| *** adiantum has joined #openstack | 00:02 | |
| *** rlucio has joined #openstack | 00:04 | |
| vishy | thief! | 00:04 |
|---|---|---|
| vishy | :) | 00:04 |
| vishy | JordanRinke: it is possible that the cache stuff should move all the way into glance, but I'm not sure if there are issues with how many images you can back to one cow2 image | 00:05 |
| vishy | s/cow2/base | 00:05 |
| rlucio | hey all... i've been a bit out of the loop lately, but now that i am back on HOL i am getting errors about config section headers for nova-api. does it use a special format now? the ppa repo doesn't seem to be changed so i have to imagine others have bumped into this | 00:05 |
| rlucio | by 'it' i mean the nova-api configuration file | 00:06 |
| *** jimbaker has quit IRC | 00:06 | |
| *** zul has quit IRC | 00:10 | |
| *** sandywalsh has quit IRC | 00:10 | |
| openstackhudson | Project nova build #460: SUCCESS in 1 min 31 sec: http://hudson.openstack.org/job/nova/460/ | 00:12 |
| openstackhudson | Tarmac: Fix for LP Bug #699654 | 00:12 |
| uvirtbot | Launchpad bug 699654 in nova "i18n - Terminating instance with invalid instance id causes error in the response" [Medium,Fix committed] https://launchpad.net/bugs/699654 | 00:12 |
| *** Tushar has quit IRC | 00:13 | |
| *** hazmat has quit IRC | 00:13 | |
| *** westmaas has joined #openstack | 00:13 | |
| *** hazmat has joined #openstack | 00:14 | |
| *** zul has joined #openstack | 00:15 | |
| *** MarkAtwood has quit IRC | 00:16 | |
| openstackhudson | Project swift build #179: SUCCESS in 30 sec: http://hudson.openstack.org/job/swift/179/ | 00:16 |
| openstackhudson | Tarmac: container-updater: temporary account update suppression on errors | 00:16 |
| *** rpedde has quit IRC | 00:16 | |
| JordanRinke | vishy: I would think you would run out of cpu/memory before the number of relations to a cow image would be an issue | 00:16 |
| *** blakeyeager has quit IRC | 00:17 | |
| vishy | nova-api.conf should be the same, but there was a point where it wasn't included in the packages | 00:17 |
| *** aimon has quit IRC | 00:18 | |
| *** kashyapc has quit IRC | 00:18 | |
| JordanRinke | vishy: unless using actual volume storage with some type of SAN.. at that point it is a dif game | 00:19 |
| *** kashyapc has joined #openstack | 00:19 | |
| *** stewart has quit IRC | 00:19 | |
| *** algorist has quit IRC | 00:21 | |
| *** pvo is now known as pvo_away | 00:22 | |
| *** aimon has joined #openstack | 00:24 | |
| rlucio | vishy: the nova-api.conf file I use just contains '--flagfile=/etc/nova/nova.conf' | 00:24 |
| rlucio | and i get this error: ConfigParser.MissingSectionHeaderError: File contains no section headers. | 00:24 |
| rlucio | file: /etc/nova/nova-api.conf, line: 1 | 00:24 |
| rlucio | '--flagfile=/etc/nova/nova.conf' | 00:24 |
| vishy | yes you need /etc/nova/nova-api.conf | 00:25 |
| vishy | it is the paste deploy config for nova api | 00:25 |
| *** troytoman has quit IRC | 00:27 | |
| rlucio | right... i don't see note about formatting though.. | 00:28 |
| rlucio | [default] | 00:28 |
| rlucio | mt | 00:29 |
| *** johnbergoon has quit IRC | 00:29 | |
| *** kashyapc has quit IRC | 00:30 | |
| *** zedas has joined #openstack | 00:32 | |
| *** JordanRinke is now known as JordanRinke_away | 00:32 | |
| *** westmaas1 has joined #openstack | 00:38 | |
| *** adiantum has quit IRC | 00:42 | |
| *** kashyapc has joined #openstack | 00:43 | |
| *** zul has quit IRC | 00:45 | |
| *** adiantum has joined #openstack | 00:46 | |
| *** MarkAtwood has joined #openstack | 00:46 | |
| rlucio | vishy: so does the nova-api process no longer need any reference to the gflags config options? | 00:51 |
| vishy | it still needs nova.conf | 00:51 |
| vishy | as well | 00:51 |
| *** zul has joined #openstack | 00:52 | |
| *** pvo_away is now known as pvo | 00:54 | |
| *** westmaas1 has quit IRC | 00:58 | |
| *** adiantum has quit IRC | 00:59 | |
| *** adiantum has joined #openstack | 01:03 | |
| *** londo_ has quit IRC | 01:14 | |
| rlucio | vishy: thanks for the info, i have things working now | 01:15 |
| *** Ryan_Lane is now known as Ryan_Lane|food | 01:17 | |
| *** jtran has quit IRC | 01:19 | |
| *** londo_ has joined #openstack | 01:20 | |
| openstackhudson | Project nova build #461: SUCCESS in 1 min 32 sec: http://hudson.openstack.org/job/nova/461/ | 01:22 |
| openstackhudson | Tarmac: Add DescribeInstanceTypes to admin api. This lets the dashboard know what sizes can be launched (using the -t flag in euca-run-instances, for example) and what resources they provide. | 01:22 |
| *** pre-war has joined #openstack | 01:29 | |
| *** prewarmoney has joined #openstack | 01:31 | |
| prewarmoney | A quick question about authentication for any devs out there... I've got two users; one a projectmanager with a bucket full of images, and the other some dude who wants to run the images | 01:34 |
| prewarmoney | I've added the second user to the first's project, and given him the developer role both globally and w/in the project | 01:34 |
| prewarmoney | I can euca-describe the images w/in the bucket, but recv a 401 when attempting to publish to the bucket or run any of the images | 01:35 |
| prewarmoney | has anyone here come across this before? | 01:36 |
| vishy | prewarmoney: that is odd | 01:36 |
| vishy | are the images registered? | 01:37 |
| vishy | usually you upload the image and then register | 01:37 |
| prewarmoney | ah that's probably what I'm forgetting | 01:37 |
| prewarmoney | thanks :) | 01:37 |
| vishy | and are you sure you are using the right credentials for the other user | 01:38 |
| vishy | meaning that he is accessing the proper project | 01:38 |
| *** rlucio has quit IRC | 01:38 | |
| prewarmoney | that I'm sure of | 01:38 |
| prewarmoney | vishy: no, that wasn't it | 01:40 |
| prewarmoney | are there any bucket-specific permissions I have to set | 01:41 |
| prewarmoney | running as the secondary user with 'euca-run-instances ami-410p7006 -k mykey -t m1.tiny' | 01:46 |
| *** pvo is now known as pvo_away | 01:47 | |
| prewarmoney | are there any other steps that have to be taken when running an image from a bucket you're not the projectmanager of? | 01:47 |
| prewarmoney | there is euca-modify-image-attribute, but it just seems to complain that ''user or group not specified | 01:50 |
| prewarmoney | euca-modify-image-attribute ami-410p7006 --launch-permission --add second_user | 01:50 |
| xtoddx | prewarmoney: i think by default the image is visible to your project, and you can use modify-image to set it to public, which means everyone can see it | 01:55 |
| *** londo_ has quit IRC | 01:58 | |
| prewarmoney | yeah, it's private at the moment, and the secondary user can see it with euca-describe... | 01:58 |
| *** stewart has joined #openstack | 01:58 | |
| prewarmoney | would he have to set it to public to run it, even though he's a member of the project? | 01:59 |
| xtoddx | prewarmoney: if it shows up in euca-describe-images, then it should launch | 01:59 |
| *** londo_ has joined #openstack | 02:00 | |
| *** hazmat has quit IRC | 02:01 | |
| prewarmoney | hmm. how would I go about tracing the authentication then? | 02:02 |
| *** GasbaKid has quit IRC | 02:04 | |
| prewarmoney | I'm seeing this in the nova-api logs: | 02:05 |
| prewarmoney | [SK8Y1F9KX063PKQ6CMKL USER PROJECT] Unauthorized request for controller=CloudController and action=RunInstances | 02:05 |
| prewarmoney | [SK8Y1F9KX063PKQ6CMKL USER PROJECT] 0.18225s 172.20.4.6 POST / None:None 401 [Boto/1.9b (linux2)] application/x-www-form-urlencoded text/html | 02:05 |
| *** miclorb_ has joined #openstack | 02:05 | |
| *** littleidea has quit IRC | 02:06 | |
| prewarmoney | that 'None:None' makes me think there'e something I missed when I set up the user | 02:06 |
| *** littleidea has joined #openstack | 02:12 | |
| *** stewart has quit IRC | 02:14 | |
| *** littleidea_ has joined #openstack | 02:17 | |
| *** hazmat has joined #openstack | 02:17 | |
| *** kpepple has joined #openstack | 02:19 | |
| kpepple | devcamcar: thx for the nova patch, it fixed my /projects/IRT/images error from earlier today. one question: what does "nebula images" mean on that page ? | 02:21 |
| *** littleidea has quit IRC | 02:21 | |
| *** littleidea_ is now known as littleidea | 02:21 | |
| prewarmoney | ok, here's how I am creating a new user and attaching it to an existing project: "nova-manage user create USER" -> "nova-manage project add PROJECT USER" -> "nova-manage role add USER developer" -> "nova-manage role add USER developer PROJECT" -> "nova-manage project zipfile PROJECT USER file.zip" -> "su - USER" -> source novarc and euca-add-keypair... -> done | 02:22 |
| prewarmoney | is this how it's done or am I missing a step | 02:22 |
| *** MarkAtwood has quit IRC | 02:25 | |
| prewarmoney | btw I'm running the stock maverick packages; should I be running the debs from ppa:nova-core/trunk instead? | 02:28 |
| *** adiantum has quit IRC | 02:29 | |
| *** johnbergoon has joined #openstack | 02:29 | |
| *** vvuksan has joined #openstack | 02:31 | |
| *** johnbergoon has left #openstack | 02:32 | |
| *** adiantum has joined #openstack | 02:35 | |
| *** burris has joined #openstack | 02:36 | |
| *** burris has quit IRC | 02:39 | |
| *** burris has joined #openstack | 02:40 | |
| *** burris has quit IRC | 02:41 | |
| kpepple | prewarmoney: the stock packages are Austin based (i think) … there are many, many bugfixes in the ppa | 02:45 |
| *** adiantum has quit IRC | 02:46 | |
| *** littleidea has quit IRC | 02:51 | |
| *** adiantum has joined #openstack | 02:59 | |
| *** kpepple has left #openstack | 03:03 | |
| *** maplebed has quit IRC | 03:05 | |
| *** syah has quit IRC | 03:13 | |
| *** adiantum has quit IRC | 03:15 | |
| *** nati has quit IRC | 03:15 | |
| *** opengeard has joined #openstack | 03:16 | |
| *** prewarmoney has quit IRC | 03:16 | |
| *** adiantum has joined #openstack | 03:21 | |
| *** burris has joined #openstack | 03:26 | |
| *** syah has joined #openstack | 03:32 | |
| vishy | prewarmoney: you don't really need to su to a different user. The stock packages are really old yes | 03:32 |
| *** burris has joined #openstack | 03:33 | |
| *** MarkAtwood has joined #openstack | 03:33 | |
| *** burris has quit IRC | 03:35 | |
| *** burris has joined #openstack | 03:36 | |
| *** Ryan_Lane|food is now known as Ryan_Lane | 03:40 | |
| *** burris has quit IRC | 03:41 | |
| *** j05h has quit IRC | 03:42 | |
| *** jimbaker has joined #openstack | 03:43 | |
| *** vvuksan has quit IRC | 03:45 | |
| *** algorist has joined #openstack | 03:45 | |
| *** burris has joined #openstack | 03:52 | |
| *** adiantum has quit IRC | 03:55 | |
| *** adiantum has joined #openstack | 03:56 | |
| *** burris has joined #openstack | 03:56 | |
| *** burris has quit IRC | 03:58 | |
| *** burris has joined #openstack | 03:58 | |
| *** burris has joined #openstack | 04:00 | |
| *** burris has joined #openstack | 04:02 | |
| *** burris has joined #openstack | 04:04 | |
| *** burris has quit IRC | 04:10 | |
| *** burris has joined #openstack | 04:11 | |
| *** miclorb_ has quit IRC | 04:12 | |
| *** MarkAtwood has quit IRC | 04:17 | |
| *** burris has quit IRC | 04:18 | |
| *** algorist has quit IRC | 04:18 | |
| *** MarkAtwood has joined #openstack | 04:20 | |
| *** burris has joined #openstack | 04:24 | |
| *** opengeard has quit IRC | 04:24 | |
| *** hadrian has quit IRC | 04:36 | |
| *** tomo_bot has quit IRC | 04:40 | |
| *** adiantum has quit IRC | 04:49 | |
| *** burris has quit IRC | 04:52 | |
| *** omidhdl has joined #openstack | 04:53 | |
| *** stewart has joined #openstack | 04:53 | |
| *** MarkAtwood has quit IRC | 04:55 | |
| *** jimbaker has quit IRC | 04:56 | |
| *** jimbaker has joined #openstack | 04:56 | |
| *** kpepple has joined #openstack | 05:00 | |
| *** adiantum has joined #openstack | 05:02 | |
| *** kpepple has left #openstack | 05:02 | |
| *** MarkAtwood has joined #openstack | 05:06 | |
| *** stewart has quit IRC | 05:12 | |
| *** adiantum has quit IRC | 05:20 | |
| *** MarkAtwood has quit IRC | 05:25 | |
| *** adiantum has joined #openstack | 05:25 | |
| *** f4m8_ is now known as f4m8 | 05:41 | |
| *** stewart has joined #openstack | 05:45 | |
| *** MarkAtwood has joined #openstack | 05:50 | |
| *** hazmat has quit IRC | 05:54 | |
| *** littleidea_ has joined #openstack | 06:03 | |
| *** adiantum has quit IRC | 06:04 | |
| *** adiantum has joined #openstack | 06:09 | |
| *** hazmat has joined #openstack | 06:10 | |
| *** adiantum has quit IRC | 06:20 | |
| *** adiantum has joined #openstack | 06:20 | |
| *** aimon has quit IRC | 06:24 | |
| *** stewart has quit IRC | 06:30 | |
| *** adiantum has quit IRC | 06:32 | |
| *** aimon_ has joined #openstack | 06:35 | |
| *** aimon_ has left #openstack | 06:36 | |
| *** adiantum has joined #openstack | 06:38 | |
| *** MarkAtwood has quit IRC | 06:39 | |
| *** littleidea_ has quit IRC | 06:59 | |
| *** aimon has joined #openstack | 07:06 | |
| *** j05h has joined #openstack | 07:11 | |
| *** aimon has quit IRC | 07:12 | |
| *** allsystemsarego has joined #openstack | 07:13 | |
| *** thimble has joined #openstack | 07:22 | |
| vishy | soren: I have a very odd networking problem if you happen to get up early. I could use some insight | 07:35 |
| *** adiantum has quit IRC | 07:45 | |
| *** adiantum has joined #openstack | 07:50 | |
| *** omidhdl has left #openstack | 07:57 | |
| ttx | wow 199 bugmail email in 10 hours | 08:00 |
| thimble | hello. I knew, I saw this some days ago but I can´t find it: OSError: [Errno 2] No such file or directory: '/usr/lib/python2.6/site-packages/nova/..//CA'. Any hint? This occurs when I want to create project zipfile. | 08:03 |
| ttx | thimble: see https://bugs.launchpad.net/nova/+bug/683137 | 08:04 |
| uvirtbot | Launchpad bug 683137 in nova "CA code doesn't get installed by "setup.py install"" [Medium,Triaged] | 08:04 |
| ttx | workaround in last comment | 08:04 |
| ttx | (just adapt the instructions to your deploy directory) | 08:04 |
| thimble | ttx: could be so easy. Thank you! I´ll give it a try. | 08:04 |
| ttx | (copy the CA dir from tarball to the plave where it looks for it) | 08:05 |
| ttx | place* | 08:05 |
| *** brd_from_italy has joined #openstack | 08:05 | |
| *** brd_from_italy has quit IRC | 08:09 | |
| soren | vishy: Sure, what's up? | 08:09 |
| *** brd_from_italy has joined #openstack | 08:09 | |
| vishy | soren: might take me a bit to explain | 08:10 |
| vishy | let me start with a pastie | 08:10 |
| *** brd_from_italy has quit IRC | 08:10 | |
| soren | vishy: I've got all day :) | 08:10 |
| vishy | soren: http://pastie.org/1498551 | 08:11 |
| vishy | so I'm trying to get nova working on two virtualbox vms in vlan mode | 08:11 |
| *** nati has joined #openstack | 08:11 | |
| vishy | there is some very weird stuff happening when an instance tries to hit metadata | 08:12 |
| vishy | dhcp seems to work fine | 08:12 |
| vishy | but the arp traffic does not make any sense to me | 08:12 |
| * soren looks | 08:12 | |
| vishy | it appears that the network host is attempting to make an arp who-has to a mac address instead of broadcast | 08:13 |
| vishy | which i didn't think was possible | 08:13 |
| vishy | the first set of arps gets a reply, but the second set doesn't...and the destination of the second set of requests is an actual mac address... | 08:14 |
| soren | Interesting. | 08:14 |
| vishy | i have no idea what the ipv6 reqs are either | 08:15 |
| vishy | so I'm trying to figure out if this is a nova problem or a multilayered vifrtualization + vlan problem | 08:15 |
| soren | Ok, looking at the kernel code now. | 08:18 |
| soren | Trying to work out the conditions under which it would send a who-has to a specific mac. | 08:18 |
| *** adiantum has quit IRC | 08:18 | |
| *** adiantum has joined #openstack | 08:24 | |
| *** omidhdl has joined #openstack | 08:25 | |
| soren | vishy: Hm... So far, all I can say is that it seems valid. | 08:27 |
| vishy | soren: so there is a situation where that can occur? | 08:27 |
| soren | vishy: I don't understand its purpose yet, but I'm wokring on it. | 08:27 |
| soren | vishy: Yeah. | 08:28 |
| soren | vishy: So it's not necessarily getting mangled somewhere along the way.. | 08:28 |
| vishy | soren: ok...so it could be something else that is eating the response then | 08:28 |
| vishy | i'm not totally sure why it needs to arp in the first place | 08:29 |
| vishy | it has the mac of that ip in the arp cache on the network host | 08:29 |
| vishy | (which i guess is obvious from the fact that it is sending the packet to that mac) | 08:29 |
| *** calavera has joined #openstack | 08:34 | |
| vishy | something that may help: if i set the default gateway of both the compute and network host to 10.0.0.1 (which is the ip that the network host has on br100, it works | 08:34 |
| *** adiantum has quit IRC | 08:34 | |
| soren | vishy: Ok, let's ignore the arp thing for a little bit. I think it's too early for me to be reading kernel networking code and RFC's. :) | 08:34 |
| vishy | but i don't see any arp packets, metadata just connects directly without issue | 08:34 |
| *** kashyapc has quit IRC | 08:35 | |
| vishy | soren: word. I've been getting smacked around by this for a few hours | 08:35 |
| vishy | soren: frustrating because it is so close to working | 08:35 |
| soren | So the problem is reaching the meta data service from a VM. You're saying this VM is running inside a VM itself? | 08:35 |
| vishy | soren: yes | 08:35 |
| soren | QEMu inside VirtualBox? | 08:36 |
| vishy | correct | 08:36 |
| vishy | works fine in single host mode | 08:36 |
| soren | Ok. | 08:36 |
| soren | Right, ok. | 08:36 |
| *** adiantum has joined #openstack | 08:36 | |
| soren | Which component is supposed to be doing the DNAT'ing? | 08:37 |
| soren | I forget. | 08:37 |
| vishy | network | 08:37 |
| soren | Ok. | 08:37 |
| soren | So the packets for 169.254.169.254 has to get there somehow. | 08:38 |
| soren | (Just talking to myself) | 08:38 |
| vishy | and the dnatting works from compute host if i give it a route | 08:38 |
| soren | A route? | 08:39 |
| vishy | for example route add -net 169.254.169.254 gw 10.0.0.1 | 08:39 |
| soren | I see. | 08:39 |
| vishy | but the default gateway for the vm should be 10.0.0.1, so the route shouldn't be necessary | 08:39 |
| soren | Can you pastebin the output of "ip addr list" on the compute node as well as in the guest? | 08:40 |
| soren | vishy: Depends. | 08:40 |
| vishy | the guest is hard to get into because i have to wait for metadata to fail | 08:40 |
| soren | vishy: Oh, and "ip route list", too. | 08:40 |
| soren | vishy: Oh, right. Crap. | 08:40 |
| soren | vishy: Well, a tcpdump fromthe compute node should do ok, I suppose. | 08:41 |
| *** thimble has quit IRC | 08:41 | |
| soren | vishy: tcpdump -i <interface of guest> | 08:41 |
| *** thimble has joined #openstack | 08:41 | |
| soren | vishy: "route add -net 169.254.169.254 gw 10.0.0.1" does actually make a difference compared to just leaving it to the default route. | 08:41 |
| soren | vishy: ...becuase it takes precedent over implicit routes (such as those that defined based on the addresses of your network interfaces). | 08:42 |
| *** Nacx has joined #openstack | 08:42 | |
| soren | "precedence"? | 08:42 |
| soren | Yeah, "precedence". | 08:42 |
| vishy | first pastie | 08:43 |
| vishy | http://pastie.org/1498622 | 08:43 |
| *** omidhdl has left #openstack | 08:44 | |
| *** fabiand_ has joined #openstack | 08:44 | |
| vishy | http://pastie.org/1498625 | 08:45 |
| soren | vishy: Could you include macs in that tcpdump, please? I can sort of guess what they're going to say based on the arp traffic, but I'd like to be sure. | 08:46 |
| vishy | looks like no ack coming back | 08:46 |
| vishy | http://pastie.org/1498626 | 08:48 |
| *** kashyapc has joined #openstack | 08:50 | |
| soren | vishy: And 10.0.0.1 is what? The network node? | 08:50 |
| vishy | yes | 08:50 |
| vishy | br100 on the network node | 08:50 |
| soren | Ok. | 08:52 |
| soren | And br100 (on the compute node) is made up of which interfaces? | 08:53 |
| vishy | vlan100 and vnet0 | 08:53 |
| soren | Ok. | 08:54 |
| soren | I assume "tcpdump -i any host 169.254.169.254" on the network host doesn't say anything at all? | 08:54 |
| vishy | i will try it (btw i just noticed the network host has a huge amount of traffic on lo | 08:55 |
| vishy | 163.3 MB, more than eth1 even, which seems a little odd | 08:55 |
| soren | Yeah, that does sound a bit out of the ordinary. :) | 08:56 |
| * vishy wonders if vlan tags are getting stripped off of some packets | 08:57 | |
| * soren has no clue about VirtualBox networking. | 08:57 | |
| soren | There be dragons, for all I know. | 08:57 |
| vishy | sure | 08:57 |
| vishy | :) | 08:57 |
| * vishy makes a kvm version of vagrant | 08:58 | |
| soren | I think I've solved the nbd lockup problem from last night. | 08:58 |
| soren | s/last night/~10 hours ago/g (for more TZ clarity) | 08:59 |
| vishy | had to run another instance because the first timed out | 08:59 |
| vishy | ah yes | 08:59 |
| vishy | and no i'm seeing nothing from that tcpdump | 08:59 |
| vishy | so it appears the packets are not getting there at all...hmmm | 09:01 |
| *** adiantum has quit IRC | 09:03 | |
| vishy | soren: if you have any more ideas of things to check, let me know | 09:08 |
| *** adiantum has joined #openstack | 09:08 | |
| *** kashyapc has quit IRC | 09:12 | |
| soren | vishy: If you could reproduce this on real hardware, I'd be more than happy to help you debug it. Right now, my inclination is to blame virtualbox. | 09:13 |
| vishy | fair enough | 09:13 |
| vishy | soren: I suppose i should at least isolate it to that | 09:13 |
| vishy | soren: 02:16:3e:17:9f:99 > 08:00:27:ba:c9:54, ARP, length 42: Reply 10.0.0.4 is-at 02:16:3e:17:9f:99, length 28 | 09:14 |
| vishy | arp reply! but it isn't making it out to the host so it must be taking the wrong route out | 09:15 |
| *** adiantum has quit IRC | 09:18 | |
| vishy | soren: there isn't anything in nwfilter rules that could be blocking the outgoing arp packet is there? | 09:19 |
| *** adiantum has joined #openstack | 09:22 | |
| *** irahgel has joined #openstack | 09:25 | |
| *** xEU has quit IRC | 09:26 | |
| soren | vishy: Sure. | 09:29 |
| soren | vishy: But how would nwfilter factor into this? | 09:29 |
| soren | vishy: It's virtualbox? | 09:29 |
| vishy | i meant on the compute host with firewall rules | 09:30 |
| soren | Hm.. | 09:31 |
| soren | It shouldn't, no. | 09:31 |
| soren | What you can do, though, is dump the output of iptables, ebtables with counters and everything. Wait a bit, and see which counters have increased. Maybe that can give you a hint about where traffic is going to. | 09:32 |
| *** adiantum has quit IRC | 09:33 | |
| *** adiantum has joined #openstack | 09:38 | |
| vishy | soren: i found a workaround, although i don't know why it makes a difference | 09:42 |
| vishy | soren: nor do i know how to make it work with multiple hosts | 09:43 |
| vishy | s/hosts projects | 09:43 |
| vishy | if i add this route on compute host: sudo route add 169.254.169.254 gw 10.0.0.1 it works | 09:43 |
| *** adiantum has quit IRC | 09:45 | |
| Adri2000 | devcamcar: hi, I'm trying dashboard trunk right now, and get a traceback when trying to access a project; want to see it? | 09:45 |
| vishy | soren: i suppose that is good enough for now, but i'd really like to understand why it needs that explicit route, and how i can make it work with all of the bridges | 09:52 |
| *** kashyapc has joined #openstack | 09:59 | |
| *** herki has quit IRC | 10:00 | |
| Adri2000 | reading about your network problems... I have some too :) so maybe you can help: I'm using nova.network.manager.VlanManager, and it seems that the compute node doesn't get correctly configured, at least the br100 interface doesn't get an ip address...any idea? | 10:00 |
| soren | vishy: That is odd indeed. | 10:00 |
| soren | vishy: Especially considering: | 10:01 |
| soren | 02:16:3e:43:ec:23 > 08:00:27:ba:c9:54, IPv4, length 74: 10.0.0.3.43470 > 169.254.169.254.80: tcp 0 | 10:01 |
| soren | 08:00:27:ba:c9:54 is 10.0.0.1. | 10:01 |
| soren | ...so it's already forwarding those packets to the gateway. | 10:01 |
| vishy | yeah but it doesn't seem to make it out over the vlan | 10:01 |
| soren | Ooohh. | 10:01 |
| vishy | without the route in they don't show up on vlan100 | 10:02 |
| soren | vishy: Can you show me "ip route list" now? | 10:02 |
| vishy | 169.254.169.254 via 10.0.0.1 dev br100 | 10:03 |
| vishy | only addition | 10:03 |
| soren | Can you show me the actual line? | 10:03 |
| soren | Is that it? | 10:04 |
| soren | I'm just surprised at the lack of scope (and potentially other stuff) | 10:05 |
| *** herki has joined #openstack | 10:05 | |
| vishy | yes that was the line | 10:05 |
| soren | vishy: Ok, just for giggles: | 10:06 |
| soren | Can you show me "ip route get 169.254.169.254" before and after you add the explicit route? | 10:07 |
| vishy | 169.254.169.254 via 10.0.2.2 dev eth0 src 10.0.2.15 | 10:09 |
| *** k0p has quit IRC | 10:09 | |
| vishy | oh whoops i have to add an ip on the bridge to actually get the route to add | 10:10 |
| vishy | 169.254.169.254 via 10.0.0.1 dev br100 src 10.0.0.29 | 10:12 |
| * ttx is testing a two-server setup on bare hardware, fwiw, will see if I hit strange metadata stuff | 10:12 | |
| vishy | cool ttx thanks | 10:12 |
| vishy | this works perfectly well without giving the compute host an ip and a special route in nebula | 10:12 |
| *** k0p has joined #openstack | 10:13 | |
| * ttx makes mental note, not using wifi networking between your two hosts speeds up the test | 10:13 | |
| vishy | somehow it seems as if the vlan isn't actually working properly | 10:13 |
| *** ddumitriu has joined #openstack | 10:14 | |
| soren | vishy: Oh.. | 10:14 |
| *** ddumitriu has quit IRC | 10:14 | |
| vishy | perhaps my setup of the vlan on the host that is running virtualbox is missing something, or perhaps the vbox nic drivers aren't handling vlan tagging properly | 10:14 |
| soren | I wonder if we need to have separate routing tables for guests. | 10:15 |
| soren | I can see why this fails now. | 10:15 |
| soren | The guest only knows about 10.0.0.1, but the packets for it land on the host and gets routed by the host. | 10:16 |
| soren | ...but the host wants to route differently than the guest. | 10:16 |
| vishy | soren: so how come it works in nebula? | 10:16 |
| ttx | vishy: maybe alien autofixing technology. | 10:16 |
| soren | vishy: I can't say. Maybe your default gateway on compute nodes actually end up either being what the guest expects, too? | 10:17 |
| soren | s/either// | 10:17 |
| soren | vishy: If you don't have an IP defined on br100... Does it work? | 10:17 |
| soren | I'm not sure how that works, tbh. | 10:18 |
| vishy | soren: no i tried that | 10:18 |
| vishy | soren in nebula, 169.254.169.254 happily routes to the network node | 10:18 |
| vishy | from the guests | 10:18 |
| vishy | even though the host routes over a different interface | 10:19 |
| ttx | I get some "Network is unreachable" errors from guest trying to reach metadata server, investigating | 10:19 |
| soren | ttx: You said they're connected via wifi? | 10:20 |
| vishy | ttx: you need a switch with vlans set up on real hardware | 10:20 |
| soren | ttx: briding and wifi => Lots of pain. It very often just doesn't work at all. | 10:21 |
| soren | vishy: I'm not sure why it works on Nebula. I'd have to investigate. When can I stop by? :) | 10:21 |
| ttx | vishy/soren: let me see if my switch supports that | 10:21 |
| vishy | soren: I'll run it by dean tomorrow, maybe he has some idea how it can be working | 10:22 |
| soren | vishy: I guess doing policy routing is reasonable. That way, we can even leave the host's "own" routing table completely alone. | 10:23 |
| vishy | soren: i do see these on the vnet interface STP 802.1w, Rapid STP, Flags [Learn, Forward], bridge-id 8c43.00:22:55:79:be:c1.808b, length 42 | 10:24 |
| * vishy has no idea how to do that | 10:25 | |
| soren | vishy: That's expected. | 10:25 |
| vishy | soren: i only see those in nebula though, not in the vbox version | 10:25 |
| soren | vishy: It's reasonably simple, really. | 10:25 |
| vishy | and i don't see the weird ipv6 packets in nebula either | 10:25 |
| vishy | regardless i really need to sleep | 10:26 |
| soren | vishy: Sure. Enjoy. | 10:27 |
| vishy | I'll try a couple of real hardware boxes tomorrow and see if i get the same issue | 10:27 |
| *** wedjat has joined #openstack | 10:39 | |
| *** nati has quit IRC | 10:48 | |
| ttx | ok, obviously my entry-line crappy Gb switches do not support VLAns ;) | 10:54 |
| ttx | though I could certainly abuse my DD-WRT setup to do so. | 10:57 |
| uvirtbot | New bug: #707908 in nova "Cannot restart nova-network under use_ipv6=true and with VlanManager" [Undecided,New] https://launchpad.net/bugs/707908 | 11:07 |
| *** aryan has quit IRC | 11:33 | |
| *** aryan has joined #openstack | 11:36 | |
| *** ddumitriu has joined #openstack | 11:44 | |
| soren | ttx: https://lkml.org/lkml/2011/1/26/131 <--- This is a big problem for natty. | 11:53 |
| *** arthurc has joined #openstack | 11:55 | |
| *** ddumitriu has quit IRC | 11:58 | |
| czajkowski | Announcing the ODFPlugfest UK http://www.odfplugfest.co.uk/ all communities and developers are welcome to take part, we have space for talks and it is a freeevent to take part in | 12:01 |
| *** fabiand_ has quit IRC | 12:04 | |
| *** TheHutch has joined #openstack | 12:17 | |
| *** reldan has joined #openstack | 12:18 | |
| *** ctennis has quit IRC | 12:25 | |
| *** adiantum has joined #openstack | 12:31 | |
| *** ctennis has joined #openstack | 12:43 | |
| *** ctennis has joined #openstack | 12:43 | |
| *** Dweezahr has joined #openstack | 12:47 | |
| *** westmaas has quit IRC | 12:49 | |
| *** aryan has quit IRC | 12:50 | |
| *** reldan has quit IRC | 12:52 | |
| *** aryan has joined #openstack | 12:53 | |
| Dweezahr | Hello! Next week thursday 3rd of February is the Bexar release, and this is sort of an Beta release as in not for production setup? Cactus is the stable release? | 13:03 |
| *** MarkAtwood has joined #openstack | 13:06 | |
| thimble | hello again. we´re trying to use openstack on sles11sp1 and are packaging. while running euca-describe-images the nova-api complains an unknown error. the last line in trace out is "<type 'exceptions.AttributeError'>: 'module' object has no attribute 'provider'", the line before is referring to boto. python-boto are we using in v2.03b. Any ideas? Thanks :) | 13:08 |
| *** ddumitriu has joined #openstack | 13:09 | |
| flashn | /win 21 | 13:09 |
| *** paulmillar has quit IRC | 13:13 | |
| *** ddumitriu has quit IRC | 13:33 | |
| ttx | thimble: you might have a boto that is too fresh. ISTR we are using 1.9 | 13:35 |
| thimble | ttx: we will try with 1.9. thanks | 13:36 |
| ttx | Dweezahr: it's a "featureful" release, so it might have rough edges | 13:36 |
| *** reldan has joined #openstack | 13:39 | |
| *** dprince has joined #openstack | 13:43 | |
| *** zul has quit IRC | 13:47 | |
| *** zul has joined #openstack | 13:48 | |
| soren | ttx: Ok, so dhcp requests are not reaching the network host. | 13:53 |
| soren | ttx: Let's start from the compute node and see how far they get. | 13:54 |
| *** zul_ has joined #openstack | 13:54 | |
| soren | ttx: Do you see them in "tcpdump -i vnet0 -vn" (assuming vnet0 is the tap device of the guest, which it probably is) | 13:54 |
| *** zul has quit IRC | 13:55 | |
| ttx | hhhm | 13:55 |
| ttx | I've no vnet0 on the compute node | 13:55 |
| soren | Do you have a vm? | 13:56 |
| soren | :) | 13:56 |
| soren | virsh list, find the name or id, then virsh dumpxml <id or name>. | 13:56 |
| soren | Try pastebining that. | 13:56 |
| ttx | I see BOOTP/DHCP stuff on that tcpdump alright | 13:57 |
| soren | So you found the right tap interface? | 13:57 |
| ttx | yes, was missig an instance | 13:57 |
| soren | Ok. | 13:57 |
| soren | Just to be sure, there's still nothing interesting on the network hosts' syslog? | 13:58 |
| ttx | nothing more than what you already got | 13:58 |
| *** laurensell has left #openstack | 13:58 | |
| soren | Ok. | 13:58 |
| soren | Ok, can you pastebin "brctl show"? | 13:58 |
| ttx | from the compute node ? | 13:59 |
| soren | Yes,please. | 13:59 |
| ttx | http://pastebin.com/tsvr92mH | 13:59 |
| soren | Yeah, that's not going to work. | 13:59 |
| soren | I'm rather embarassed to say so, but I'm not sure how this is supposed to work. | 14:00 |
| ttx | right, I narrowed it down to the bridges not talking to each other... but my lack of bridging foo started to stall me | 14:00 |
| *** hadrian has joined #openstack | 14:00 | |
| ttx | soren: for kicks I tried to specify a --flat_interface=eth0 but that just seemed to prevent bridge creation. | 14:01 |
| *** guynaor has joined #openstack | 14:02 | |
| *** reldan has quit IRC | 14:02 | |
| *** guynaor has left #openstack | 14:03 | |
| *** ppetraki has joined #openstack | 14:03 | |
| soren | vishy: The dhcp_start attribute on Networks is only used as an argument to dnsmasq's dhcp-range, correct? | 14:04 |
| ttx | soren: what were you expecting in "brctl show" output ? | 14:07 |
| *** vvuksan has joined #openstack | 14:08 | |
| *** iRTermite has quit IRC | 14:09 | |
| soren | ttx: If the VM is supposed to be able to get a dhcp response from the network node, they have to either be on the same broadcast domain or we need a dhcp relay agent. | 14:09 |
| soren | ttx: We don't have a dhcp relay agent. | 14:10 |
| *** mdomsch has joined #openstack | 14:10 | |
| soren | ttx: So they have to be in the same broadcast domain. That means that the bridge at least has to be connected to your physical network interface. | 14:10 |
| soren | From there, there might be more trickery on the network host. | 14:10 |
| soren | ttx: I'm pretty sure that once Vishy turns up again, he's going to something like Flat{,DHCP}Manager and a separate network host doesn't work or something. | 14:11 |
| ttx | soren: right, i suspect that's what --flat_interface is supposed to do | 14:11 |
| soren | ttx: Yeah. | 14:12 |
| ttx | I'll dig in that direction | 14:12 |
| soren | You're probably right. | 14:12 |
| ttx | but I failed to enable it succesfully so far. | 14:12 |
| soren | ttx: I was trying it earlier, but the subnet I have for VM's is not the same as my hosts's subnet. | 14:12 |
| soren | ...so I couldn't use it. | 14:12 |
| *** hggdh has quit IRC | 14:13 | |
| ttx | soren: Is it normal that this is full of virbr0 references: http://pastebin.com/w9Vvbw3r ? | 14:13 |
| ttx | I thought virbr0 wouldn't be used at all | 14:14 |
| ttx | same for 192.168.122.0 | 14:14 |
| soren | libvirt sets it up by default. It shouldn't interfere with anything. | 14:15 |
| ttx | ok | 14:15 |
| soren | You see them because I'm also (moreso, even) careful about not destroying existing rules. | 14:16 |
| *** iRTermite has joined #openstack | 14:16 | |
| *** hggdh has joined #openstack | 14:20 | |
| ttx | soren: so I don't think it can work without eth0 being added to the bridge, but if you do so it screws up networking | 14:21 |
| thimble | ttx: we downgraded to boto 1.9 now, we need the exakt version of euca2ools you use please. | 14:22 |
| ttx | boto 1.9b, euca2ools 1.2 | 14:23 |
| thimble | ttx: thank you | 14:23 |
| ttx | though in ubuntu euca2ools 1.2 was heavily patched. | 14:23 |
| thimble | please don´t say this :( | 14:23 |
| ttx | thimble: s/patched/fixed/ | 14:24 |
| *** reldan has joined #openstack | 14:24 | |
| ttx | i.e. we couldn't grow a dep on boto2 but still had to fix issues in boto1.9 | 14:24 |
| *** ddumitriu has joined #openstack | 14:26 | |
| soren | I'm taking off for the day. | 14:27 |
| *** gondoi has joined #openstack | 14:28 | |
| *** westmaas has joined #openstack | 14:28 | |
| *** alekibango has quit IRC | 14:32 | |
| *** alekibango has joined #openstack | 14:33 | |
| *** arthurc has quit IRC | 14:36 | |
| uvirtbot | New bug: #708025 in nova "VLAN field in networks table should be constrained" [Undecided,New] https://launchpad.net/bugs/708025 | 14:37 |
| *** arthurc has joined #openstack | 14:37 | |
| rcc | has glance sphinx dependency fix been merged ? i can't run tests right now :/ | 14:37 |
| ttx | rcc: if you mean https://code.launchpad.net/~mordred/nova/conditional-sphinx/+merge/47450, it wasn't yet | 14:39 |
| rcc | y, that one | 14:39 |
| *** aliguori has joined #openstack | 14:44 | |
| zul_ | ttx: you mean unbroken | 14:44 |
| *** zul_ is now known as zul | 14:44 | |
| *** zul has joined #openstack | 14:44 | |
| *** cynb has joined #openstack | 14:44 | |
| *** MarkAtwood has quit IRC | 14:46 | |
| *** jaypipes has joined #openstack | 14:48 | |
| *** ddumitriu has quit IRC | 14:50 | |
| rcc | :( , my pep8 'issue' will have to wait then | 14:53 |
| rcc | btw, whoever wrote that openstack dashboard, it's pretty awesome, kudos | 14:53 |
| *** f4m8 is now known as f4m8_ | 14:54 | |
| *** littleidea has joined #openstack | 14:54 | |
| ianweller | rackerhacker: ping :) | 14:55 |
| *** olivier_ has quit IRC | 14:57 | |
| *** olivier_ has joined #openstack | 14:57 | |
| jaypipes | mtaylor: see https://code.launchpad.net/~mordred/glance/conditional-sphinx/+merge/47452... | 14:58 |
| *** littleidea has quit IRC | 14:58 | |
| *** sandywalsh has joined #openstack | 15:01 | |
| *** MarcMorata has joined #openstack | 15:03 | |
| *** dendrobates is now known as dendro-afk | 15:03 | |
| *** dendro-afk is now known as dendrobates | 15:05 | |
| rackerhacker | ianweller: ttl exceeded ;) | 15:05 |
| rackerhacker | ianweller: what's up? | 15:05 |
| ianweller | rackerhacker: nah, i set the ttl for 9001 | 15:06 |
| ianweller | rackerhacker: pm? | 15:06 |
| rackerhacker | ianweller: sure ;) | 15:06 |
| sandywalsh | Another core have a sec to approve https://code.launchpad.net/~sandy-walsh/nova/lp705082/+merge/46895 | 15:09 |
| ttx | if someone can explain to me how traffic is supposed to flow from compute nodes to network node in FlatDHCPManager mode, I'll pay in beer. | 15:12 |
| *** maplebed has joined #openstack | 15:12 | |
| *** vvuksan has quit IRC | 15:19 | |
| ttx | Reviews needed on https://code.launchpad.net/~ntt-pf-lab/nova/lp703037/+merge/47028 <-- contains several needed ipv6 bugfixes | 15:20 |
| henrichrubin | i am using the latest nova branch w/ mysql and 'nova.sh run.' i am getting this error http://paste.org/pastebin/view/27975 on euca-run-instances, even though the users table was created. any idea? | 15:29 |
| henrichrubin | ? | 15:29 |
| ttx | rcc: you can test by running pep8 nova/volume/api.py | 15:31 |
| ttx | just removing the extra line works for me. | 15:32 |
| rcc | y, but you know...i wanted to run the tests hudson runs so I don't fall into the same problem with yet another test :p . just trying to cover everything to not bother you again with reviews... | 15:33 |
| *** dirakx has joined #openstack | 15:36 | |
| *** westmaas has quit IRC | 15:41 | |
| *** pharkmillups has joined #openstack | 15:42 | |
| *** aliguori has quit IRC | 15:43 | |
| *** littleidea has joined #openstack | 15:44 | |
| flashn | lslsls | 15:46 |
| flashn | ¨¨ | 15:46 |
| flashn | lslsls | 15:46 |
| *** aliguori has joined #openstack | 15:46 | |
| *** daleolds has joined #openstack | 15:47 | |
| *** thimble has quit IRC | 15:50 | |
| *** maplebed has quit IRC | 15:51 | |
| *** DigitalFlux has joined #openstack | 15:52 | |
| *** DigitalFlux has joined #openstack | 15:53 | |
| *** DigitalFlux has joined #openstack | 15:53 | |
| *** dragondm has joined #openstack | 15:53 | |
| *** DigitalFlux has quit IRC | 15:53 | |
| *** daleolds has quit IRC | 15:54 | |
| *** thimble has joined #openstack | 15:54 | |
| *** manasa has joined #openstack | 15:56 | |
| *** westmaas has joined #openstack | 15:57 | |
| *** mray has joined #openstack | 15:59 | |
| *** littleidea_ has joined #openstack | 16:00 | |
| *** littleidea has quit IRC | 16:04 | |
| *** littleidea_ is now known as littleidea | 16:04 | |
| *** thimble has quit IRC | 16:08 | |
| *** arthurc has quit IRC | 16:09 | |
| *** arthurc has joined #openstack | 16:09 | |
| openstackhudson | Project nova build #462: SUCCESS in 1 min 30 sec: http://hudson.openstack.org/job/nova/462/ | 16:12 |
| openstackhudson | Tarmac: Simple little changes related to openstack api to work better with glance. | 16:12 |
| openstackhudson | Fixes ImageID missing from Glance and int->string id problem. | 16:12 |
| Ryan_Lane | ttx: traffic flows via DNAT and SNAT | 16:12 |
| Ryan_Lane | ttx: the network node is the gateway for the instances | 16:13 |
| Ryan_Lane | ttx: it gets an IP and the gateway via DHCP from the network node | 16:13 |
| Ryan_Lane | ttx: when floating IPs are assigned to the instances, the network node sets the SNAT and DNAT rules | 16:13 |
| Ryan_Lane | so the network node needs ip forwarding enabled | 16:13 |
| *** troytoman has joined #openstack | 16:14 | |
| ttx | Ryan_Lane: I'm blocked in my test: the DHCP requests from the instance never reach the network node. | 16:14 |
| Ryan_Lane | ah. I had this problem the other day | 16:14 |
| Ryan_Lane | there's a workaround | 16:14 |
| ttx | Ryan_Lane: how does that DHCP request reach the network node ? | 16:15 |
| Ryan_Lane | lemme see if I can find it | 16:15 |
| Ryan_Lane | there's a bug... | 16:15 |
| ttx | I'm confused with how the bridge setup is supposed to make the two communicate with each other. | 16:15 |
| Ryan_Lane | ah | 16:15 |
| Ryan_Lane | so in flat mode, you make the bridge | 16:16 |
| Ryan_Lane | in flat dhcp mode, you do not | 16:16 |
| Ryan_Lane | you give it a device, and it makes the bridge | 16:16 |
| Ryan_Lane | on both the network node and the compute nodes | 16:16 |
| ttx | Ryan_Lane: precise "you give it a device" | 16:16 |
| Ryan_Lane | ttx: https://answers.launchpad.net/nova/+question/141960 | 16:16 |
| ttx | I think that's the part I'm missing, my bridges don't have interfaces linked | 16:17 |
| Ryan_Lane | that's the bug fix I'm mentioning, in case you missed it | 16:17 |
| Ryan_Lane | gimme a sec though | 16:17 |
| Ryan_Lane | I'll show you my config | 16:17 |
| Ryan_Lane | ttx: http://pastebin.com/VrvbD3ba | 16:18 |
| Ryan_Lane | I'm using that on both compute nodes and network nodes | 16:18 |
| Ryan_Lane | a few of those options are likely unnecessary | 16:19 |
| ttx | Ryan_Lane: you have multiple ethX on the machine you use ? | 16:19 |
| Ryan_Lane | yes | 16:19 |
| Ryan_Lane | I'm using ESXi, so I just add one | 16:20 |
| Ryan_Lane | that'll be more difficult when I switch to nova :) | 16:20 |
| ttx | Ryan_Lane: ah. So the setup supposes two interfaces on each node, connected to each other ? | 16:20 |
| Ryan_Lane | I think you can do it with a single interface too | 16:21 |
| Ryan_Lane | I'd likely set the public_interface to the bridge that'll be created | 16:21 |
| ttx | I only have eth0 and when I use --flat_interface=eth0 it kinda breaks | 16:21 |
| ttx | (maybe an extra config step is required to unbreak the "regular" network) | 16:21 |
| Ryan_Lane | I wonder if it would work with virtual devices | 16:22 |
| ttx | Ryan_Lane: my nodes are connected through their eth0, with IPs in 192.168.0.0/24. When I specify --flat_interface=eth0, it assigns eth0 to the bridge and I lose networking between the nodes | 16:22 |
| Ryan_Lane | meaning, make a virtual nic from eth0 (like eth0:1), make that the public interface | 16:23 |
| Ryan_Lane | then use eth0 for flat_interface, and eth0:1 for the public one | 16:23 |
| ttx | If I don't do --flat_interface, instance traffic just doesn't get from one to the other | 16:23 |
| sandywalsh | ttx jaypipes sirp- perhaps another quick approve on https://code.launchpad.net/~sandy-walsh/nova/lp705082/+merge/47548 ... there was a line that got lost on a trunk conflict (sorry) | 16:24 |
| Ryan_Lane | yeah, the flat interface is added to a bridge, that bridge is used for the virtual nics | 16:24 |
| Ryan_Lane | the instance's virtual nics that is | 16:25 |
| ttx | Ryan_Lane: ok, then it looks like we are missing some network topology docs. I was under the impression that I could test multiple nodes using a single interface | 16:25 |
| *** rpedde has joined #openstack | 16:25 | |
| Ryan_Lane | you can | 16:25 |
| Ryan_Lane | I just don't know how :) | 16:25 |
| Ryan_Lane | though I'd have to imagine that using nic aliases (like eth0:1) would work | 16:26 |
| ttx | and VlanManager needs sharper switches than I have :) | 16:26 |
| Ryan_Lane | vlanmanager is a little too much for my tastes :) | 16:26 |
| ttx | -ENOTENOUGHHARDWARE fail | 16:26 |
| Ryan_Lane | heh | 16:26 |
| Ryan_Lane | getting my users to use a vpn would not fly. hell, getting my coworkers to use a vpn wouldn't fly | 16:27 |
| ttx | sandywalsh: looks like most of your patch disappeared ? | 16:27 |
| ttx | sandywalsh: or my brain was fried by too much testing ? | 16:28 |
| sandywalsh | ttx the main part was merged with trunk ... this is just the missing part | 16:28 |
| ttx | sandywalsh: oh! | 16:28 |
| ttx | sandywalsh: go for it | 16:29 |
| ttx | (can't flip the switch, not nova-core) | 16:29 |
| sandywalsh | ttx thx | 16:29 |
| * annegentle makes it back to the land of the living | 16:32 | |
| jaypipes | annegentle: welcome back :) | 16:32 |
| ttx | annegentle: how was the other side ? | 16:32 |
| annegentle | it's early for the Californians, but does anyone know if django-nova should be local to the database server or local to the cloud controller when you install it? | 16:33 |
| annegentle | thanks - it was foggy :) | 16:33 |
| ttx | annegentle: dunno, but I'd be very surprised if it needed to be local | 16:33 |
| *** kashyapc has quit IRC | 16:33 | |
| annegentle | there's a lotta mention of "local/locally" in the readme at http://bazaar.launchpad.net/~devcamcar/openstack-dashboard/trunk/view/head:/README - just not sure what the context is. Local to? | 16:34 |
| mtaylor | jaypipes: I signed the thing a while ago but can't find the email with the results of that | 16:37 |
| mtaylor | jaypipes, ttx, dendrobates, soren: anybody know if there is a master list in which I can look for my signature id? | 16:38 |
| *** pvo_away is now known as pvo | 16:38 | |
| dendrobates | mtaylor: sorry, what signature id? | 16:39 |
| mtaylor | dendrobates: for signing the CLA | 16:39 |
| ttx | mtaylor: you mean the echosign logs ? | 16:39 |
| mtaylor | dendrobates: I'm not in our CLA list on the wiki, but I did sign the fool thing | 16:39 |
| mtaylor | ttx: yeah | 16:39 |
| ttx | dendrobates, jbryce: ^ | 16:39 |
| jbryce | mtaylor: i think i might be able to find you | 16:40 |
| mtaylor | jbryce: you are awesome | 16:40 |
| *** vvuksan has joined #openstack | 16:41 | |
| openstackhudson | Project nova build #463: SUCCESS in 1 min 28 sec: http://hudson.openstack.org/job/nova/463/ | 16:42 |
| openstackhudson | Tarmac: Simple little changes related to openstack api to work better with glance. | 16:42 |
| openstackhudson | Fixes ImageID missing from Glance and int->string id problem. | 16:42 |
| jbryce | mtaylor: https://secure.echosign.com/public/audit?tx=XBGUYJI59M5Y6G - that think at the end is your echosign id | 16:42 |
| colinnich | Got an odd problem with the swift stats system. The final processed files are in a strange format that most things can't read. Seems to be UTF-32, but with no headers. Haven't really got any experience of working with different encodings, so not sure if I'm doing something silly. The only way I can read the files with all tools is by running iconv -f utf32 -t ascii 75ea619fc096c906407914728e3a2dc9.csv | 16:44 |
| colinnich | 'cat' displays them, but pretty much everything else fails | 16:44 |
| creiht | notmyname: -^ ? | 16:44 |
| mtaylor | jbryce: THANKS! | 16:46 |
| dendrobates | mtaylor: you should have received an email from echosign | 16:46 |
| mtaylor | dendrobates: yeah - but that was forever ago and I don't seem to have it anymore :) | 16:46 |
| dendrobates | mtaylor: if you can't find it I can look for it. | 16:46 |
| mtaylor | dendrobates: jbryce found it for me | 16:46 |
| *** zenmatt has quit IRC | 16:47 | |
| *** GasbaKid has joined #openstack | 16:48 | |
| *** zenmatt has joined #openstack | 16:49 | |
| notmyname | colinnich: utf-8 is good, so utf-32 must be 4 times better, right? | 16:50 |
| *** calavera has quit IRC | 16:51 | |
| *** kashyapc has joined #openstack | 16:52 | |
| creiht | colinnich: have you checked to see if the raw logs are in the same format? | 16:52 |
| colinnich | creiht, notmyname: raw logs (/var/log/swift/proxy.log anyway) are fine. more and vi open them fine | 16:54 |
| colinnich | I'm processing them on my ubuntu desktop, but I tried downloading them (with st) on the proxy server and it was the same issue | 16:55 |
| *** cynb has quit IRC | 16:55 | |
| *** jcannava has joined #openstack | 16:57 | |
| notmyname | anything odd like the locale on the machine? | 16:57 |
| *** kbringard has joined #openstack | 16:57 | |
| colinnich | notmyname: how do you check (lucid)? | 16:57 |
| notmyname | "locale" | 16:58 |
| colinnich | notmyname: looks ok - http://paste.openstack.org/show/571/ | 16:59 |
| jk0 | jaypipes: have a sec to look at https://code.launchpad.net/~jk0/glance/pep8/+merge/47556 ? | 16:59 |
| colinnich | notmyname: my laptop is the same, but en_GB | 16:59 |
| notmyname | colinnich: the final uploaded csv should end with ".csv.gz". you mentioned ".csv". did you decompress the file? | 17:00 |
| colinnich | notmyname: yeah, I'm decompressing it with gunzip | 17:00 |
| *** jdurgin has joined #openstack | 17:01 | |
| jaypipes | jk0: looking.. | 17:02 |
| notmyname | colinnich: odd. a quick run on my SAIO and everything worked | 17:04 |
| colinnich | notmyname: just checked my old cluster which is running Austin (I think!) and it's the same | 17:07 |
| colinnich | notmyname: haven't noticed before as I only ever opened it in OpenOffice which works fine. Now I'm trying to access them programatically... | 17:07 |
| notmyname | colinnich: do you see only old csvs do this or even ones created now? | 17:08 |
| creiht | colinnich: what does it say if you run | 17:09 |
| * ttx calls it a day | 17:09 | |
| creiht | file LOGFILENAME | 17:09 |
| colinnich | creiht: b42164667b281f9a3ed7b73f7b887f68.csv: data | 17:11 |
| creiht | hrm | 17:11 |
| * creiht is stumped :/ | 17:12 | |
| colinnich | notmyname: the one I was looking at was from this morning | 17:13 |
| creiht | colinnich: Do the logs have sensitive data in them? | 17:13 |
| colinnich | creiht: nope | 17:13 |
| creiht | colinnich: could you upload one somewhere where we could download and take a look at? | 17:14 |
| *** dubsquared has joined #openstack | 17:14 | |
| colinnich | creiht: http://downloads.netintelligence.com/temp/swiftlog.csv.gz | 17:16 |
| creiht | well that is interesting | 17:17 |
| vvuksan | kbringard: ping | 17:23 |
| kbringard | vvuksan: ahoy | 17:24 |
| vvuksan | i have made progress | 17:24 |
| kbringard | vvuksan: I'm stuck in dependency hell | 17:24 |
| vvuksan | compiled 0.8.7 libvirtd | 17:24 |
| vishy | ttx, soren: sent an email to openstack regarding flatdhcp | 17:24 |
| *** TheHutch has quit IRC | 17:25 | |
| notmyname | colinnich: the csv is created by putting the data into a cStringIO.StringIO object, then compressed as it is sent to the swift cluster | 17:26 |
| vvuksan | kbringard: let me upload the RPMS | 17:26 |
| vvuksan | kbringard: http://vuksan.com/centos/nova/x86_64/ | 17:26 |
| colinnich | notmyname: yeah, I had a look at that earlier to see if I could see anything | 17:26 |
| kbringard | wow, that's a lot of compiling | 17:27 |
| vvuksan | kbringard: downloaded RHEL6 libvirtd RPMS, then downloaded 0.8.7 | 17:27 |
| *** maple_bed has joined #openstack | 17:28 | |
| vvuksan | kbringard: had to tweak the configure to use python26 and couple other changes but it was straight forward after that | 17:28 |
| vvuksan | now i need to debug this | 17:28 |
| vvuksan | (nova.exception): TRACE: Command: qemu-img create -f qcow2 -o cluster_size=2M,backing_file=/opt/nova-state/instances/_base/ami-hvxr8c1j_sm /opt/nova-state/instances/instance-00000009/disk | 17:28 |
| vvuksan | (nova.exception): TRACE: Stderr: 'qemu-img: invalid option -- o\n' | 17:28 |
| vvuksan | but things are going in the right direction :-) | 17:28 |
| *** manasa has quit IRC | 17:29 | |
| kbringard | indeed so, good work sir | 17:29 |
| kbringard | vvuksan: did you remove libvirt 0.6.3? | 17:33 |
| vvuksan | yes | 17:33 |
| vvuksan | removed all the remanants | 17:33 |
| vvuksan | even virt-manager and all the other junkj | 17:33 |
| *** daleolds has joined #openstack | 17:34 | |
| *** pharkmillups has quit IRC | 17:38 | |
| vvuksan | this is just great | 17:39 |
| kbringard | vvuksan: yea, I don't see the -o option in qemu-img on cent, but it's there in the one on debian | 17:39 |
| vvuksan | looks like libvirtd 0.8.7 has yanked -o from qemu-img | 17:39 |
| kbringard | ubuntu rather | 17:39 |
| vvuksan | yeah | 17:39 |
| vvuksan | noticed it myself | 17:39 |
| kbringard | qemu-img-0.10.5-1.el5.2 is what I have installed | 17:41 |
| vvuksan | good point | 17:41 |
| kbringard | on ubuntu it's not a separate package, but qemu is version 0.12 | 17:41 |
| kbringard | I'd wager that's the issue | 17:41 |
| vvuksan | # rpm -qf /usr/bin/qemu-img | 17:41 |
| vvuksan | kvm-qemu-img-83-164.0.1.el5_5.25 | 17:41 |
| vvuksan | hmmm | 17:41 |
| kbringard | yea, I was just noticing that | 17:42 |
| vvuksan | time to download http://ftp.redhat.com/redhat/rhel/beta/6.0/source/SRPMS/qemu-kvm-0.12.1.2-2.17.el6.src.rpm | 17:43 |
| kbringard | indeed :-/ I tried using the kvm-qemu-img instead and it still doesn't support it | 17:45 |
| *** Nacx has quit IRC | 17:48 | |
| *** kpepple has joined #openstack | 17:50 | |
| *** GasbaKid has quit IRC | 18:11 | |
| *** kashyapc has quit IRC | 18:14 | |
| dubsquared | http://paste.openstack.org/show/574/ — trying to use sqllite but i want to use mysql? | 18:14 |
| vvuksan | dubsquared: did you do db sync ? | 18:16 |
| dubsquared | yeah | 18:17 |
| vvuksan | did it give you output :-) ? | 18:17 |
| dubsquared | yeah :D | 18:17 |
| vvuksan | mm | 18:17 |
| vvuksan | try rerunning db sync with flagfile | 18:18 |
| dubsquared | neg | 18:19 |
| dubsquared | my mysql tables are populated... | 18:20 |
| dubsquared | with network, fixedips, etc | 18:20 |
| vvuksan | when do you then get the error ? | 18:20 |
| *** MarcMorata has quit IRC | 18:21 | |
| vvuksan | dubsquared: does the networks table exist ? | 18:21 |
| *** MarcMorata has joined #openstack | 18:21 | |
| dubsquared | ja | 18:21 |
| dubsquared | nova-network is complaining | 18:21 |
| *** vvuksan has quit IRC | 18:23 | |
| *** Ryan_Lane is now known as Ryan_Lane|food | 18:24 | |
| *** MarcMorata has quit IRC | 18:26 | |
| dubsquared | hmm ok got that figured out, now getting another error from api | 18:26 |
| dubsquared | http://paste.openstack.org/show/575/ | 18:27 |
| dubsquared | felt like i got through all these long ago, now they are all popping up again heh | 18:27 |
| dubsquared | api is running and listeing | 18:28 |
| dubsquared | novarc is sourced | 18:28 |
| dubsquared | all info in it is correc5t | 18:28 |
| kpepple | dubsquared: what did the nova-api log say ? | 18:28 |
| kbringard | are you sure the novarc you sourced has the correct API key? | 18:28 |
| kpepple | dubsquared: looks like a key problem | 18:28 |
| kbringard | that happens to me sometimes... I create new projects and source the wrong rc | 18:28 |
| dubsquared | 2011-01-26 12:23:03,716 AUDIT nova.api [-] Authentication Failure: No project with id dubproj | 18:29 |
| dubsquared | odd.. | 18:29 |
| dubsquared | i surely made that project | 18:29 |
| dubsquared | lemme try again | 18:29 |
| dubsquared | :facepalm: | 18:31 |
| kbringard | hehe | 18:31 |
| dubsquared | dubproj != dubproject | 18:31 |
| dubsquared | :| | 18:32 |
| dubsquared | :D | 18:32 |
| dubsquared | port 22 is now allowed by default? | 18:33 |
| *** vvuksan has joined #openstack | 18:36 | |
| *** mdomsch has quit IRC | 18:36 | |
| vvuksan | dubsquared: did you find the problem? | 18:38 |
| kbringard | vvuksan: I'm grabbing the qemu source, compiling it, and then I'm going to try to drop in replace qemu-img... | 18:41 |
| kbringard | ballsy... stupid... but ballsy | 18:42 |
| *** retr0h has joined #openstack | 18:42 | |
| * kbringard slaps retr0h around a bit with a large trout | 18:43 | |
| vvuksan | yeah | 18:44 |
| *** bcwaldon has quit IRC | 18:48 | |
| *** arthurc has quit IRC | 18:53 | |
| *** mdomsch has joined #openstack | 18:56 | |
| *** bcwaldon has joined #openstack | 19:02 | |
| *** dendrobates is now known as dendro-afk | 19:02 | |
| *** dendro-afk is now known as dendrobates | 19:03 | |
| *** Ryan_Lane|food is now known as Ryan_Lane | 19:05 | |
| *** nelson has quit IRC | 19:05 | |
| *** pvo is now known as pvo_away | 19:06 | |
| *** nelson has joined #openstack | 19:06 | |
| *** irahgel has quit IRC | 19:06 | |
| *** rlucio has joined #openstack | 19:07 | |
| *** MarkAtwood has joined #openstack | 19:07 | |
| kbringard | libvirtError: internal error no supported architecture for os type 'hvm' | 19:11 |
| kbringard | anyone seen this error before? | 19:12 |
| kpepple | kbringard: are you using Xen ? | 19:12 |
| kbringard | nope, should be using KVM | 19:12 |
| kbringard | the internets seem to think it means hardware acceleration for VMs is disabled in the bios | 19:13 |
| kbringard | which I suppose it could be, but I've run other VMs on this machine before with no problems | 19:13 |
| ttx | vishy: very enlightening, many thanks | 19:14 |
| kbringard | is there a way to disable using hvm? | 19:16 |
| vvuksan | kbringard: i think this has to do with libvirt set up | 19:17 |
| *** maple_bed is now known as maplebed | 19:17 | |
| kpepple | kbringard: ubuntu or centos ? | 19:17 |
| kbringard | kpepple: cent | 19:17 |
| kbringard | it works on ubuntu :-) | 19:17 |
| ttx | kbringard: that's the error you get when your compute node doesn't support kvm | 19:18 |
| ttx | kbringard: what os are you on ? | 19:18 |
| kbringard | ttc centos 5 | 19:18 |
| kbringard | err | 19:18 |
| kbringard | ttx* | 19:18 |
| kpepple | kbringard: yeah, that's probably libvirt shennigans … try seeing if you have kvm loaded | 19:18 |
| ttx | make sure you have VT exensions and they are enabled in BIOS | 19:19 |
| ttx | (should see that in dmesg kernel output) | 19:19 |
| ttx | on Ubuntu kvm-ok does that for you. | 19:20 |
| kpepple | kbringard: check you cpu at cat /proc/cpuinfo and look for vt or vmx in the flags | 19:21 |
| kbringard | I don't see anything about KVM in dmesg, but I do have vmx in my cpuinfo | 19:21 |
| *** littleidea has quit IRC | 19:21 | |
| kpepple | try to # modprobe kvm | 19:21 |
| kbringard | from what I read that doesn't necessarily mean it's enabled, but that I should see something about the bios disabling KVM in dmesg if it was disabled | 19:21 |
| kbringard | lsmod shows kvm loaded, but I did a modprobe on it again | 19:22 |
| vvuksan | # dmesg | grep kvm | 19:22 |
| kbringard | vvuksan: nothing | 19:22 |
| *** pvo_away is now known as pvo | 19:23 | |
| vvuksan | mine says disabled by bios | 19:23 |
| vvuksan | so that sounds good :-) | 19:23 |
| kbringard | I'm going to try rebooting to see if I can catch some KVM stuff... | 19:24 |
| kbringard | now I see some kvm stuff in dmesg | 19:27 |
| kbringard | kvm: virtualization flags detected on this hardware: vmx tpr_shadow vnmi flexpriority | 19:27 |
| kbringard | oaded kvm module (kvm-83-164.el5_5.25) | 19:27 |
| kbringard | so lets try this again... | 19:27 |
| *** jtran has joined #openstack | 19:29 | |
| jtran | hey all. any guide on the API-Server rest API? | 19:30 |
| *** westmaas has quit IRC | 19:30 | |
| kpepple | vvuksan: OT question - are you using stock centos libvirt (0.6.3) or did you recompile to get to 0.8.7 ? | 19:30 |
| kbringard | kpepple: recompiled | 19:30 |
| vvuksan | kpepple: I built RPMS | 19:31 |
| kbringard | 0.6.3 segfaults | 19:31 |
| vvuksan | http://vuksan.com/centos/nova/ | 19:31 |
| kbringard | or well, causes nova-compute to segfault | 19:31 |
| kpepple | vvuksan: thx | 19:31 |
| *** imsplitbit has joined #openstack | 19:33 | |
| kbringard | ok, so rebooting seems to have resolved that hvm issue... weird | 19:34 |
| vvuksan | and ... | 19:37 |
| vvuksan | don't hold us in suspense :-) | 19:37 |
| kbringard | oh, sorry | 19:37 |
| kbringard | getting a new qemu error | 19:38 |
| vvuksan | k | 19:38 |
| kbringard | libvir: QEMU error : internal error Process exited while reading console log output: qemu: could not open disk image /var/lib/nova/instances/instance-00000011/disk | 19:38 |
| kbringard | but the disk is there | 19:38 |
| *** joearnold has joined #openstack | 19:38 | |
| kbringard | so I am wondering if I boned something when I upgraded qemu | 19:38 |
| vvuksan | kpepple: got qemu-img RPMS ? | 19:39 |
| kpepple | vvuksan: not a new one … | 19:41 |
| kbringard | vvuksan: rpm -qf /usr/bin/qemu-img --> qemu-0.12.4-1.el5.rf | 19:41 |
| vvuksan | is this fromepel ? | 19:41 |
| kbringard | hmmm, maybe not | 19:42 |
| kbringard | one moment | 19:42 |
| kbringard | actually, not likely... because of the .rf | 19:42 |
| kbringard | I didn't yum install it, I grabbed the RPM directly from somehwee | 19:42 |
| kbringard | somwhere* | 19:42 |
| kbringard | rpmforge | 19:44 |
| kbringard | http://apt.sw.be/redhat/el5/en/x86_64/rpmforge/RPMS/qemu-0.12.4-1.el5.rf.x86_64.rpm | 19:44 |
| *** kpepple has quit IRC | 19:45 | |
| *** hggdh has quit IRC | 19:45 | |
| vishy | kbringard: cold be app armor | 19:46 |
| *** reldan has quit IRC | 19:47 | |
| vishy | s/cold/could | 19:47 |
| vvuksan | selinux | 19:47 |
| vishy | there are specific exceptions in the debian packaging for files named disk | 19:47 |
| vishy | kbringard: might need something like that...easy test is to turn off security in /etc/libvirt/qemu.conf and restart libvirt-bin and see if it fixes the issue | 19:49 |
| dubsquared | vvuksan: sorry stepped away for a meeting, i think it's resolved…project name in novarc | 19:51 |
| *** brd_from_italy has joined #openstack | 19:51 | |
| *** adiantum has quit IRC | 19:52 | |
| kbringard | vishy: I don't think I'm running SELinux on the host, but I set security_driver = "none" anyway, but I am still seeing the same error | 19:53 |
| vishy | possibly a permissions issue then | 19:53 |
| vishy | try chmodding the disk file and the backing file | 19:54 |
| vishy | libvirt on debian changes ownership of the file to libvirt-qemu, not sure how it works on centos | 19:54 |
| vvuksan | vishy: apparently i have exhausted my instance quota | 19:55 |
| vvuksan | InstanceLimitExceeded: InstanceLimitExceeded: Instance quota exceeded. You can only run 0 more instances of this type. | 19:55 |
| kbringard | vishy: I've been running everything as root to try to avoid permission problems... it looks like it's all owned by root and set to 644 | 19:55 |
| vvuksan | due to all the unsuccesful launches | 19:55 |
| vvuksan | how do I reset that ? | 19:55 |
| kbringard | have you been terminating them? | 19:55 |
| vvuksan | no | 19:56 |
| vvuksan | i guess I should :-D | 19:56 |
| kbringard | that's what I've been doing | 19:56 |
| kbringard | they just sit in pending, or whatever | 19:56 |
| vvuksan | yep | 19:56 |
| vvuksan | what's the command? | 19:56 |
| kbringard | euca-terminate-instances $instanceid | 19:56 |
| vishy | kbringard: hmm oddd | 19:56 |
| kbringard | is what I've been using | 19:56 |
| vishy | anything interesting in syslog? | 19:56 |
| dubsquared | any ideas? UnknownError: An unknown error has occurred. Please try your request again. — from euca-describe-instances | 19:57 |
| dubsquared | er | 19:57 |
| dubsquared | s/instances/images | 19:57 |
| vvuksan | i got that | 19:57 |
| vvuksan | i retry then it works | 19:57 |
| soren | mtaylor: did you find it yet? | 19:57 |
| mtaylor | soren: yes. all good | 19:57 |
| soren | mtaylor: Cool. | 19:57 |
| soren | vishy: Did you see my lp:~soren/nova/assume-dhcp-start-is-gateway branch? | 19:58 |
| *** hggdh has joined #openstack | 19:58 | |
| soren | (I forget if I mentioned it here) | 19:58 |
| *** gondoi has quit IRC | 19:58 | |
| kbringard | vishy: http://paste.openstack.org/show/577/ | 19:59 |
| kbringard | the qemu-kvm that is installed is from kvm-83-164.el5_4.25 | 19:59 |
| kbringard | I wonder if that is having problems with the qemu I installed from rpmforge | 20:00 |
| vishy | soren: no | 20:00 |
| vvuksan | kbringard: do you have qemu-nbd ? | 20:00 |
| *** gondoi has joined #openstack | 20:00 | |
| kbringard | rpm -qf /usr/bin/qemu-nbd | 20:01 |
| kbringard | qemu-0.12.4-1.el5.rf | 20:01 |
| vvuksan | oh my | 20:01 |
| *** fysa has quit IRC | 20:01 | |
| vishy | soren: does that work? | 20:01 |
| vvuksan | kbringard: i got farther than you | 20:02 |
| vvuksan | by compiling the qemu by hand | 20:02 |
| vvuksan | then replacing qemu-img and qemu-nbd in /usr/bin | 20:02 |
| vvuksan | error I'm getting is this | 20:03 |
| vvuksan | 2011-01-26 12:01:38,655 WARNING nova.virt.libvirt_conn [-] instance instance-0000000e: ignoring error injecting data into image ami-hvxr8c1j (nbd device /dev/nbd14 did not show up) | 20:03 |
| vvuksan | libvir: Domain Config error : internal error invalid domain type %(type)s | 20:03 |
| kbringard | the injecting data error is normal(ish) I get that on ubuntu | 20:03 |
| vvuksan | if I run libvirtd in verbose mode and foreground | 20:04 |
| vishy | vvuksan modprobe nmd to stop the first error | 20:04 |
| vvuksan | should I see any activity ? | 20:04 |
| *** clauden has joined #openstack | 20:05 | |
| vvuksan | you mean nbd | 20:05 |
| vishy | aye | 20:05 |
| vvuksan | k | 20:05 |
| *** clauden has quit IRC | 20:05 | |
| vishy | funny error message on domain type | 20:06 |
| *** clauden has joined #openstack | 20:06 | |
| vvuksan | that didn't seem to do the trick :-( | 20:06 |
| vishy | what is the domain type in xml? | 20:06 |
| vvuksan | which file would that be ? | 20:06 |
| vishy | libvirt.xml in the instances dir | 20:06 |
| soren | vishy: It does for me. | 20:07 |
| vishy | soren: does it fix any particular issue? or is it just a simplification? | 20:07 |
| soren | vishy: I'm not sure I understand it all well enough to say whether there are situations where it won't work. | 20:07 |
| vvuksan | vishy: i am not sure which directory instances dir is :-( | 20:08 |
| vvuksan | would that be /usr/lib/python2.6/site-packages/nova-2011.1-py2.6.egg/nova/virt/libvirt.xml.template | 20:08 |
| vvuksan | ? | 20:08 |
| soren | vishy: I'm not sure, really. | 20:08 |
| kbringard | probably /var/lib/nova/instances/? | 20:09 |
| soren | vishy: Do we ever have multiple dnsmasq's? | 20:09 |
| vishy | vvuksan: no the actual file | 20:09 |
| vishy | soren: yes there is one per project in vlan mode | 20:09 |
| vvuksan | it doesn't appear I have the libvirt.xml :-( | 20:09 |
| soren | vishy: Ok... Why does that work? :) | 20:09 |
| kbringard | vvuksan: once you're in that dir you need to cd to the name of the instance | 20:10 |
| soren | vishy: I mean... If they all specify the same IP to --dhcp-range. | 20:10 |
| kbringard | so like, /var/lib/nova/instances/instance-00000014 | 20:10 |
| kbringard | or whatever | 20:10 |
| *** larstobi has joined #openstack | 20:10 | |
| vvuksan | is there a variable that sets the instances dir ? | 20:11 |
| kbringard | not sure, I don't recall setting something | 20:11 |
| vishy | soren: they don't each network has a different dhcp_start | 20:11 |
| kbringard | bbiab, food | 20:11 |
| vishy | soren: 344 net['dhcp_start'] = str(project_net[2]) | 20:12 |
| vvuksan | vishy: I'm unsure which directory is that :-( | 20:13 |
| vvuksan | any hints | 20:13 |
| vishy | vvuksan: it is wherever your instances directory is set to | 20:14 |
| vvuksan | it's not set :-( | 20:14 |
| vvuksan | ie. it's not set in nova.conf | 20:14 |
| soren | vishy: Oh, right. My mind is elsewhere :( | 20:15 |
| vishy | then probably nova/instances | 20:16 |
| soren | vishy: Oh, that just sets it to gateway+1? | 20:16 |
| soren | vishy: I'm pretty sure it'll work. The actual IP doesn't seem to matter, as long as it's in the right subnet (because where the leases file). | 20:17 |
| soren | It's pretty easy to test, though. | 20:17 |
| soren | I'm just working on something else right now.. I'll get to it in a little bit. | 20:17 |
| *** zenmatt has quit IRC | 20:21 | |
| dubsquared | random question: using FlatManger, can I use internal IP's that are being NAT'd on a firewall as my CC_ADDR, and pointing to all my services? | 20:23 |
| *** vvuksan has quit IRC | 20:24 | |
| *** zenmatt has joined #openstack | 20:24 | |
| dubsquared | i don't see why not…but i am getting random issues here and there | 20:24 |
| *** jtran has left #openstack | 20:26 | |
| *** phymata has joined #openstack | 20:30 | |
| vishy | dubsquared: no routing back to the vms will probably fail | 20:36 |
| phymata | can anyone confirm that the Django based management UI for Nova will be in bexar? i'm 99% sure it will be given it's status on the openstack blueprints (https://blueprints.launchpad.net/openstack?searchtext=bexar+django) but i need to be 100% sure. | 20:36 |
| vishy | phymata it is a separate project | 20:36 |
| vishy | and it already exists | 20:36 |
| dubsquared | vishy: damn, i was hoping that was not the case | 20:36 |
| *** joearnold has quit IRC | 20:36 | |
| phymata | vishy: okay. so is it reasonable to say that it's implemented and will work with the bexar release. | 20:37 |
| *** zenmatt has quit IRC | 20:39 | |
| *** pvo is now known as pvo_away | 20:39 | |
| phymata | vishy: sorry. there should be a ? at the end of my last sentence. | 20:39 |
| uvirtbot | New bug: #708278 in nova "Prevent duplicated VLAN IDs " [Undecided,New] https://launchpad.net/bugs/708278 | 20:42 |
| vishy | phymata: correct | 20:44 |
| phymata | vishy: thanks! | 20:44 |
| *** pvo_away is now known as pvo | 20:53 | |
| JordanRinke_away | ttx around or can you answer this vishy... | 20:55 |
| JordanRinke_away | the current hyperv code and the bugs I am filing/fixing will all be a part of the bexar release correct? | 20:56 |
| *** kpepple has joined #openstack | 20:57 | |
| *** barton has joined #openstack | 20:58 | |
| *** reldan has joined #openstack | 20:59 | |
| *** JordanRinke_away is now known as JordanRinke | 21:00 | |
| vishy | JordanRinke: bugs can still go in | 21:00 |
| *** jfluhmann_ has joined #openstack | 21:01 | |
| vishy | current hyperv code is in already | 21:01 |
| JordanRinke | right, and the code I am filing against, is already in.. yeah cool | 21:01 |
| JordanRinke | just wanted to make sure | 21:01 |
| JordanRinke | since I am doing hyper-v blueprints for cactus | 21:01 |
| *** jfluhmann_ has quit IRC | 21:01 | |
| dubsquared | soooo this euca-describe-images thing…here is what i am seeing http://paste.openstack.org/show/581/ | 21:03 |
| *** fysa_ has joined #openstack | 21:07 | |
| *** vvuksan has joined #openstack | 21:08 | |
| *** adamsih300u has joined #openstack | 21:13 | |
| vishy | haven't seen that, are you trying to use glance with euca? | 21:14 |
| dubsquared | negative, euca w/ objectstore | 21:15 |
| *** jbryce has quit IRC | 21:16 | |
| dprince | Hey guys. I'm hitting this with the latest 623: | 21:16 |
| dprince | https://bugs.launchpad.net/nova/+bug/708301 | 21:16 |
| uvirtbot | Launchpad bug 708301 in nova "s3.py global name '_fix_image_id' is not defined" [Undecided,New] | 21:16 |
| uvirtbot | New bug: #708301 in nova "s3.py global name '_fix_image_id' is not defined" [Undecided,New] https://launchpad.net/bugs/708301 | 21:16 |
| *** littleidea has joined #openstack | 21:16 | |
| dubsquared | did you file that vishy? | 21:17 |
| dubsquared | that was awesome timing.. | 21:17 |
| *** littleidea_ has joined #openstack | 21:20 | |
| *** metoikos has joined #openstack | 21:21 | |
| *** littleidea has quit IRC | 21:23 | |
| *** littleidea_ is now known as littleidea | 21:23 | |
| *** ctennis has quit IRC | 21:23 | |
| *** dprince has quit IRC | 21:27 | |
| *** jcannava has left #openstack | 21:28 | |
| *** pvo is now known as pvo_away | 21:29 | |
| uvirtbot | New bug: #708309 in nova "New localized strings with >1 placeholder still being added to trunk not using dict formating" [Undecided,New] https://launchpad.net/bugs/708309 | 21:31 |
| rlucio | oh im glad someone filed 708301, i was just about to do that (i saw it before lunch( | 21:33 |
| vvuksan | anyone have a clue why I would get this error when launching an instance | 21:34 |
| vvuksan | libvir: QEMU error : internal error Process exited while reading console log output: qemu: could not open disk image /opt/nova-state/instances/instance-00000016/disk | 21:34 |
| vvuksan | this is RHEL and SELINUX is not running | 21:34 |
| JordanRinke | in ubuntu there was a problem with the apparmor config that caused a similar issue | 21:35 |
| JordanRinke | but i think it could access disk and not local | 21:36 |
| vvuksan | hmm | 21:36 |
| *** Cybo has joined #openstack | 21:36 | |
| rlucio | out of disk space maybe? | 21:36 |
| barton | does anyone know how well ElasticFox will work with Bexar? | 21:36 |
| vvuksan | nope plenty of disk | 21:36 |
| JordanRinke | what do permissions look like on that file? | 21:36 |
| vvuksan | 644 owned by root | 21:36 |
| vvuksan | everything is running as root | 21:37 |
| JordanRinke | ah ok | 21:37 |
| vvuksan | even nova-computer | 21:37 |
| vvuksan | compute | 21:37 |
| JordanRinke | my apparmor issue presented even as root | 21:37 |
| vvuksan | right | 21:37 |
| vvuksan | selinux would too | 21:37 |
| vvuksan | but i disabled it | 21:37 |
| JordanRinke | can you manually start that instance now? | 21:38 |
| vvuksan | how would I manually start it | 21:38 |
| sandywalsh | I made a silly mistake on my previous bug fix, dan-prince fixed it up for me. Perhaps a review? https://code.launchpad.net/~dan-prince/nova/fix_image_id_s3/+merge/47589 | 21:39 |
| JordanRinke | virsh console | 21:40 |
| JordanRinke | err | 21:41 |
| vvuksan | JordanRinke: any pointers how to start it from virsh :-) | 21:41 |
| rlucio | sandywalsh: i tested & approved the bug | 21:41 |
| JordanRinke | virsh create libvirt.xml | 21:41 |
| rlucio | ... merge proposal, w/e | 21:41 |
| JordanRinke | or whatever the xml is called | 21:41 |
| vvuksan | JordanRinke: k | 21:41 |
| *** irahgel has joined #openstack | 21:41 | |
| JordanRinke | i think there is a verbose option on that virsh create too | 21:41 |
| JordanRinke | actually | 21:42 |
| *** irahgel has left #openstack | 21:42 | |
| JordanRinke | if the imag you are working on an ubuntu image? | 21:42 |
| JordanRinke | the instance | 21:42 |
| vvuksan | JordanRinke: same error | 21:42 |
| JordanRinke | ok, well at least you have an easy way to test it now | 21:42 |
| vvuksan | right | 21:42 |
| *** rpedde has quit IRC | 21:43 | |
| kpepple | blueprint question: when i create a blueprint (via launchpad), do i also need to create a corresponding wiki page ? | 21:43 |
| JordanRinke | you seem compotent so I will refrain from suggesting standard troubleshooting now :-D | 21:43 |
| vvuksan | :-) | 21:43 |
| vvuksan | bizarre | 21:44 |
| JordanRinke | bah i always incorrectly spell competent... the irony | 21:44 |
| *** pvo_away is now known as pvo | 21:44 | |
| uvirtbot | New bug: #708320 in nova "associate floating ip address to an instance works without allocating address" [Undecided,New] https://launchpad.net/bugs/708320 | 21:45 |
| *** miclorb has joined #openstack | 21:46 | |
| annegentle | devcamcar: are you around for a question or two about installing the dashboard (Django-nova)? | 21:46 |
| *** littleidea has quit IRC | 21:47 | |
| *** GasbaKid has joined #openstack | 21:48 | |
| *** troytoman has quit IRC | 21:49 | |
| *** metoikos has quit IRC | 21:50 | |
| kpepple | annegentle: i've got the dashboard running if devcamcar isn't around | 21:50 |
| *** pvo is now known as pvo_away | 21:50 | |
| *** pvo_away is now known as pvo | 21:50 | |
| annegentle | kpepple: great, thanks! do you know if you'd install it on the database node in a multi-server install? Or always on the cloud controller node? | 21:52 |
| *** pvo is now known as pvo_away | 21:53 | |
| kpepple | annegentle: in the localsettings, it has an option for NOVA_DEFAULT_ENDPOINT which takes a URL pointing to nova-api …. so it doesn't need to be collocated with nova-api or nova-compute. It also doesn't have any login for the nova-db, so I don't think it needs to be on the nova database server. | 21:55 |
| phymata | annegentle: interesting timing, i have a couple of questions about the dashboard too. | 21:56 |
| uvirtbot | New bug: #708329 in nova "DescribeGroups: Groups are not sorted properly" [Undecided,New] https://launchpad.net/bugs/708329 | 21:56 |
| kpepple | annegentle: as such, i don't believe it needs to be either server — but needs to contact the nova-api server via HTTP | 21:56 |
| phymata | annegentle: is development on the dashboard more or less complete and will it work with the bexar release? | 21:56 |
| kpepple | annegentle: i also have a short wiki page on this (with SCREENSHOTS!) but don't know how to get it up on wiki.openstack.com | 21:57 |
| annegentle | kpepple: ok, thanks. And what is this about "create a site in your database and set the SITE_ID setting" - I get that when I run the tests | 21:57 |
| annegentle | kpepple: woo! You should be able to create a wiki account, or just email it to me anne@openstack.org if you like | 21:57 |
| annegentle | kpepple: and I can create the page, attach the screenshots, etc. | 21:57 |
| phymata | kpepple: i'd like to see that wiki page too when it's ready. | 21:58 |
| kpepple | annegentle: did you create your local sqlite db for the dashboard ? | 21:58 |
| annegentle | kpepple: nope! Ok, how did you know to do that? | 21:59 |
| kpepple | annegentle: after you have installed the virtualenv with tools/install_venv.py <path to django-nova/trunk> (to pul in the django-nova libraries), you need to db sync with something like $ tools/with_venv.sh dashboard/manage.py syncdb | 22:01 |
| kpepple | annegentle: remember to create your local/local_settings.py file first tho | 22:01 |
| kpepple | annegentle: with normal settings the sqlite db should be in local/dashboard_openstack ... | 22:02 |
| annegentle | ok, I have gotten confused between django-nova and openstack-dashboard. Whee! | 22:02 |
| annegentle | one is a reference implementation! | 22:03 |
| *** daleolds has quit IRC | 22:03 | |
| kpepple | annegentle: yeah … it's confusing as they are very tightly coupled | 22:03 |
| annegentle | kpepple: ok, I don't have a tools/ directory. Even though I thought I branched the django-nova project. | 22:05 |
| vvuksan | JordanRinke: it is possible that the disk image got munged somehow | 22:07 |
| kpepple | annegentle: sorry. the install_venv.py and syncdb commands are to be done in the openstack-dashboard repo. | 22:07 |
| vvuksan | is there a command line I could use to generate the disk image by hand | 22:07 |
| vishy | check the permissions of instances/_base/**** as well | 22:08 |
| annegentle | kpepple: no need to apologize :) I finally "get it" why I didn't have a tools dir :) | 22:08 |
| vvuksan | vishy: it may not be a permissions issue. kbringard tells me instance/disk file on his Ubuntu is 22 Mb and on Centos 8 MB | 22:09 |
| kpepple | annegentle: so the high level plan to build the dashboard is (1) build django-nova (with the README files boostrap script and buildout), (2) change over to the openstack-dashboard repo, (3) do the install_venv, (4) create your local/local_settings.py file, (5) syncdb and finally run the server | 22:09 |
| vvuksan | that would imply something funny happens on Centos | 22:09 |
| vishy | ivvuksan: it starts out as 8 mb | 22:10 |
| vvuksan | oh | 22:10 |
| vvuksan | this is ttylinux | 22:10 |
| vishy | it is copy on write, so it grows once it launches | 22:10 |
| vvuksan | i see | 22:10 |
| vvuksan | there goes that theory | 22:10 |
| vishy | tha backing image is in _base so it could be a permission issue to the backing image as well | 22:11 |
| vvuksan | i should be able to run kvm with -hda disk -m 512 right ? | 22:11 |
| vvuksan | vishy: permissions look fine :-( | 22:11 |
| phymata | annegentle: so is it safe to say development on the dashboard is more or less complete and will it work with the bexar release? | 22:12 |
| *** MarkAtwood has quit IRC | 22:12 | |
| vishy | perhaps with kvm, but i test with virsh create myself | 22:12 |
| vvuksan | that is what I tried | 22:13 |
| vvuksan | virsh bails out | 22:13 |
| vvuksan | I assume it calls kvm | 22:13 |
| vvuksan | # virsh create libvirt.xml | 22:13 |
| vvuksan | error: Failed to create domain from libvirt.xml | 22:13 |
| vvuksan | error: internal error Process exited while reading console log output: qemu: could not open disk image /opt/nova-state/instances/instance-00000018/disk | 22:13 |
| vvuksan | # file /opt/nova-state/instances/instance-00000018/disk | 22:14 |
| vvuksan | /opt/nova-state/instances/instance-00000018/disk: QEMU Copy-On-Write disk image version 2, size 0 + 25165824 | 22:14 |
| kpepple | phymata: http://blog.rabbityard.com/post/2937203639/openstack-dashboard-video | 22:14 |
| vishy | and nothing in syslog about it? | 22:14 |
| annegentle | phymata: yes, that's a safe statement as far as I know :) | 22:14 |
| vvuksan | vishy: nothing but can't access the disk | 22:15 |
| *** reldan has quit IRC | 22:15 | |
| vvuksan | open it that is | 22:15 |
| vishy | also check for any other errrors in /var/log/libvirt/qemu/instance-00000018 | 22:16 |
| vvuksan | this is the beautiful thing that apparently gets executed | 22:17 |
| vvuksan | /usr/libexec/qemu-kvm -S -M rhel5.4.0 -m 512 -smp 1 -name instance-00000018 -uuid 610e3e89-040b-5e90-2479-f64f6adf552c -nographic -monitor unix:/var/lib/libvirt/qemu/instance-00000018.monitor,server,nowait -boot c -kernel /opt/nova-state/instances/instance-00000018/kernel -initrd /opt/nova-state/instances/instance-00000018/ramdisk -append root=/dev/vda console=ttyS0 -drive file=/opt/nova-state/instances/instance-00000018/disk,if=vir | 22:17 |
| phymata | kpepple: looks good. thanks! | 22:17 |
| vvuksan | if i try that by hand I get | 22:17 |
| vvuksan | qemu: could not open disk image console=ttyS0 | 22:17 |
| vvuksan | so that may be it | 22:17 |
| phymata | annegentle: thanks for the info. | 22:17 |
| vvuksan | apparently not | 22:18 |
| vvuksan | still same thing :-( | 22:18 |
| vvuksan | I straced it | 22:18 |
| vvuksan | it apparently reads first kilobyte in 512 byte chunks | 22:19 |
| annegentle | phymata: I hope to get a wiki page going soon, and I'm trying to write up some "official" docs as well. | 22:19 |
| vvuksan | http://paste.openstack.org/show/582/ | 22:19 |
| vishy | probably reading the image metadata looking for backing image | 22:19 |
| vvuksan | but one would think it would show up in the strace input | 22:20 |
| vishy | try qemu-img info <img> | 22:20 |
| vishy | see what it shows for the backing image and make sure that it is readable | 22:20 |
| vishy | is it possible that the image was created with an older version of qemu-img that is somehow incompatible? | 22:21 |
| phymata | annegentle: great. once bexar is released we're going to take it for a spin and try to get the dashboard working with it. i'll keep an eye out for the wiki page then and contribute if there's anything worthwhile for me to contribute. | 22:21 |
| uvirtbot | New bug: #708347 in nova "RunInstances: Invalid instance type gives improper error message" [Undecided,New] https://launchpad.net/bugs/708347 | 22:21 |
| vvuksan | vishy: http://paste.openstack.org/raw/583/ | 22:21 |
| *** stewart has joined #openstack | 22:22 | |
| *** brd_from_italy has quit IRC | 22:22 | |
| *** londo_ has quit IRC | 22:24 | |
| vvuksan | vishy: i got qemu-img version 0.12.1 | 22:25 |
| uvirtbot | New bug: #708351 in nova "RunInstances: Invalid availability zone doesn't give error message" [Undecided,New] https://launchpad.net/bugs/708351 | 22:26 |
| *** littleidea has joined #openstack | 22:26 | |
| *** littleidea_ has joined #openstack | 22:27 | |
| *** GasbaKid has quit IRC | 22:27 | |
| *** GasbaKid has joined #openstack | 22:28 | |
| *** fysa_ is now known as fysa | 22:29 | |
| *** littleidea has quit IRC | 22:31 | |
| *** littleidea_ is now known as littleidea | 22:31 | |
| vishy | vvuksan: I'm stumped... | 22:34 |
| vvuksan | i'm stumped too | 22:36 |
| vvuksan | i just compiled 0.13 | 22:36 |
| vvuksan | qemu-img | 22:37 |
| vvuksan | still same problem | 22:37 |
| vvuksan | is there a way to recreate the disk by hand ? | 22:37 |
| *** kbringard has quit IRC | 22:37 | |
| vvuksan | i'm sure there is | 22:37 |
| vvuksan | I guess I'll search through source code | 22:38 |
| aliguori | vvuksan, what does qemu-img info show? | 22:41 |
| vvuksan | aliguori: shows this http://paste.openstack.org/raw/583/ | 22:42 |
| uvirtbot | New bug: #708360 in nova "xml namespace should match API version" [Undecided,New] https://launchpad.net/bugs/708360 | 22:42 |
| aliguori | vvuksan, looks like you've got a corrupt image fwiw | 22:43 |
| vvuksan | could be | 22:44 |
| aliguori | what you see in the strace | 22:44 |
| vvuksan | http://paste.openstack.org/raw/582/ | 22:44 |
| *** ppetraki has quit IRC | 22:44 | |
| aliguori | we try to open(O_RDWR), if that fails, we open again O_RDONLY | 22:44 |
| *** mdomsch has quit IRC | 22:44 | |
| aliguori | yeah | 22:44 |
| aliguori | qcow2_open is failing | 22:44 |
| vvuksan | this is RHEL5 | 22:45 |
| vvuksan | i compiled qemu-img by hand | 22:45 |
| aliguori | the format hasn't changed | 22:45 |
| vvuksan | whereas kvm is the stock | 22:45 |
| vvuksan | hmm | 22:45 |
| aliguori | there's a bunch of sanity checking in open, one of those is failing | 22:45 |
| *** dfg has joined #openstack | 22:46 | |
| rlucio | sandywalsh: hey i found another bug with your _fix_image_id method | 22:46 |
| rlucio | sandywalsh: do you just want to fix it in your current branch? | 22:46 |
| uvirtbot | New bug: #708367 in nova "euca-run-instances fails due to KeyError for imageId" [Undecided,New] https://launchpad.net/bugs/708367 | 22:51 |
| *** zenmatt has joined #openstack | 22:52 | |
| *** londo_ has joined #openstack | 22:52 | |
| rlucio | sandywalsh: yea that's the bug, 708367, i put the patch there | 22:52 |
| vvuksan | aliguori: any clues what to do next ? | 22:52 |
| aliguori | vvuksan, give qemu-img check a try | 22:53 |
| aliguori | you may get lucky | 22:53 |
| vvuksan | No errors were found on the image. | 22:54 |
| aliguori | otherwise, take the new qemu, and put a printf in qcow_open() at every goto fail | 22:54 |
| aliguori | and then move the discussion to #qemu on OFTC :-) | 22:54 |
| *** bcwaldon has left #openstack | 22:54 | |
| *** allsystemsarego has quit IRC | 22:56 | |
| uvirtbot | New bug: #708371 in nova "Runinstance : Image available status not checked" [Undecided,New] https://launchpad.net/bugs/708371 | 22:56 |
| *** MarkAtwood has joined #openstack | 23:03 | |
| *** iRTermite has quit IRC | 23:04 | |
| *** matiu has joined #openstack | 23:06 | |
| *** iRTermite has joined #openstack | 23:07 | |
| *** imsplitbit has quit IRC | 23:07 | |
| *** matiu has quit IRC | 23:12 | |
| *** sandywalsh has quit IRC | 23:13 | |
| *** hisaharu has quit IRC | 23:21 | |
| *** gondoi has quit IRC | 23:21 | |
| uvirtbot | New bug: #708383 in nova "Terminate instance: disassociates only one floating ip address" [Undecided,New] https://launchpad.net/bugs/708383 | 23:22 |
| *** MarcMorata has joined #openstack | 23:22 | |
| *** barton has quit IRC | 23:23 | |
| *** rpedde has joined #openstack | 23:24 | |
| *** daleolds has joined #openstack | 23:32 | |
| *** sandywalsh has joined #openstack | 23:33 | |
| *** dprince has joined #openstack | 23:33 | |
| sandywalsh | rlucio, sure, do you have a patch? | 23:35 |
| rlucio | sanywalsh: ya, its in the bug log for bug 708367 | 23:36 |
| uvirtbot | Launchpad bug 708367 in nova "euca-run-instances fails due to KeyError for imageId" [Undecided,New] https://launchpad.net/bugs/708367 | 23:36 |
| dprince | sandywalsh: I saw the merge for the image_id_s3 fix failed because I was missing from Authors. | 23:37 |
| sandywalsh | rlucio, thx, looking | 23:37 |
| sandywalsh | dprince, yup | 23:37 |
| dprince | sandywalsh: I pushed again to that branch with my name added to Authors. | 23:37 |
| dprince | That should fix it. | 23:37 |
| *** larstobi has quit IRC | 23:37 | |
| dprince | Anything I need to do to get that pushed and in? | 23:37 |
| sandywalsh | dprince, care to add https://bugs.launchpad.net/nova/+bug/708367 | 23:38 |
| uvirtbot | Launchpad bug 708367 in nova "euca-run-instances fails due to KeyError for imageId" [Undecided,New] | 23:38 |
| dprince | sandywalsh: sure. You want that in the fix_image_id_s3 branch then? | 23:39 |
| *** MarcMorata has quit IRC | 23:39 | |
| *** rpedde has quit IRC | 23:43 | |
| *** mray has quit IRC | 23:46 | |
| dprince | Looks like the request to merge fix_image_id_s3 is getting reapproved. Thanks guys. | 23:47 |
| *** dprince has quit IRC | 23:47 | |
| *** GasbaKid has quit IRC | 23:47 | |
| *** ranger57 has joined #openstack | 23:49 | |
| openstackhudson | Project nova build #464: SUCCESS in 1 min 29 sec: http://hudson.openstack.org/job/nova/464/ | 23:52 |
| openstackhudson | Tarmac: Fixes a stupid mistake I made when I moved this method from a module into a class. | 23:52 |
| openstackhudson | Thanks Dan for spotting it. | 23:52 |
| sandywalsh | dprince awesome thx again | 23:52 |
| *** ctennis has joined #openstack | 23:53 | |
| sandywalsh | rlucio, thx for the fix ... didn't do any s3 testing | 23:53 |
| rlucio | sandywalsh: np, glad i could help | 23:53 |
| *** jaypipes has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!