 | |
| *** mgoldmann has joined #openstack | 07:24 | |
| *** Nacx has joined #openstack | 07:24 | |
| *** lionel_ has quit IRC | 07:27 | |
| *** sebastianstadil has joined #openstack | 07:28 | |
| *** lionel_ has joined #openstack | 07:31 | |
| *** jano_ has joined #openstack | 07:38 | |
| *** jano has quit IRC | 07:38 | |
| *** lionel_ has quit IRC | 07:39 | |
| *** CloudChris has joined #openstack | 07:39 | |
| *** lionel_ has joined #openstack | 07:40 | |
| uvirtbot | New bug: #737368 in nova "s3 object store not using proper names for image attributes" [Medium,In progress] https://launchpad.net/bugs/737368 | 07:42 |
| *** toluene has joined #openstack | 07:51 | |
| *** jano_ has quit IRC | 08:00 | |
| *** jano has joined #openstack | 08:00 | |
| *** kalonike has joined #openstack | 08:02 | |
| *** jeffjapan has quit IRC | 08:06 | |
| *** yamahata_home has joined #openstack | 08:06 | |
| *** yamahata_home is now known as yamahata_laptop | 08:07 | |
| *** yamahata_laptop is now known as yamahata_i | 08:08 | |
| *** MarcMorata has joined #openstack | 08:15 | |
| *** skiold has joined #openstack | 08:19 | |
| *** azneita has quit IRC | 08:20 | |
| *** zaitcev has quit IRC | 08:23 | |
| *** uksysadmin has joined #openstack | 08:27 | |
| HugoKuo_ | bonjour | 08:38 |
| nerens | howdie | 08:46 |
| nerens | i'm trying to start new instances and it hangs at scheduling, any ideas? is this a network issue? | 08:48 |
| nerens | Mar 18 10:44:01 hera dhclient: No DHCPOFFERS received. | 08:48 |
| nerens | Mar 18 10:44:01 hera dhclient: No working leases in persistent database - sleeping. | 08:48 |
| soren | If it's stuck in scheduling, I doubt it's a network thing. | 08:50 |
| soren | Try looking at your nova-scheduler.log | 08:50 |
| *** daveiw has joined #openstack | 08:50 | |
| nerens | ah.. you're right: ERROR nova.rpc [-] Failed to fetch message from queue | 08:50 |
| *** olafont has quit IRC | 08:54 | |
| *** allsystemsarego has joined #openstack | 08:59 | |
| *** allsystemsarego has joined #openstack | 08:59 | |
| soren | Never seen that before | 09:07 |
| nerens | I'm seeing similar errors as this post: https://bugs.launchpad.net/nova/+bug/718869 | 09:11 |
| uvirtbot | Launchpad bug 718869 in nova "nova-compute: can't spawn new instances after restarting RabbitMQ" [High,Confirmed] | 09:11 |
| nerens | was also seeing the error : CONNECTION_FORCED - broker forced connection closure with reason 'shutdown' | 09:13 |
| nerens | so does that mean I shouldn't restart rabbitmq? Or if I do do I need to restart everything else? | 09:13 |
| *** omidhdl has quit IRC | 09:15 | |
| nerens | does rabbit have a config file? /etc/rabbitmq is empty | 09:16 |
| *** drico has joined #openstack | 09:19 | |
| *** ggy has joined #openstack | 09:22 | |
| perra_ | I must admit I haven't got much experience working with ebtables. Can I find somewhere through documentation what rules are configured (their consequence) in the nat table (ebtables) of the host for each instance? | 09:23 |
| *** ramkrsna has joined #openstack | 09:25 | |
| *** ramkrsna has quit IRC | 09:30 | |
| ggy | Hi I was able to get an instance running but I am not able to ssh or ping on to it. The output of euca-get-console output is at http://paste.openstack.org/show/910/ | 09:36 |
| alekibango | nerens: with empty rabbitmq, there is still default config... having guest user... and nova is by default configured to use that. | 09:38 |
| alekibango | ggy: did you use euca-authorize ? | 09:39 |
| alekibango | google openstack ssh euca-authorize if you didnt | 09:40 |
| ggy | alekibango: yeah! had used the multinode install script. So it was done automatically. When I tried to do it again manually, I got the message that the rule already exists | 09:41 |
| *** ramkrsna has joined #openstack | 09:44 | |
| ggy | I have been stuck with this problem for almost 3-4 days now! Unable to ping or ssh onto an instance | 09:50 |
| *** olafont has joined #openstack | 10:01 | |
| soren | What's the image that you're using? | 10:01 |
| nerens | alekibango: thx | 10:03 |
| soren | ggy: What | 10:03 |
| soren | whoops | 10:03 |
| soren | ggy: What's the image you're trying to run? | 10:03 |
| ggy | +soren: I was runnign the image from http://c0179148.cdn1.cloudfiles.rackspacecloud.com/ubuntu1010-UEC-localuser-image.tar.gz . I had also tried ttylinux-uec-amd64-12.1_2.6.35-22_1.tar.gz | 10:04 |
| soren | ggy: What does euca-describe-instances say about it? | 10:05 |
| ggy | it shows the status as running | 10:05 |
| soren | Ok. | 10:05 |
| soren | I would check nova-compute.log to see if it says anything interesting. | 10:06 |
| ggy | +soren: there were no errors | 10:07 |
| ggy | +soren: I have pasted the content from nova-compute.log http://paste.openstack.org/show/924/ | 10:09 |
| *** yamahata_i has quit IRC | 10:13 | |
| *** ggy has quit IRC | 10:14 | |
| *** miclorb has joined #openstack | 10:14 | |
| *** ggy has joined #openstack | 10:28 | |
| soren | ggy: That looks fine. Hmm. | 10:29 |
| soren | ggy: Oh, you're using FlatManager. | 10:29 |
| soren | ggy: That doesn | 10:29 |
| soren | t work with ttylinux. | 10:29 |
| *** ved_ has joined #openstack | 10:29 | |
| ggy | +soren: but it should work with the maveric release right? | 10:30 |
| soren | I think. | 10:30 |
| ved_ | Hi, is there a way I can map my NFS volume (on NetApp storage for eg) for object storage ? | 10:30 |
| soren | I don't really use the FlatManager. | 10:30 |
| ggy | Actually I used the install script and so the default was FlatManager. Which one are you using?. Actually I had seen a couple of other guys with the same problem too | 10:32 |
| soren | ggy: I use VlanManager. | 10:34 |
| soren | ggy: It shouldn't be breaking this this badly for you, though. | 10:34 |
| soren | ggy: How are you uploading your images to Nova? | 10:34 |
| soren | ttx: netadmin_smoketests and sysadmin_smoketests now work beautifully. | 10:35 |
| ggy | +soren: I used uec-publish-image | 10:35 |
| soren | Ok, that's what I do, too. | 10:35 |
| *** shrewduh has quit IRC | 10:43 | |
| *** fayce has joined #openstack | 10:47 | |
| HugoKuo_ | Just finish 1 front end + 1 compute node XD | 10:47 |
| HugoKuo_ | looking good | 10:47 |
| HugoKuo_ | but I got a question .........I install nova-compute in front too | 10:48 |
| HugoKuo_ | and my front end has only 2 core + 2G ram and compute node has 4cores with 4GB ram | 10:48 |
| HugoKuo_ | how come that I can run up 6 instance on frontend , but only 4 instance on compute node!! | 10:49 |
| HugoKuo_ | it's really wired ... | 10:49 |
| soren | The current schedulers are really primitive. | 10:50 |
| perra_ | where should I look if I want to do some alteration how the dnsmasq and nova-dhcp are handing out IP-addresses. What component of nova is responsible of keeping the database consistent reflecting what addresses has been handed out within a project ? | 10:50 |
| HugoKuo_ | /var/lib/nova/networks/nova-br100.conf | 10:51 |
| HugoKuo_ | would you want this ? | 10:51 |
| HugoKuo_ | perra_ : | 10:51 |
| soren | perra_: What are you wanting to do? | 10:52 |
| perra_ | that only provides information about what has been handed out by the dhcp so far, right? Editing that conf-file won't result in any changes, or am I wrong? | 10:52 |
| soren | Any changes will likely be overwritten by nova quite soon. What are you trying to do? | 10:53 |
| perra_ | soren: I want to bypass the euca2ool, letting a user choose whatever IP-address available in the project address space. | 10:54 |
| soren | Oh, that again. | 10:54 |
| perra_ | haha :) | 10:54 |
| perra_ | more people been asking? | 10:54 |
| soren | Why do you want to do this? | 10:54 |
| HugoKuo_ | sound interesting | 10:55 |
| soren | It sounds horribly broken t ome. | 10:55 |
| soren | But go on. Convince me otherwise :) | 10:55 |
| HugoKuo_ | :\ | 10:55 |
| soren | perra_: What is the use case? | 10:58 |
| perra_ | I'm on a project which would benefit multiple networks within a cloud and also connecting them together using a VM as a router. Basically building networks within the cloud. Are multiple networks among the objectives of OpenStack in the nearby future? There are some old server implementations which requires specific IP-addresses which we would like to place in the cloud aswell. Therefore we would like to be able to set IP addresses, specificly. | 10:58 |
| *** z0 has joined #openstack | 10:59 | |
| perra_ | use case: short term -> networks within a cloud environment, where users can manipulate networks settings as they like within a project. long term -> distributed clouds | 11:00 |
| *** ved_ has quit IRC | 11:00 | |
| soren | Multiple networks are on the roadmap. | 11:01 |
| soren | Not for the release in April, but very, very likely later this year. | 11:01 |
| perra_ | soren: sound great! | 11:01 |
| soren | At that point, you can probably do whatever you want with IP-addresses. | 11:02 |
| HugoKuo_ | current max instances number only 10 ? | 11:02 |
| perra_ | soren: but you still don't like my horrible idea? :) | 11:02 |
| soren | HugoKuo_: I forget what the defaults limits are, but you can change them if you want. | 11:02 |
| soren | perra_: Not one bit, no. | 11:03 |
| perra_ | fair enough | 11:03 |
| soren | perra_: If you need to assign specific ip addresses to something, you're doing it wrong. | 11:03 |
| soren | perra_: Cloud != virtualisation. | 11:03 |
| HugoKuo_ | oh thanks | 11:04 |
| perra_ | or, they did something wrong in the past, which I have to live with :/ | 11:04 |
| soren | Well, it's fine as such. It's just not something that belong in the cloud, then. | 11:04 |
| HugoKuo_ | any flag to define the maxmum instances? | 11:04 |
| soren | There are plenty of use cases that simply don't belong in the cloud. That's fine. | 11:05 |
| *** mustfg2 has joined #openstack | 11:05 | |
| *** doude has quit IRC | 11:05 | |
| HugoKuo_ | perra_: did you ever try to seperate nova-network component from all-in one box ? | 11:05 |
| mustfg2 | hi, i got still an error trying to allocate an address to an instance, can someone help me ? | 11:06 |
| soren | mustfg2: If you describe the problem, somone might be able to help you. | 11:06 |
| *** ggy has quit IRC | 11:07 | |
| perra_ | HugoKuo_: can you elaborate on that, what do you mean? | 11:07 |
| HugoKuo_ | perra_ : I install all components in one machine . and then I add one more compute node , everything is working fine | 11:08 |
| HugoKuo_ | but I think the all-in-one box is too busy | 11:08 |
| HugoKuo_ | after that I want to seperate nova-network and nova-scheduler from all-in-one box ..... | 11:09 |
| perra_ | Hugo_Kuo_: that's my current setup as well | 11:09 |
| HugoKuo_ | ok got it ... | 11:09 |
| HugoKuo_ | thanks body | 11:09 |
| mustfg2 | http://paste.openstack.org/show/916/ heres the problem shown in nova-network.log any ideas ? | 11:10 |
| mustfg2 | i have 10 floating addresses in the database, but i cant allocate them... | 11:10 |
| mustfg2 | when i try euca-allocate-address the error is: "UnknownError: An unknown error has occurred. Please try your request again." | 11:11 |
| soren | mustfg2: Look in nova-network.log | 11:12 |
| mustfg2 | the link above shows the network log | 11:12 |
| soren | HugoKuo_: nova-manage project quota 0 instances 100 <---- Will let project 0 run 100 instances. | 11:13 |
| soren | mustfg2: How did you create your network? (not the floating addressses, the other ones) | 11:14 |
| CloudChris | Hello together, I am trying to bring all components in a nice graphic - I appreciate any feedback: | 11:14 |
| CloudChris | http://img145.imageshack.us/i/standjetzt.png/ | 11:14 |
| soren | CloudChris: nova-manage does not speak to the openstack api. | 11:14 |
| soren | CloudChris: It usually tweaks the database directly. | 11:15 |
| HugoKuo_ | soren : thank u so much ~! | 11:15 |
| mustfg2 | the script created the network / bridge and nova.conf | 11:15 |
| *** gdusbabek has quit IRC | 11:15 | |
| CloudChris | soren: good hint, if that's all i'm happy with it so far ;P | 11:15 |
| *** gdusbabek has joined #openstack | 11:15 | |
| soren | CloudChris: Hang on, I've only just glanced at it so far. | 11:16 |
| CloudChris | I will | 11:16 |
| *** gdusbabek has quit IRC | 11:16 | |
| *** gdusbabek has joined #openstack | 11:16 | |
| soren | CloudChris: Pretty much everything speaks to the database, actually. | 11:16 |
| mustfg2 | i running flatdhcp mode | 11:17 |
| soren | CloudChris: ..and nova-objectstore speaks to the NovaAuth thing you have, too. | 11:17 |
| mustfg2 | and the instance gets a private ip | 11:17 |
| soren | mustfg2: What is "the script"? | 11:17 |
| soren | mustfg2: Oh, ok. | 11:17 |
| mustfg2 | nova-CC-... .sh | 11:17 |
| mustfg2 | branch bexar | 11:17 |
| soren | HugoKuo_: How did you add the floating ips? | 11:18 |
| soren | Err... | 11:18 |
| soren | mustfg2: How did you add the floating ips? | 11:18 |
| *** nerens has quit IRC | 11:18 | |
| mustfg2 | nova-manage floating add or something like that | 11:22 |
| *** nerens has joined #openstack | 11:22 | |
| mustfg2 | the ip´s are shown in the database | 11:22 |
| mustfg2 | but i cant allocate them | 11:22 |
| soren | What does euca-describe-addresses say? | 11:22 |
| HugoKuo_ | wow , it's so crazy , I change instance limit to 100 , nova runs 10 instance on one nova-compute box , I got 20 instances @_@ | 11:23 |
| mustfg2 | it shows the addresses | 11:24 |
| mustfg2 | euca-describe-addresses | 11:24 |
| mustfg2 | ADDRESS172.25.10.90None (None) | 11:24 |
| mustfg2 | ... | 11:24 |
| soren | mustfg2: Ok, and what's the exact command you're running (the euca-allocate-addresss one) | 11:26 |
| *** fayce has quit IRC | 11:26 | |
| mustfg2 | euca-allocate-address <instance-id> | 11:27 |
| soren | mustfg2: That's not how it works, IIRC. | 11:28 |
| soren | mustfg2: Try just "euca-allocate-address" | 11:28 |
| mustfg2 | omg :D | 11:28 |
| mustfg2 | euca-allocate-address doesnt work either | 11:29 |
| soren | ...and then: | 11:29 |
| soren | euca-associate-address -i $instance_id $ip | 11:29 |
| soren | How does it fail? | 11:29 |
| mustfg2 | ok i check | 11:29 |
| mustfg2 | :D | 11:29 |
| mustfg2 | option -i not recognized | 11:31 |
| mustfg2 | wait i wrote allocate | 11:33 |
| mustfg2 | sry | 11:33 |
| mustfg2 | worked | 11:33 |
| mustfg2 | thx | 11:33 |
| *** reldan has joined #openstack | 11:37 | |
| *** adiantum has joined #openstack | 11:38 | |
| HugoKuo_ | instance ip is assigned by nova-network or nova-compute ? | 11:41 |
| soren | nova-network | 11:42 |
| soren | mustfg2: Sure | 11:42 |
| HugoKuo_ | by nova-network ....so why that I have to install dnsmasq on compute node ? | 11:43 |
| HugoKuo_ | I really don't understand the reason . | 11:43 |
| soren | You don't? | 11:43 |
| soren | Oh, well, yes, you do. | 11:43 |
| soren | ...but not because of Nova. Because of libvirt. | 11:43 |
| HugoKuo_ | !!!!! | 11:44 |
| openstack | HugoKuo_: Error: "!!!!" is not a valid command. | 11:44 |
| soren | I agree with openstack. "!!!!" is not a valid command. | 11:44 |
| HugoKuo_ | :> | 11:44 |
| HugoKuo_ | I had better learn more basic knowledge about nova :< | 11:45 |
| HugoKuo_ | how shame I am :< | 11:45 |
| *** toluene has quit IRC | 11:48 | |
| *** sebastianstadil has quit IRC | 11:48 | |
| *** mustfg2 has quit IRC | 11:53 | |
| uksysadmin | Are there any good guides or whitepapers on openstack for using openstack in earnest? what I'm looking for is something on best practices outside of a small set up - for example if I'm to design a hosting environment with redundancy, highlighting SPOFs and what could be done about them etc? | 11:57 |
| *** stewart has quit IRC | 11:57 | |
| uksysadmin | e.g. openstack on X number of servers hosting 1,000 instances - where to put mysql, api - does api become a SPOF? what do people do for HA? | 12:00 |
| uksysadmin | any rackspace guys here or astronauts ;-) | 12:00 |
| *** kapil has joined #openstack | 12:06 | |
| *** DigitalFlux has joined #openstack | 12:11 | |
| *** stewart has joined #openstack | 12:14 | |
| kalonike | mustfg2, ur vm's running now? | 12:16 |
| kalonike | what is the difference between nova 2011.1 and 2011.1.1? which one should i install? | 12:18 |
| kalonike | i just would like a single machine demo environment | 12:19 |
| *** miclorb has quit IRC | 12:22 | |
| *** Hexadecimal has joined #openstack | 12:40 | |
| *** BinaryBlob has joined #openstack | 12:41 | |
| *** bcwaldon has joined #openstack | 12:44 | |
| kalonike | if i install from tar.gz where should i put in the filesystem? is it important where i put?? | 12:45 |
| kalonike | please, anyone who had a succesful nova install, could u help me a little??? | 12:46 |
| *** mustfg2 has joined #openstack | 12:54 | |
| *** bkkrw has joined #openstack | 12:56 | |
| soren | kalonike: If you can, I highly recommend using Ubuntu packages instead. | 12:57 |
| kalonike | ok, which one? nova-core/release works? | 12:59 |
| kalonike | or is there any bexar repository? | 12:59 |
| *** dprince has joined #openstack | 13:00 | |
| kalonike | ubuntu packages have 0.9.1 version | 13:02 |
| kalonike | that is still austen, no? | 13:02 |
| soren | nova-core/release is bexar | 13:06 |
| soren | Our most recent release is bexar, so that's what's in nova-core/release. | 13:06 |
| kalonike | thanks for ur answer | 13:07 |
| *** hadrian has joined #openstack | 13:07 | |
| *** hggdh has quit IRC | 13:09 | |
| *** Cen has joined #openstack | 13:17 | |
| *** patcoll has joined #openstack | 13:20 | |
| *** hggdh has joined #openstack | 13:21 | |
| *** dendro-afk is now known as dendrobates | 13:22 | |
| kalonike | have u met with this? /usr/bin/nova-manage does not match any options: user project role shell vpn floating network service log db volume | 13:24 |
| kalonike | when i try to use nova-manage user admin testuser | 13:24 |
| kalonike | after nova-manage db sync | 13:25 |
| *** hggdh has joined #openstack | 13:25 | |
| kalonike | which was successful first | 13:25 |
| kalonike | but i cant repeat | 13:25 |
| kalonike | and no matter what kind of nova-manage command i use, get the same answer | 13:26 |
| *** mahadev has joined #openstack | 13:29 | |
| *** h0cin has joined #openstack | 13:31 | |
| *** pvo has joined #openstack | 13:31 | |
| *** Lantizia has quit IRC | 13:32 | |
| kalonike | i have installed 2011.1 bzr645 | 13:34 |
| *** jano has quit IRC | 13:35 | |
| *** berlin2011 has joined #openstack | 13:36 | |
| *** bcwaldon has quit IRC | 13:40 | |
| *** fysa has quit IRC | 13:44 | |
| jaypipes | sirp_, dabo: hope you can assist me. I've asked for your reviews on the VMWare VSphere merge proposal, specifically. | 13:45 |
| * soren also takes a look | 13:46 | |
| jaypipes | soren: ty | 13:46 |
| soren | "62 branches proposed for merging into this one."... It's going to be a busy week. | 13:47 |
| ttx | nah | 13:48 |
| ttx | soren: only 18 *have to* be reviewed this week | 13:49 |
| *** cerdea has joined #openstack | 13:49 | |
| *** Shentonfreude has joined #openstack | 13:49 | |
| kalonike | hm, i have removed all nova packages, and reinstalled them, and now nova-manage works :) | 13:50 |
| *** cerdea has quit IRC | 13:50 | |
| *** mahadev has quit IRC | 13:51 | |
| *** patcoll has quit IRC | 13:51 | |
| soren | ttx: What's the best view for finding that out? The releasestatus page? | 13:52 |
| *** patcoll has joined #openstack | 13:52 | |
| ttx | soren: at this point, my email gives the list... Working on my priority generator as we speak. | 13:53 |
| soren | ttx: Cool. | 13:54 |
| ttx | soren: my evil scoring algorithm currently gives that: http://pastebin.com/PSRVyZGZ | 13:55 |
| soren | ttx: That is way cool! | 13:57 |
| westmaas | ttx: yeah, looks pretty good | 13:57 |
| *** Shentonfreude has quit IRC | 13:57 | |
| ttx | i'll probably release the python code before I HTMLize the output. I suck so much at that that it could take time | 13:58 |
| soren | Cool beans | 13:59 |
| jaypipes | wow, someone's actually trying to use the Hyper-V virt layer... this should be interesting. | 14:00 |
| soren | Yikes. | 14:00 |
| *** ryker has joined #openstack | 14:03 | |
| *** zenmatt has joined #openstack | 14:04 | |
| soren | This is starting to come together quite nicely: | 14:04 |
| soren | https://jenkins.linux2go.dk/ | 14:05 |
| pvo | looks good soren | 14:07 |
| soren | hggdh: You'll probably appreciate it as well.. ^ | 14:07 |
| * soren likes the green | 14:07 | |
| * hggdh goes look | 14:08 | |
| *** kbringard has joined #openstack | 14:08 | |
| *** MarcMorata has quit IRC | 14:08 | |
| hggdh | ooohhhhh | 14:08 |
| * hggdh notes down the location, and the master behind it | 14:08 | |
| kbringard | hey guys, quick ? | 14:08 |
| soren | hggdh: nova-install runs whenever there's a new package in the ppa. | 14:08 |
| westmaas | soren: nice | 14:08 |
| soren | hggdh: ...and that happens automatically once there's a commit on trunk. | 14:08 |
| kbringard | is there a reason 120s was chosen as the default lease time in the --dhcp-range flag | 14:09 |
| kbringard | in linux_net.py | 14:09 |
| hggdh | soren: this is *very* cool | 14:09 |
| soren | kbringard: Someone rolled a 120 sided die. | 14:09 |
| kbringard | soren: lol | 14:09 |
| kbringard | I thought I was awesome because I roll 20s | 14:09 |
| kbringard | the reason I ask is because, if I understand dnsmasq correctly, once the lease expires, it'll no longer return DNS info | 14:10 |
| kbringard | but, since we're essentially using static addressing (since nova-network manages it) | 14:10 |
| kbringard | changing it to an infinite lease time (in theory) shouldn't hurt anything, and would cause DNS lookups to work for as long as the VM is alive | 14:11 |
| kbringard | but I wanted to ask if there was a reason it was set to 2 minutes... since y'all know the entire chain of events far better than me | 14:11 |
| soren | Not sure about the history | 14:12 |
| kbringard | coo, no worries | 14:12 |
| soren | I think it's somewhat convenient, though, that the instances check in with the dhcp server at least once every two minutes. | 14:13 |
| kbringard | yea, good point | 14:13 |
| kbringard | maybe I've just built my instances wrong | 14:13 |
| kbringard | the leases seem to be expiring and DNS stops working for them | 14:13 |
| kbringard | or maybe I'm smoking crack and have the whole thing setup wrong ;-) | 14:14 |
| soren | They should be refreshing their dhcp lease. | 14:14 |
| *** dspano has joined #openstack | 14:15 | |
| * soren calls it a week | 14:16 | |
| kbringard | have a good weekend soren | 14:17 |
| kbringard | thanks for the insight | 14:17 |
| soren | Likewise | 14:17 |
| *** gondoi has joined #openstack | 14:17 | |
| *** guigui has quit IRC | 14:18 | |
| *** dendrobates is now known as dendro-afk | 14:18 | |
| *** Shentonfreude has joined #openstack | 14:19 | |
| *** pvo has quit IRC | 14:19 | |
| uvirtbot | New bug: #737574 in nova "'module' object has no attribute 'id_to_ec2_id'" [Undecided,In progress] https://launchpad.net/bugs/737574 | 14:21 |
| *** Paul_Miller has joined #openstack | 14:23 | |
| *** dendro-afk is now known as dendrobates | 14:26 | |
| *** msassak has joined #openstack | 14:32 | |
| *** bcwaldon has joined #openstack | 14:36 | |
| *** BinaryBlob has quit IRC | 14:37 | |
| *** koolhead17 is now known as koolhead17|afk | 14:37 | |
| *** rnirmal has joined #openstack | 14:40 | |
| *** hggdh has quit IRC | 14:40 | |
| *** perra_ has quit IRC | 14:40 | |
| kalonike | my new problem: failed to parse error message from AWS: unknown:1:0: syntax error EC2ResponseError: 403 forbidden Access denied to this resource | 14:41 |
| kalonike | i am running as root | 14:41 |
| kalonike | and i have sourced novarc | 14:41 |
| kalonike | what the hell is the problem? | 14:42 |
| kalonike | i did everything in the same way as earlier | 14:42 |
| *** hggdh has joined #openstack | 14:43 | |
| *** ppetraki has joined #openstack | 14:43 | |
| *** maplebed has joined #openstack | 14:43 | |
| kalonike | is there any nova system which i could login?? | 14:44 |
| kalonike | i am on the way going crazy, not being able to install nova | 14:44 |
| kalonike | but i would need to my work very urgent a working nova system | 14:44 |
| *** oldbam has joined #openstack | 14:46 | |
| *** pvo has joined #openstack | 14:47 | |
| oldbam | is there a way in swauth to assign user to a custom group (not .reseller_admin, account:user, account) via ReST? | 14:48 |
| *** dfg has joined #openstack | 14:50 | |
| jaypipes | dendrobates: reminder.. it's your review day today. | 14:54 |
| ttx | soren, jaypipes, justinsb: basic version of the prioritizer at https://code.launchpad.net/~ttx/+junk/reviewlist | 14:54 |
| ttx | verrry slow. | 14:54 |
| jaypipes | dendrobates: sorry, not it's not. n,m. | 14:54 |
| ttx | verrry raw. | 14:54 |
| dendrobates | jaypipes: did it change? | 14:54 |
| jaypipes | dendrobates: no, I'm an idiot. | 14:54 |
| dendrobates | it was 3/22 | 14:54 |
| dendrobates | oh | 14:54 |
| jaypipes | dendrobates: it's josh mckenty's day today... and I doubt we will hear from him somehow.. | 14:55 |
| dendrobates | my RD is my 2nd day at cisco, so I might have to do it in the evening | 14:56 |
| *** kashyap has quit IRC | 14:56 | |
| *** romain_lenglet has joined #openstack | 14:57 | |
| *** mahadev has joined #openstack | 14:58 | |
| *** littleidea has joined #openstack | 15:00 | |
| *** zaitcev has joined #openstack | 15:06 | |
| *** maplebed has quit IRC | 15:08 | |
| *** Lantizia has joined #openstack | 15:09 | |
| *** f4m8 is now known as f4m8_ | 15:12 | |
| *** dragondm has joined #openstack | 15:12 | |
| *** reldan has quit IRC | 15:13 | |
| pvo | dendrobates: you could switch with someone | 15:14 |
| pvo | take today instead of mckenty? :) | 15:14 |
| *** Paul_Miller has quit IRC | 15:15 | |
| *** guigui has joined #openstack | 15:16 | |
| *** reldan has joined #openstack | 15:17 | |
| *** dragondm has quit IRC | 15:18 | |
| *** littleidea has quit IRC | 15:19 | |
| *** mahadev has quit IRC | 15:21 | |
| *** dragondm has joined #openstack | 15:24 | |
| *** j05h has joined #openstack | 15:24 | |
| *** dragondm has quit IRC | 15:24 | |
| *** romain_lenglet has left #openstack | 15:24 | |
| *** dragondm has joined #openstack | 15:25 | |
| *** littleidea has joined #openstack | 15:26 | |
| *** bcwaldon has quit IRC | 15:26 | |
| *** woleium has joined #openstack | 15:29 | |
| *** CloudChris has quit IRC | 15:29 | |
| *** imsplitbit has joined #openstack | 15:30 | |
| *** daveiw has quit IRC | 15:33 | |
| *** oldbam has left #openstack | 15:39 | |
| *** mahadev has joined #openstack | 15:40 | |
| *** littleidea has quit IRC | 15:40 | |
| *** omidhdl has joined #openstack | 15:42 | |
| *** ppetraki has quit IRC | 15:48 | |
| *** mahadev has quit IRC | 15:51 | |
| *** enigma1 has joined #openstack | 15:52 | |
| *** sophiap has joined #openstack | 15:52 | |
| *** enigma1 has left #openstack | 15:53 | |
| *** enigma1 has joined #openstack | 15:53 | |
| *** bcwaldon has joined #openstack | 15:54 | |
| *** uksysadmin has quit IRC | 16:02 | |
| *** mahadev has joined #openstack | 16:09 | |
| *** maplebed has joined #openstack | 16:09 | |
| *** berlin2011 has quit IRC | 16:18 | |
| *** littleidea has joined #openstack | 16:23 | |
| *** dendrobates is now known as dendro-afk | 16:24 | |
| *** littleidea has quit IRC | 16:25 | |
| *** dendro-afk is now known as dendrobates | 16:28 | |
| kbringard | hmmm, nova-network doesn't seem to be updating my iptables rules... has anyone seen this before? | 16:28 |
| openstackjenkins | Project nova build #665: SUCCESS in 2 min 7 sec: http://hudson.openstack.org/job/nova/665/ | 16:28 |
| openstackjenkins | Tarmac: Mark instance metadata as deleted when we delete the instance | 16:28 |
| *** skiold has quit IRC | 16:28 | |
| *** [Crono] has joined #openstack | 16:31 | |
| *** [Crono] has left #openstack | 16:32 | |
| *** stewart has quit IRC | 16:33 | |
| *** stewart has joined #openstack | 16:33 | |
| *** guigui has quit IRC | 16:34 | |
| *** littleidea has joined #openstack | 16:34 | |
| *** woleium_ has joined #openstack | 16:39 | |
| *** guigui1 has joined #openstack | 16:39 | |
| *** woleium has quit IRC | 16:41 | |
| *** woleium_ is now known as woleium | 16:41 | |
| jk0 | Happy Friday everyone! :) http://bit.ly/ifqyIF | 16:44 |
| *** reldan has quit IRC | 16:44 | |
| ttx | More confusion on the Openstack API python bindings with https://github.com/jacobian/openstack.compute | 16:47 |
| justinsb | ttx: What is our position at the moment? Are we pulling a set of bindings into trunk? | 16:48 |
| ttx | justinsb: no, they are shipped seperately as python-novaclient | 16:48 |
| ttx | but my effort duplication detector went wild | 16:49 |
| *** z0 has quit IRC | 16:49 | |
| *** dprince has quit IRC | 16:52 | |
| ttx | jk0: happy Friday too -- my weekend starts now. | 16:52 |
| *** kapil_ has joined #openstack | 16:53 | |
| *** kapil has quit IRC | 16:53 | |
| *** bkkrw has quit IRC | 16:55 | |
| RobertLaptop | jk0, Now I have that stupid song in my head | 16:57 |
| jk0 | :) | 16:57 |
| *** littleidea has quit IRC | 16:57 | |
| *** joshuamckenty has joined #openstack | 16:59 | |
| joshuamckenty | jk0: Are you going to review 53064, or do you want me to? | 17:00 |
| joshuamckenty | It's my day | 17:00 |
| joshuamckenty | :) | 17:00 |
| jaypipes | sirp_: I'd love to get some reviews from you for the Glance bug fixes if you have spare cycles today.. | 17:00 |
| *** joearnold has joined #openstack | 17:01 | |
| *** Nacx has quit IRC | 17:01 | |
| sirp_ | jaypipes: sure, *been focused on the nova backlog as of late* | 17:01 |
| jaypipes | joshuamckenty: hella backlog. as many reviews as you can do would be appreciated. | 17:01 |
| jaypipes | sirp_: me 2 until yesterday... | 17:01 |
| joshuamckenty | Understood. I'm on spotty internets today, so I'm going to grab a ton of branches and review them offline | 17:02 |
| *** joearnold has quit IRC | 17:02 | |
| jk0 | joshuamckenty: I've been slowly perusing and testing it the past few days, but you could have a look too if you want :) | 17:02 |
| joshuamckenty | I had just seen that jk0 had started the review on that one already, didn't want to cross wires | 17:02 |
| jaypipes | joshuamckenty: either that, or re-assign your review day to a different Anso-er that can pick up the slack. We're really struggling. | 17:02 |
| *** joearnold has joined #openstack | 17:03 | |
| jk0 | joshuamckenty: actually, you can skip 53064. I will take it and try to get soren to look too | 17:03 |
| *** z0 has joined #openstack | 17:03 | |
| joshuamckenty | I'll see if I can get some back-up, but I should be able to get a bunch of these done | 17:03 |
| joshuamckenty | jk0: k | 17:03 |
| joshuamckenty | jk0: anyway, the ntt folks should be putting an NTT copyright line in, unless they've got a mechanism to assign to OpenStack LLC. | 17:04 |
| *** joshuamckenty has quit IRC | 17:04 | |
| *** reldan has joined #openstack | 17:06 | |
| justinsb | ttx: Your script is awesome | 17:06 |
| *** littleidea has joined #openstack | 17:07 | |
| *** aliguori has quit IRC | 17:07 | |
| justinsb | ttx: Though I don't like the score it gives to my favorite MP: https://code.launchpad.net/~justin-fathomdb/nova/more-focused-pylint/+merge/53711 :-) | 17:07 |
| *** littleidea has left #openstack | 17:10 | |
| *** bkkrw has joined #openstack | 17:10 | |
| *** ramd has quit IRC | 17:14 | |
| pvo | justinsb: Should we discuss moving the pylint requirements? Seems something we shoudl talk about as a community rather than just merge in, no? | 17:14 |
| justinsb | pvo: Sorry, don't really understand what you mean by "moving". Do you mean moving like "changing"? | 17:15 |
| pvo | s/moving/removing/ requirements | 17:15 |
| justinsb | pvo: I think this is actually a bit of a no-brainer... the big change is not to warn about "_" | 17:15 |
| justinsb | pvo: Ah | 17:16 |
| justinsb | pvo: I think we can all agree that pylint has way too many false positives at the moment | 17:16 |
| *** joearnold has quit IRC | 17:16 | |
| pvo | maybe I misread.. but weren't you suggesting we remove the require docstrings part? | 17:16 |
| *** kbringard has quit IRC | 17:16 | |
| justinsb | pvo: I think the "_" fix should be 100% non-controversial. That's more working around a pylint bug | 17:16 |
| justinsb | pvo: The docstrings warning I agree is not cut and dry | 17:17 |
| justinsb | pvo: My thought was that nobody is checking pylint at the moment | 17:17 |
| pvo | I do when I merge branches | 17:17 |
| justinsb | pvo: Ideally we'd just have docstrings warn on public functions | 17:17 |
| btorch | jaypipes: I downloaded lp:~jaypipes/glance/bug736295 , built and installed it. When I do the update it still won't set it to true | 17:17 |
| justinsb | pvo: But I don't think we can do that at the moment | 17:17 |
| justinsb | pvo: I'm glad you're checking :-) | 17:18 |
| *** ramkrsna has quit IRC | 17:18 | |
| justinsb | pvo: I don't really care about the docstrings... eday set up an "errors only" job, and with "_", that then shows things that look like real errors only | 17:18 |
| justinsb | pvo: So that's top priority for me | 17:18 |
| openstackjenkins | Project nova build #666: SUCCESS in 2 min 8 sec: http://hudson.openstack.org/job/nova/666/ | 17:18 |
| openstackjenkins | Tarmac: Re-implementation (or just implementation in many cases) of Limits in the OpenStack API. Limits is now available through /limits and the concept of a limit has been extended to include arbitrary regex / http verb combinations along with correct XML/JSON serialization. Tests included. | 17:18 |
| justinsb | pvo: I think the right thing to do on docstrings is to have yet another pylint run which tells us about docstrings and other non-code issues | 17:19 |
| pvo | fair enough. | 17:19 |
| mustfg2 | can someone tell me if a user using nova-compute is supposed to use the nova-manage commands ? | 17:19 |
| justinsb | pvo: But unless we whittle down the lists first, it's just too intimidating to attack | 17:19 |
| justinsb | pvo: Certainly let's talk about this at the design conference... I have a proposal along those lines | 17:20 |
| justinsb | pvo: But my vote would be to get ignore-docstrings in for the moment, until the pylint list is shorter | 17:20 |
| *** joearnold has joined #openstack | 17:21 | |
| *** kalonike has quit IRC | 17:22 | |
| jaypipes | btorch: error or log output? | 17:23 |
| jaypipes | btorch: isn't the functional test case I added in http://bazaar.launchpad.net/~jaypipes/glance/bug736295/revision/96 exactly the series of commands you run? | 17:24 |
| mustfg2 | to concrete my question can i use nova-manage commands remote? | 17:24 |
| *** z0 has quit IRC | 17:25 | |
| dragondm | mustfg2: no. There is some of that functionality available thru the api if you have the admin api turned on, tho. | 17:28 |
| mustfg2 | ok | 17:28 |
| jaypipes | sirp_: I don't think our test case for migrations works at all... when I run it with debug, I don't see a single SQL statement echo'd from sqlalchemy... | 17:32 |
| *** anticw has quit IRC | 17:33 | |
| *** anticw has joined #openstack | 17:33 | |
| sirp_ | jaypipes: uhhh, i thought i had it working when i committed *scratches head* | 17:33 |
| sirp_ | jaypipes: although migrations are so weird with sa-migrate, i'm honest not 100% sure | 17:34 |
| jaypipes | sirp_: yeah... | 17:34 |
| jaypipes | sirp_: I'm not sure either :( | 17:34 |
| jaypipes | sirp_: because I get the impressions that DDL statements aren't echo'd in the same way as DML statements in SQLAlchemy. :( | 17:35 |
| sirp_ | jaypipes: possible,, guess you could make sure the test db is not deleted after the test, and then inspect it | 17:36 |
| jaypipes | sirp_: http://paste.openstack.org/show/929/ | 17:36 |
| btorch | jaypipes: ok it works .. was having problems due to my pythonpath but all good now, do you want me to place any comments on the ticket ? my output ? | 17:37 |
| jaypipes | sirp_: see how when I execute the INSERTs and SELECTs that those statements get echo'd, but the migrate doesn't echo anything? | 17:37 |
| jaypipes | btorch: on the merge proposal, just note that you tested it and it works :) | 17:37 |
| *** z0 has joined #openstack | 17:37 | |
| jaypipes | btorch: https://code.launchpad.net/~jaypipes/glance/bug736295/+merge/54002 | 17:38 |
| sirp_ | jaypipes: hmm i wonder if sa-migrate is not passing through a echo=True to the engine when it applies the DDL | 17:39 |
| sirp_ | jaypipes: but in our code we do use a echo=True | 17:39 |
| jaypipes | sirp_: hmm, lemme check... | 17:40 |
| *** mustfg2 has quit IRC | 17:41 | |
| *** guigui1 has quit IRC | 17:41 | |
| jaypipes | sirp_: shit, no idea. it's not clear how upgrade() even received its param (migrate_engine). It's not passed via anything in glance/registry/db/migration.py :( | 17:42 |
| *** bkkrw has quit IRC | 17:43 | |
| sirp_ | jaypipes: it's a hook passed from sa-migrate, no? | 17:43 |
| sirp_ | jaypipes: can you pprint the engine.__dict__ and see if echo is set? | 17:43 |
| jaypipes | sirp_: where is that hook? | 17:45 |
| *** MotoMilind has joined #openstack | 17:45 | |
| *** Ryan_Lane has joined #openstack | 17:46 | |
| sirp_ | jaypipes: no idea, would have to hunt around sa-migrate source-code to find where it kicks it off, but imagine itsomething like MigrationRunner or some such | 17:46 |
| jaypipes | sirp_: http://paste.openstack.org/show/930/ is the pprint(migrate_engine.__dict__)... | 17:47 |
| *** joearnol_ has joined #openstack | 17:47 | |
| sirp_ | jaypipes; the engine has an engine (guess it wraps it), can you pprint migrate_enging.engine | 17:48 |
| *** joearnold has quit IRC | 17:49 | |
| *** joearnol_ has quit IRC | 17:49 | |
| jaypipes | sirp_: hehe, already tried that. prints the exact same thing: http://paste.openstack.org/show/931/ | 17:49 |
| jaypipes | retarded. | 17:49 |
| sirp_ | haha, obj graphs shouldn't have cycles *ugh* | 17:50 |
| jaypipes | yeah, ugh. | 17:50 |
| sirp_ | welll | 17:50 |
| *** pmullaney has joined #openstack | 17:50 | |
| jaypipes | sirp_: I did a migrate_engine.engine.echo = True and got more output... | 17:52 |
| sirp_ | jaypipes: cool | 17:52 |
| *** joearnold has joined #openstack | 17:57 | |
| openstackjenkins | Project nova build #667: SUCCESS in 2 min 6 sec: http://hudson.openstack.org/job/nova/667/ | 17:58 |
| openstackjenkins | Tarmac: Change cloud.id_to_ec2_id to ec2utils.id_to_ec2_id. Fixes EC2 API error handling when invalid instances and volume names are specified. | 17:58 |
| *** kyzh is now known as kyzh|off | 17:58 | |
| *** joearnold has quit IRC | 18:03 | |
| openstackjenkins | Project nova build #668: SUCCESS in 2 min 5 sec: http://hudson.openstack.org/job/nova/668/ | 18:04 |
| openstackjenkins | Tarmac: Replaced all pylint "disable-msg=" with "disable=" and "enable-msg=" with "enable=". | 18:04 |
| *** kapil_ is now known as kapil | 18:06 | |
| *** kapil is now known as hazmat | 18:06 | |
| *** hazmat has joined #openstack | 18:06 | |
| *** kashyap has joined #openstack | 18:08 | |
| jaypipes | sirp_: so, this test case: http://paste.openstack.org/show/933/ is producing this loveliness: http://paste.openstack.org/show/932/ | 18:09 |
| *** adiantum has quit IRC | 18:11 | |
| jaypipes | sirp_: the ALTER TABLE ... RENAME TO migrate_temp is the hack needed because SQLite doesn't support ALTER TABLE .. DROP COLUMN. (and we drop the type column on images..) | 18:11 |
| jaypipes | sirp_: note: the hack is in sa-migrate, not in our code.... | 18:12 |
| jaypipes | sirp_: if I add back the index=True on the deleted and is_public columns in Images, that hack dies a horrible death (and that's what the bug report http://code.google.com/p/sqlalchemy-migrate/issues/detail?id=117 is all about) | 18:13 |
| openstackjenkins | Project nova build #669: SUCCESS in 2 min 5 sec: http://hudson.openstack.org/job/nova/669/ | 18:13 |
| openstackjenkins | Tarmac: Cleanup of FakeAuthManager | 18:13 |
| jaypipes | sirp_: oh, and why all log output is doubled-up is a complete mystery to me... | 18:14 |
| sirp_ | jaypipes: wonder if somehow we attached to StreamHandlers to the logging object | 18:15 |
| sirp_ | *two | 18:15 |
| *** joearnold has joined #openstack | 18:15 | |
| *** dendrobates is now known as dendro-afk | 18:23 | |
| *** woleium has quit IRC | 18:25 | |
| *** beepboop has joined #openstack | 18:25 | |
| beepboop | hi all, got kind of an urgent help request for any devs on the channel | 18:26 |
| beepboop | rabbitmq's beam.smp process borked; high CPU usage and no messages placed in the networking queue are getting consumed | 18:27 |
| *** woleium has joined #openstack | 18:28 | |
| *** spectorclan_ has joined #openstack | 18:28 | |
| beepboop | has anyone seen this before? found the/any root cause/s? is there any good way to restart rabbitmq without restarting all the nova management services? | 18:28 |
| spectorclan_ | OpenStack Governance Nominations Close Tonight at Midnight CST - check http://etherpad.openstack.org/Spring2011-Elections for nominations and let me know if you wish to be removed. Thanks, | 18:30 |
| pvo | beepboop: we do see it from time to time. Have to bounce rabbit services to get it to recover. | 18:30 |
| pvo | I haven't done it with nova though yet. | 18:30 |
| beepboop | last time I just restarted the rabbitmq service things got strange and had to eventually stop/start the nova api & scheduler services | 18:31 |
| *** dweimer has joined #openstack | 18:32 | |
| *** lionel_ has quit IRC | 18:32 | |
| pvo | beepboop: yea, thats usually what we do when rabbit freaks out | 18:32 |
| *** lionel_ has joined #openstack | 18:32 | |
| beepboop | ok. just restarted everything and waiting for the network node to re-register itself | 18:33 |
| beepboop | pvo: yeah, everything came back fine. thanks man | 18:38 |
| pvo | beepboop: np. | 18:38 |
| *** woleium_ has joined #openstack | 18:39 | |
| *** clauden has joined #openstack | 18:40 | |
| *** woleium has quit IRC | 18:41 | |
| *** woleium_ is now known as woleium | 18:41 | |
| *** aliguori has joined #openstack | 18:49 | |
| dweimer | I just updated swift to trunk (branch 251) and I'm now getting 403 errors from the proxy-server. So far I have tracked it down to the swauth 'authorize' function and it is hitting the denied_response call at the end. parse_acl is also being called with 'None' so there are no groups to match against. Any thoughts where I should look next? | 18:50 |
| *** dweimer is now known as dweimer_ | 18:57 | |
| *** DigitalFlux has joined #openstack | 18:57 | |
| *** Ryan_Lane has quit IRC | 19:04 | |
| *** kashyap has quit IRC | 19:12 | |
| *** imsplitbit has quit IRC | 19:29 | |
| *** aliguori has quit IRC | 19:30 | |
| *** littleidea has joined #openstack | 19:32 | |
| *** photron has joined #openstack | 19:34 | |
| *** jbryce_ has joined #openstack | 19:38 | |
| *** reldan has quit IRC | 19:38 | |
| *** jbryce has quit IRC | 19:40 | |
| *** jbryce_ is now known as jbryce | 19:40 | |
| *** aliguori has joined #openstack | 19:42 | |
| btorch | ok I have setup my nova.conf to use glance and also changed my S3_URL to point to my glance box. using euca-describe-images I see that it can reach the glance box but it keeps on failing when I try to uec-publish-tarball. I've tried giving bucket as images and glance and bucket-1 but nothing works | 19:45 |
| btorch | http://paste.openstack.org/show/934/ | 19:45 |
| btorch | the glance bucket exists in swift and I can upload an image fine using the glance tool | 19:45 |
| *** DigitalFlux has quit IRC | 19:48 | |
| *** joearnold has quit IRC | 19:48 | |
| *** spectorclan_ has quit IRC | 19:49 | |
| *** bcwaldon has quit IRC | 19:49 | |
| *** bcwaldon has joined #openstack | 19:50 | |
| *** dweimer_ is now known as dweimer | 19:58 | |
| kpepple_ | anyone running nova under postgres and has a moment to help me set it up ? trying to test my fix for lp734680 | 19:59 |
| btorch | kpepple_: I'm not right now but I might be able to help you with postgres | 20:00 |
| *** brd_from_italy has joined #openstack | 20:01 | |
| btorch | kpepple_: pg 8 or 9 ? | 20:02 |
| kpepple_ | btorch: thanks. i've created a database "nova" and assigned the nova user a password "adgth7". is this the correct nova.conf parameters -- http://paste.openstack.org/show/935/ ? | 20:02 |
| *** MotoMilind has quit IRC | 20:02 | |
| kpepple_ | btorch: pg8.4 on ubuntu 10.10 | 20:02 |
| *** woleium has quit IRC | 20:06 | |
| *** littleidea has quit IRC | 20:06 | |
| btorch | kpepple_: give me a sec trying out here | 20:08 |
| *** MotoMilind has joined #openstack | 20:09 | |
| *** ianw has joined #openstack | 20:09 | |
| btorch | kpepple_: what part is failing for you ? | 20:12 |
| kpepple_ | btorch: when i do nova-manage db sync, it doesn't seem to connect to the DB. do i need a different driver ? | 20:13 |
| btorch | kpepple_: have you made any modifications to pg_hba.conf ? | 20:13 |
| kpepple_ | btorch: i'm a noob on postgres ... what is pg_hba.conf ? | 20:14 |
| btorch | kpepple_: ok that's the authentication file .. but it's cool if you haven't modified | 20:14 |
| btorch | kpepple_: then make sure you have python-psycopg2 installed | 20:14 |
| btorch | kpepple_: I changed my setup from mysql to pg and that was the problem | 20:15 |
| btorch | kpepple_: after installing the pg python module the tables were loaded | 20:15 |
| *** reldan has joined #openstack | 20:15 | |
| dragondm | yah, you will need that to talk to postgres. | 20:15 |
| kpepple_ | btorch: yes, i installed that | 20:16 |
| kpepple_ | do i need to reference that driver in the nova.conf file ? | 20:17 |
| *** romain_lenglet_ has joined #openstack | 20:17 | |
| btorch | kpepple_: try this and see if you can actually access the database psql -U nova -h localhost nova -l | 20:17 |
| *** romain_lenglet_ is now known as romain_lenglet | 20:17 | |
| btorch | kpepple_: I haven't .. I'm using 2011.2~bzr758 version of nova though | 20:17 |
| kpepple_ | btorch: so that works ... but i don't have anything listed under Access priviledges for the nova user | 20:18 |
| btorch | kpepple_: that's fine | 20:18 |
| btorch | kpepple_: what nova version are you using ? | 20:18 |
| kpepple_ | btorch: trunk ... with my bug fixes | 20:19 |
| btorch | kpepple_: cool | 20:19 |
| *** jheiss_ has joined #openstack | 20:19 | |
| *** jheiss has quit IRC | 20:19 | |
| *** jheiss_ is now known as jheiss | 20:19 | |
| btorch | kpepple_: try removing that --auth_driver flag and restarting nova and see if that helps | 20:19 |
| *** nerens has quit IRC | 20:20 | |
| *** aliguori has quit IRC | 20:20 | |
| *** elasticdog has quit IRC | 20:20 | |
| *** romain_lenglet_ has joined #openstack | 20:21 | |
| *** romain_lenglet has quit IRC | 20:21 | |
| *** romain_lenglet_ is now known as romain_lenglet | 20:21 | |
| *** hazmat has quit IRC | 20:22 | |
| kpepple_ | btorch: didn't help :( | 20:24 |
| *** joshuamckenty has joined #openstack | 20:25 | |
| btorch | kpepple_: hmm ... we know your pg_hba.conf is correct cause you can access the db | 20:26 |
| *** omidhdl has left #openstack | 20:26 | |
| *** joshuamckenty has quit IRC | 20:26 | |
| *** joearnold has joined #openstack | 20:27 | |
| btorch | kpepple_: I'm not sure how well nova has been tested with pg cause even though I got the db sync to populate the database nova .. nova-manage gives me quite a few errors and nova-manage user delete doesn't work | 20:27 |
| *** dendro-afk is now known as dendrobates | 20:27 | |
| kpepple_ | btorch: yeah, i'm actually trying to fix a pg problem with nova-manage instance_types ... let me start up a clean virtual server and try this | 20:28 |
| btorch | yeah in the nova-manage.log I see pg errors when trying to delete the user | 20:28 |
| btorch | but then again I'm using an older nova revision than you so it might be fixed on yours | 20:28 |
| kpepple_ | btorch: yeah, it's probably the issue with setting deleted=0 when we've defined deleted as a boolean. it gets cast fine for us in mysql but postgres b0rks on it. | 20:29 |
| btorch | yep | 20:29 |
| *** nerens has joined #openstack | 20:30 | |
| *** taihen has quit IRC | 20:32 | |
| btorch | kpepple_: yeah postgres likes values to be between single quotes | 20:32 |
| btorch | kpepple_: http://paste.openstack.org/show/936/ | 20:32 |
| *** arun_ has joined #openstack | 20:35 | |
| *** arun_ has joined #openstack | 20:35 | |
| *** z0 has quit IRC | 20:36 | |
| *** aliguori has joined #openstack | 20:37 | |
| *** CloudChris has joined #openstack | 20:39 | |
| *** arun_ has quit IRC | 20:39 | |
| *** joshuamckenty has joined #openstack | 20:39 | |
| *** dendrobates is now known as dendro-afk | 20:45 | |
| *** arun_ has joined #openstack | 20:45 | |
| *** sophiap has quit IRC | 20:49 | |
| *** littleidea has joined #openstack | 20:49 | |
| *** kapil has joined #openstack | 20:52 | |
| *** bkkrw has joined #openstack | 20:54 | |
| *** dfg has quit IRC | 21:01 | |
| *** h0cin has quit IRC | 21:04 | |
| *** bcwaldon has quit IRC | 21:09 | |
| joshuamckenty | hey ironcamel I just finished a review of https://code.launchpad.net/~rackspace-titan/nova/openstack-api-versioned-controllers/+merge/53748 | 21:09 |
| joshuamckenty | Mostly questions / clarifications | 21:09 |
| *** mgoldmann has quit IRC | 21:10 | |
| *** romain_lenglet has quit IRC | 21:12 | |
| *** bkkrw has quit IRC | 21:15 | |
| *** Ryan_Lane has joined #openstack | 21:15 | |
| westmaas | joshuamckenty: thanks a ton! he is out of town for the weekend but I'm sure he'll see it on Tuesday | 21:17 |
| westmaas | thanks for the other titan reviews too :) | 21:17 |
| joshuamckenty | cool | 21:17 |
| joshuamckenty | goin through them now | 21:17 |
| joshuamckenty | sandywalsh and xtoddx : are we ready for another review of https://code.launchpad.net/~sandy-walsh/nova/zones3/+merge/52565 ? | 21:18 |
| westmaas | ah, I was thinking you hit the others up already, but I guess that was other people! | 21:18 |
| joshuamckenty | westmaas: not that cool, I'm afraid. I've been too long from some parts of the code, and had to read quite a bit of it to catch up | 21:19 |
| westmaas | heh. Last night was the first time I got to sit down and pair with one of my guys. Was pretty fun, but yeah, I lot to read. | 21:19 |
| *** beepboop has quit IRC | 21:24 | |
| *** bkkrw has joined #openstack | 21:28 | |
| *** MotoMilind has quit IRC | 21:28 | |
| *** bcwaldon has joined #openstack | 21:35 | |
| btorch | kpepple_: hope u get it working :) later | 21:39 |
| *** dspano has quit IRC | 21:41 | |
| kpepple_ | btorch: thx | 21:42 |
| *** joshuamckenty has quit IRC | 21:44 | |
| devcamcar | mtaylor: are you around? | 21:44 |
| *** alekibango has quit IRC | 21:49 | |
| *** dirakx has quit IRC | 21:49 | |
| *** CloudChris has quit IRC | 21:51 | |
| *** allsystemsarego has quit IRC | 21:54 | |
| *** CloudChris has joined #openstack | 21:54 | |
| *** patcoll has quit IRC | 21:59 | |
| *** Shentonfreude has quit IRC | 22:00 | |
| *** jc_smith has joined #openstack | 22:04 | |
| *** jheiss has quit IRC | 22:04 | |
| *** jheiss has joined #openstack | 22:04 | |
| *** vernhart has quit IRC | 22:09 | |
| *** gondoi has quit IRC | 22:09 | |
| *** MotoMilind has joined #openstack | 22:10 | |
| *** rnirmal has quit IRC | 22:10 | |
| *** jero has quit IRC | 22:11 | |
| *** pvo has quit IRC | 22:12 | |
| *** dw_jake has quit IRC | 22:12 | |
| *** nerens has quit IRC | 22:13 | |
| *** CloudChris has quit IRC | 22:14 | |
| *** CloudChris has joined #openstack | 22:17 | |
| *** dendro-afk is now known as dendrobates | 22:22 | |
| *** CloudChris has quit IRC | 22:24 | |
| *** CloudChris has joined #openstack | 22:24 | |
| *** Lantizia has quit IRC | 22:28 | |
| *** brd_from_italy has quit IRC | 22:30 | |
| *** enigma1 has left #openstack | 22:33 | |
| *** maplebed has quit IRC | 22:40 | |
| *** msassak has quit IRC | 22:52 | |
| *** ianw has quit IRC | 23:01 | |
| *** aliguori has quit IRC | 23:04 | |
| *** CloudChris has quit IRC | 23:10 | |
| *** MotoMilind has quit IRC | 23:10 | |
| *** dendrobates is now known as dendro-afk | 23:13 | |
| *** hggdh has quit IRC | 23:13 | |
| *** hggdh has joined #openstack | 23:14 | |
| *** nerens has joined #openstack | 23:20 | |
| *** sophiap has joined #openstack | 23:27 | |
| *** dendro-afk is now known as dendrobates | 23:32 | |
| *** aliguori has joined #openstack | 23:38 | |
| *** nerens has quit IRC | 23:39 | |
| *** joshuamckenty has joined #openstack | 23:39 | |
| mtaylor | devcamcar: hey there | 23:40 |
| joshuamckenty | sandywalsh: Gave a review of 52565 | 23:40 |
| devcamcar | mtaylor: hey hey! just following up with you on my pypi probs | 23:46 |
| *** joearnold has quit IRC | 23:47 | |
| mtaylor | devcamcar: yes! I think I'll solve that here real-soon-now | 23:48 |
| *** joearnold has joined #openstack | 23:48 | |
| devcamcar | mtaylor: awesome, thanks! | 23:48 |
| devcamcar | i looked again and it still eludes me | 23:49 |
| *** dragondm has quit IRC | 23:49 | |
| *** miclorb has joined #openstack | 23:50 | |
| *** CloudChris has joined #openstack | 23:50 | |
| *** nerens has joined #openstack | 23:52 | |
| *** ryker has quit IRC | 23:53 | |
| *** elasticdog has joined #openstack | 23:53 | |
| *** joshuamckenty has quit IRC | 23:54 | |
| *** pmullaney has quit IRC | 23:55 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
