| spackest | anyone using debian with openstack? | 00:01 |
|---|---|---|
| *** lloydde has quit IRC | 00:06 | |
| *** crescendo has quit IRC | 00:08 | |
| flaviamissi | spackest: I have ** | 00:10 |
| *** oubiwann1 has quit IRC | 00:10 | |
| *** oubiwann has joined #openstack | 00:10 | |
| *** crescendo has joined #openstack | 00:12 | |
| *** rnirmal has quit IRC | 00:13 | |
| *** aculich has joined #openstack | 00:14 | |
| *** bengrue has quit IRC | 00:16 | |
| *** pixelbeat has quit IRC | 00:18 | |
| *** dwcramer has joined #openstack | 00:21 | |
| *** sleepsonthefloo has quit IRC | 00:21 | |
| *** rods has quit IRC | 00:23 | |
| spackest | flaviamissi: what flavor of debian and how did it go? | 00:23 |
| spackest | can I just follow the normal getting started guide you think? I've been through it a couple times | 00:23 |
| flaviamissi | spackest: I've used the same installation script and worked fine | 00:23 |
| flaviamissi | debian squeeze | 00:23 |
| *** willaerk has quit IRC | 00:24 | |
| *** rods has joined #openstack | 00:24 | |
| spackest | thanks, I'll take a look | 00:24 |
| flaviamissi | :) | 00:25 |
| flaviamissi | spackest: I just remembered, I had an issue with creation of log files | 00:27 |
| flaviamissi | specifically logs/startup.log | 00:27 |
| *** rods has left #openstack | 00:34 | |
| *** spackest has quit IRC | 00:36 | |
| *** whenry has quit IRC | 00:36 | |
| *** micadeyeye has quit IRC | 00:38 | |
| *** micadeyeye_ has quit IRC | 00:38 | |
| *** datajerk has joined #openstack | 00:40 | |
| uvirtbot | New bug: #914049 in nova "EC2 API CreateKeyPair writes private key to log in verbose mode" [Undecided,New] https://launchpad.net/bugs/914049 | 00:41 |
| *** p0 has quit IRC | 00:42 | |
| *** yshh has joined #openstack | 00:42 | |
| *** yshh has joined #openstack | 00:43 | |
| *** rods has joined #openstack | 00:46 | |
| *** yshh has quit IRC | 00:47 | |
| warik | what needs to be setup for the nova-network service to give ip address onto a remote nova-compute? Which flags allows that? | 00:51 |
| *** aculich has quit IRC | 00:54 | |
| *** dragondm has quit IRC | 00:58 | |
| *** livemoon has joined #openstack | 01:01 | |
| *** longhudou has joined #openstack | 01:02 | |
| *** hdms has quit IRC | 01:04 | |
| *** hdms has joined #openstack | 01:04 | |
| longhudou | does openstack support vm live_migration on ubuntu11.04? | 01:09 |
| *** ccc1 has joined #openstack | 01:10 | |
| *** jakedahn has quit IRC | 01:11 | |
| *** krow has quit IRC | 01:12 | |
| *** Binbin has joined #openstack | 01:14 | |
| *** sleepsonthefloo has joined #openstack | 01:14 | |
| *** wilmoore has quit IRC | 01:17 | |
| *** rods has quit IRC | 01:20 | |
| *** scalability-junk has quit IRC | 01:20 | |
| *** rods has joined #openstack | 01:21 | |
| *** ben_duyujie has joined #openstack | 01:22 | |
| *** floehmann has quit IRC | 01:23 | |
| *** vendemiat has quit IRC | 01:23 | |
| *** clauden_ has quit IRC | 01:24 | |
| *** mszilagyi has quit IRC | 01:24 | |
| *** ccc1 has left #openstack | 01:24 | |
| *** rods has quit IRC | 01:25 | |
| tightwork | warik: shot in the dark, all compute nodes to have the same --flat_network_dhcp_start= ? | 01:26 |
| *** krow has joined #openstack | 01:26 | |
| *** rods has joined #openstack | 01:26 | |
| tightwork | warik: which should be the nova-network dnsmasq dhcp server | 01:27 |
| *** krow has quit IRC | 01:28 | |
| *** donaldngo_hp has quit IRC | 01:28 | |
| *** dragondm has joined #openstack | 01:29 | |
| *** andrewsmedina has joined #openstack | 01:29 | |
| warik | tightwork: thanks for your answer! The cc doesn't have this flag ( the instances are getting an ip) and the other nova-compute server has this flag currently (but the instances are not getting ups) however, I tried before without this flag and the result was the same. | 01:30 |
| *** rods has quit IRC | 01:30 | |
| warik | i'm using FlatDHCP | 01:30 |
| *** rods has joined #openstack | 01:32 | |
| *** maplebed has quit IRC | 01:32 | |
| *** tmichael has quit IRC | 01:32 | |
| tightwork | warik: their building but no IP? | 01:33 |
| *** ben_duyujie has quit IRC | 01:33 | |
| warik | tightwork: here are my config files, you might spot something wrong right away! http://pastebin.com/DcHd3Xrp | 01:33 |
| warik | tightwork: that's correct | 01:33 |
| warik | no issue in the logs, whatsoever | 01:34 |
| *** ben_duyujie has joined #openstack | 01:35 | |
| *** tmichael has joined #openstack | 01:36 | |
| *** osier has joined #openstack | 01:36 | |
| tightwork | warik: I want to say you cant multi node out a single nic... | 01:37 |
| tightwork | id like to think I read that in a doc | 01:37 |
| *** krow has joined #openstack | 01:37 | |
| *** aculich has joined #openstack | 01:38 | |
| warik | tightwork: hmm, possible. Although, this setup was working under the 2011.3 version | 01:38 |
| *** krow has quit IRC | 01:39 | |
| warik | I'm running 2012.1 | 01:39 |
| *** dprince has joined #openstack | 01:41 | |
| *** spiffxp has quit IRC | 01:42 | |
| *** rustam has quit IRC | 01:46 | |
| *** lzyeval has joined #openstack | 01:48 | |
| tightwork | warik: examples im seeing show --flat_interface=eth1. You seem to be almost there with the two systems. Networking I've seen is a big part of cloud and making it right can definitely keep your head on... if you have the extra nics and switch.. do it | 01:48 |
| warik | yes! i guess it's worth the try, I'm running this test environment under virtualbox, that will be easy to test | 01:49 |
| *** igors has joined #openstack | 01:50 | |
| *** paulmillar has joined #openstack | 01:50 | |
| *** gyee has quit IRC | 01:56 | |
| *** usef has joined #openstack | 02:03 | |
| *** usef has quit IRC | 02:03 | |
| *** usef has joined #openstack | 02:04 | |
| hugokuo | anyone around ? | 02:05 |
| hugokuo | I need some more information about nova-volume in Essex ...... how to specify the iscsi target location under essex | 02:06 |
| *** msavy has quit IRC | 02:07 | |
| *** pradeep1 has joined #openstack | 02:13 | |
| *** rods has quit IRC | 02:15 | |
| *** erkules|away has joined #openstack | 02:17 | |
| *** erkules has quit IRC | 02:20 | |
| *** warik has quit IRC | 02:27 | |
| *** datajerk has quit IRC | 02:28 | |
| *** datajerk has joined #openstack | 02:28 | |
| *** datajerk has joined #openstack | 02:29 | |
| *** jfluhmann_ has joined #openstack | 02:31 | |
| *** rods has joined #openstack | 02:35 | |
| *** mrjazzcat has quit IRC | 02:35 | |
| *** winston-d has joined #openstack | 02:37 | |
| *** _adjohn has joined #openstack | 02:41 | |
| *** sandywalsh has quit IRC | 02:43 | |
| *** adjohn has quit IRC | 02:43 | |
| *** _adjohn is now known as adjohn | 02:43 | |
| *** rods has quit IRC | 02:46 | |
| *** _adjohn has joined #openstack | 02:46 | |
| *** adjohn has quit IRC | 02:49 | |
| *** flaviamissi has quit IRC | 02:50 | |
| *** _adjohn has quit IRC | 02:50 | |
| *** pradeep1 has quit IRC | 02:52 | |
| *** aliguori_ has quit IRC | 02:54 | |
| uvirtbot | New bug: #822813 in quantum "Improve pylint score" [Medium,In progress] https://launchpad.net/bugs/822813 | 02:55 |
| *** wilmoore has joined #openstack | 02:57 | |
| uvirtbot | New bug: #822519 in quantum "Quantum plugin extension should be able to load from multiple folders" [Undecided,Fix committed] https://launchpad.net/bugs/822519 | 03:02 |
| uvirtbot | New bug: #826563 in quantum "add version.py, similar to nova" [Low,Fix committed] https://launchpad.net/bugs/826563 | 03:02 |
| *** usef has quit IRC | 03:02 | |
| *** usef has joined #openstack | 03:03 | |
| *** andrewbogott has quit IRC | 03:05 | |
| uvirtbot | New bug: #837693 in quantum "Incorrect assertion in test case - _test_show_port_detail()" [Undecided,New] https://launchpad.net/bugs/837693 | 03:06 |
| uvirtbot | New bug: #832338 in quantum "Show a specific extension" [Low,New] https://launchpad.net/bugs/832338 | 03:11 |
| *** wariola has joined #openstack | 03:11 | |
| *** dprince has quit IRC | 03:11 | |
| *** jdurgin has quit IRC | 03:13 | |
| longhudou | anyone who have tested vm live_migration? | 03:17 |
| *** ejat has joined #openstack | 03:18 | |
| *** ejat has joined #openstack | 03:18 | |
| Ryan_Lane | longhudou: I've used it. it's slightly buggy for me in diablo | 03:18 |
| longhudou | nova service is running as root? | 03:18 |
| Ryan_Lane | well, I chose to use tls and certificates for authentication | 03:19 |
| Ryan_Lane | and I'm using gluster as the instance storage | 03:19 |
| longhudou | i am using NFS | 03:19 |
| Ryan_Lane | if you use qemu+ssh, then you need to install ssh keys for root and nova | 03:19 |
| Ryan_Lane | if you are using NFSv4, ensure your users map appropriately | 03:20 |
| Ryan_Lane | or just switch to NFSv3 with auth=sys | 03:20 |
| *** jfluhmann_ has quit IRC | 03:20 | |
| Ryan_Lane | NFSv4 with auth=sys and a properly mapped domain will work, though | 03:21 |
| longhudou | I think I am using just NFSv3 | 03:21 |
| longhudou | ssh keys for root and nova? | 03:21 |
| longhudou | that is what? | 03:21 |
| Ryan_Lane | if you are using qemu+ssh | 03:21 |
| Ryan_Lane | if you are not, then you don't need to worry about that | 03:21 |
| *** RobertLaptop has joined #openstack | 03:22 | |
| *** ejat has quit IRC | 03:22 | |
| longhudou | when i use ps -ef|grep nova, the user of all nova service is nova,not root | 03:22 |
| Ryan_Lane | if you didn't change --live_migration_uri, then you are using qemu+tcp | 03:22 |
| Ryan_Lane | or qemu+tls | 03:23 |
| Ryan_Lane | longhudou: yes, but, nova runs some commands via sudo | 03:23 |
| Ryan_Lane | which turns the user into root | 03:23 |
| longhudou | I don't change that live_migration_rui | 03:23 |
| Ryan_Lane | then you don't need to worry about ssh keys | 03:24 |
| Ryan_Lane | or the fact that it occasionally runs as root | 03:24 |
| Ryan_Lane | what problem are you having, exactly? | 03:24 |
| longhudou | Returning exception (404, u"NOT_FOUND - no exchange '08cb3ce723a1494b8d0969a495e6ebe2' in vhost '/'", (60, 40), 'Channel.basic_publish') to caller | 03:25 |
| *** ejat has joined #openstack | 03:25 | |
| longhudou | err log in nova-scheduler.log | 03:26 |
| longhudou | not sure what caused that | 03:26 |
| *** feiyuliu has quit IRC | 03:26 | |
| *** RobertLaptop has left #openstack | 03:26 | |
| Ryan_Lane | ah. I've seen this too | 03:26 |
| Ryan_Lane | I have no idea why this occurs | 03:27 |
| longhudou | then how did you solve this | 03:27 |
| Ryan_Lane | I haven't. that's why I mentioned it's buggy | 03:27 |
| Ryan_Lane | it works for me occasionally | 03:27 |
| Ryan_Lane | as long as I only run one migration at a time | 03:28 |
| Ryan_Lane | even then occasionally I have problems | 03:28 |
| Ryan_Lane | I'd be lying if I said it was actually usable. | 03:28 |
| longhudou | ok,I got it,you mean now you can occasionally succeed in live_migration | 03:28 |
| longhudou | am i right? | 03:28 |
| Ryan_Lane | yes | 03:29 |
| Ryan_Lane | but, as mentioned, I've run into problems: http://ryandlane.com/blog/2012/01/06/fixing-a-very-broken-instance-live-migration-manually/ | 03:29 |
| *** andrewsmedina has quit IRC | 03:30 | |
| longhudou | ok,let me see | 03:31 |
| *** longhudou_ has joined #openstack | 03:33 | |
| *** eglynn_ has joined #openstack | 03:35 | |
| *** andrewbogott has joined #openstack | 03:35 | |
| *** andrewbogott has quit IRC | 03:35 | |
| *** andrewbogott has joined #openstack | 03:35 | |
| *** longhudou has quit IRC | 03:36 | |
| *** andrewbogott has quit IRC | 03:36 | |
| *** ben_duyujie has left #openstack | 03:38 | |
| *** eglynn has quit IRC | 03:39 | |
| *** longhudou_ has quit IRC | 03:40 | |
| *** longhudou has joined #openstack | 03:41 | |
| *** judd7 has joined #openstack | 03:46 | |
| *** usef has quit IRC | 03:50 | |
| *** saberman_1 has joined #openstack | 03:51 | |
| *** koolhead17|zzZZ is now known as koolhead17 | 03:52 | |
| *** saberman_1 has quit IRC | 03:52 | |
| *** judd7 has quit IRC | 03:55 | |
| *** lzyeval has quit IRC | 03:55 | |
| *** igors has quit IRC | 03:56 | |
| *** dwcramer has quit IRC | 03:58 | |
| *** saberman_1 has joined #openstack | 04:00 | |
| *** saberman_1 has quit IRC | 04:01 | |
| *** hdms has quit IRC | 04:02 | |
| *** dwcramer has joined #openstack | 04:10 | |
| *** dragondm has quit IRC | 04:13 | |
| *** rnorwood has joined #openstack | 04:15 | |
| *** krow has joined #openstack | 04:18 | |
| *** dragondm has joined #openstack | 04:20 | |
| *** endzYme has joined #openstack | 04:25 | |
| *** dwcramer has quit IRC | 04:29 | |
| *** Rajaram has joined #openstack | 04:29 | |
| *** tmichael has quit IRC | 04:34 | |
| *** CrazyThinker has quit IRC | 04:34 | |
| *** hadrian has quit IRC | 04:35 | |
| *** tmichael has joined #openstack | 04:35 | |
| *** pradeep1 has joined #openstack | 04:37 | |
| *** deshantm_ has joined #openstack | 04:43 | |
| *** deshantm has quit IRC | 04:46 | |
| *** rnorwood has quit IRC | 04:49 | |
| *** jfluhmann has joined #openstack | 04:52 | |
| *** dragondm has quit IRC | 04:53 | |
| *** mjfork has quit IRC | 05:17 | |
| *** manish has joined #openstack | 05:24 | |
| manish | Ethernet interface of my instance is not enble | 05:25 |
| manish | please help | 05:25 |
| manish | i am able to take VNC on my instance but Ethernet hardware is not present | 05:25 |
| manish | for instance | 05:26 |
| winston-d | manish : did u check the libvirt.xml for your instance? also the log for qemu? | 05:28 |
| manish | Hello Winston | 05:30 |
| manish | thanks for replying | 05:31 |
| manish | i checked qemu log | 05:31 |
| manish | http://paste.openstack.org/show/4183/ | 05:31 |
| manish | no -nic Argument passed over it | 05:31 |
| *** koolhead17 has quit IRC | 05:34 | |
| manish | libvirt.xml log http://paste.openstack.org/show/4184/ | 05:35 |
| *** wilmoore has quit IRC | 05:37 | |
| *** dpippenger has quit IRC | 05:40 | |
| manish | winston-d libvirt.xml log http://paste.openstack.org/show/4184/ | 05:44 |
| winston-d | manish : your libvirt.xml is too simple, there's no NIC session. | 05:45 |
| winston-d | you should check libvirt.template.xml | 05:45 |
| manish | yes winston | 05:45 |
| manish | how can i fix it | 05:45 |
| manish | winston-d how can i fix it | 05:47 |
| winston-d | fix your libvirt.template | 05:48 |
| manish | where is it excalty | 05:48 |
| manish | on nova-compute | 05:48 |
| manish | winston-d where is it exacltly nova-compute | 05:50 |
| manish | winston-d my openstack archtecture http://paste.openstack.org/show/4185/ | 05:50 |
| manish | winston-d my nova.conf file http://paste.openstack.org/show/4186/ | 05:51 |
| winston-d | manish : i don't know ubuntu, i'm using RHEL. can you check /var/lib/nova ? | 05:52 |
| manish | yes sure | 05:53 |
| *** pradeep1 has quit IRC | 05:53 | |
| manish | winston-d yes sure | 05:54 |
| manish | there is no file with name libvirt.tmplet | 05:54 |
| manish | there is no file with name libvirt.templet | 05:55 |
| manish | winston-d there is no file with name libvirt.templet | 05:55 |
| *** pradeep1 has joined #openstack | 05:55 | |
| manish | winston-d which service is responsible for creating libvirt.xml | 05:56 |
| manish | winston-d for instance | 05:57 |
| winston-d | manish : how about /usr/share/nova? | 05:57 |
| winston-d | manish : nova-compute is. | 05:57 |
| *** deshantm_ has quit IRC | 05:57 | |
| manish | winston-d there is no such file /usr/share/nova | 05:58 |
| winston-d | manish : well, then I can do anything to help. I don't know where ubuntu package puts the libvirt template. sorry | 05:59 |
| manish | winston-d all nova in my system http://paste.openstack.org/show/4187/ | 05:59 |
| winston-d | manish : you can check /usr/share/pyshared/nova, /usr/lib/python2.6/dist-packages/nova and /usr/lib/python2.7/dist-packages/nova | 06:00 |
| manish | winston-d i am checking | 06:01 |
| *** anpl1 has quit IRC | 06:04 | |
| *** oubiwann has quit IRC | 06:08 | |
| manish | winston-d http://paste.openstack.org/show/4189/ | 06:13 |
| manish | winston-d lot of files is here which is exactly responsible for genreting libvirt.xml | 06:14 |
| winston-d | well, inside virt/libvirt/connection.py, there's a function call 'spawn', the thing it does, is to generate xml | 06:18 |
| *** zigo has joined #openstack | 06:21 | |
| *** smoser has quit IRC | 06:23 | |
| hugokuo | bonjour | 06:30 |
| manish | winston-d def spawn http://paste.openstack.org/show/4190/ | 06:30 |
| *** zigo-_- has joined #openstack | 06:31 | |
| *** zigo has quit IRC | 06:32 | |
| *** mohits has joined #openstack | 06:32 | |
| winston-d | manish : i think you need to crack down this issue via: 1) find your libvirt.template, 2) add debug output to see what network_info is. | 06:32 |
| *** bepernoot has joined #openstack | 06:33 | |
| manish | winston-d i in my system there is no file named libvirt.templet so better to be go with log | 06:34 |
| winston-d | manish : you better find someone who is using ubuntu to help u | 06:34 |
| winston-d | there should be one, but the naming may be different | 06:35 |
| manish | winston-d def spawn what its mean exactly block_device_info=None | 06:35 |
| manish | winston-d http://paste.openstack.org/show/4190/ | 06:36 |
| winston-d | manish : that means, the default value for block_device_info is None if there's no input | 06:36 |
| manish | winston-d k | 06:37 |
| manish | winston-d thanks winston-d i trying with logs if you find any suggestion regarding libvirt.xml please revert | 06:39 |
| winston-d | you can ask hugokuo, i think he's using ubuntu | 06:39 |
| *** smoser has joined #openstack | 06:40 | |
| hugokuo | wasup | 06:40 |
| manish | hello hugkuo | 06:40 |
| manish | hugkuo http://paste.openstack.org/show/4184/ look this file network is not attched with that | 06:41 |
| *** Ryan_Lane has quit IRC | 06:42 | |
| manish | hello hugokuo | 06:44 |
| manish | hugokuo http://paste.openstack.org/show/4184/ look this file network is not attched with that | 06:45 |
| *** mohits1 has joined #openstack | 06:47 | |
| *** mohits has quit IRC | 06:47 | |
| *** mohits1 is now known as mohits | 06:48 | |
| *** mohits has joined #openstack | 06:48 | |
| *** ejat has quit IRC | 06:49 | |
| *** ejat has joined #openstack | 06:50 | |
| *** guigui has joined #openstack | 06:50 | |
| *** bobi111 has joined #openstack | 06:51 | |
| bobi111 | Hi,I tried to upgrade glance on Ubuntu 11.10 using trunk repository. And I get this error: http://pastebin.com/upSwAsT8. Any idea how to fix? | 06:51 |
| *** ejat has quit IRC | 06:52 | |
| *** Ryan_Lane has joined #openstack | 06:54 | |
| *** ejat has joined #openstack | 06:54 | |
| manish | hello i need help on ubuntu openstack | 06:58 |
| manish | hello hugokuo i need help regarding ubuntu openstack | 06:59 |
| *** hingo has joined #openstack | 07:00 | |
| *** ejat has quit IRC | 07:03 | |
| *** adjohn has joined #openstack | 07:03 | |
| *** adjohn has quit IRC | 07:04 | |
| *** bepernoot has quit IRC | 07:08 | |
| *** Ryan_Lane has quit IRC | 07:08 | |
| *** mohits1 has joined #openstack | 07:08 | |
| *** CrazyThinker has joined #openstack | 07:11 | |
| *** mohits has quit IRC | 07:11 | |
| *** mohits has joined #openstack | 07:12 | |
| *** mohits1 has quit IRC | 07:13 | |
| *** ldlework has joined #openstack | 07:13 | |
| hugokuo | manish , there's my libvirt.xml http://paste.openstack.org/show/4191/ | 07:16 |
| *** mohits has quit IRC | 07:16 | |
| hugokuo | --flat_injected=False plz try to set it as "True" | 07:17 |
| *** mindpixel has joined #openstack | 07:17 | |
| *** Remco_ has joined #openstack | 07:17 | |
| manish | hugokuo interface is missing in my libvirt.xml | 07:19 |
| *** mohits has joined #openstack | 07:19 | |
| manish | hugokuo how can i fix it have u any idea about that | 07:20 |
| *** llang629 has joined #openstack | 07:20 | |
| *** llang629 has left #openstack | 07:21 | |
| hugokuo | manish , I'm not sure if flatmode will generate nic interface info in libvirt.xml | 07:21 |
| manish | k | 07:22 |
| hugokuo | maybe you can try to setup network manager as FlatDHCP | 07:22 |
| *** katkee has joined #openstack | 07:22 | |
| *** zigo-_- has quit IRC | 07:23 | |
| manish | hugokuo Right Hugokuo http://paste.openstack.org/show/4186/ | 07:23 |
| *** zigo has joined #openstack | 07:23 | |
| *** GheRivero has quit IRC | 07:26 | |
| *** mgoldmann has joined #openstack | 07:29 | |
| *** miclorb_ has quit IRC | 07:33 | |
| *** CrazyThinker has quit IRC | 07:34 | |
| *** CrazyThinker has joined #openstack | 07:35 | |
| *** lionel has quit IRC | 07:36 | |
| *** lionel has joined #openstack | 07:37 | |
| *** Remco_ has quit IRC | 07:38 | |
| winston-d | hugokuo : i've got some questions about multizone. maybe you can help? | 07:39 |
| *** hggdh has joined #openstack | 07:39 | |
| *** zigo has quit IRC | 07:43 | |
| *** zigo has joined #openstack | 07:45 | |
| *** popux has joined #openstack | 07:47 | |
| *** saju_m has joined #openstack | 07:48 | |
| bobi111 | Hi,What should I do If I have errors when upgrading glance from trunk repository? | 07:50 |
| *** davlap has quit IRC | 07:52 | |
| *** mohits has quit IRC | 07:52 | |
| *** mohits has joined #openstack | 07:52 | |
| longhudou | can not start instance on one of the compute node | 07:54 |
| longhudou | error is " waiting for metadata service at http://169.254.169.254/2009-04-04/meta-data/instance-id" | 07:54 |
| longhudou | anyone know that | 07:54 |
| *** rgidron2 has joined #openstack | 07:56 | |
| *** ejat has joined #openstack | 07:58 | |
| *** ejat has joined #openstack | 07:58 | |
| *** marcuz has joined #openstack | 07:59 | |
| *** lionel has quit IRC | 08:04 | |
| *** lionel has joined #openstack | 08:04 | |
| *** reidrac has joined #openstack | 08:04 | |
| *** zigo-_- has joined #openstack | 08:05 | |
| *** zigo has quit IRC | 08:05 | |
| *** ninkotech has quit IRC | 08:09 | |
| *** bepernoot has joined #openstack | 08:10 | |
| *** heyho has joined #openstack | 08:14 | |
| *** rocambole has joined #openstack | 08:16 | |
| *** zigo-_- has quit IRC | 08:18 | |
| *** zigo has joined #openstack | 08:18 | |
| *** antenagora has joined #openstack | 08:25 | |
| *** Remco_ has joined #openstack | 08:27 | |
| *** robbiew has joined #openstack | 08:30 | |
| *** ccc1 has joined #openstack | 08:31 | |
| *** mrevell has joined #openstack | 08:33 | |
| *** krow has quit IRC | 08:34 | |
| *** zul has joined #openstack | 08:34 | |
| *** thickskin has joined #openstack | 08:36 | |
| *** ccc1 has quit IRC | 08:41 | |
| *** sleepsonthefloo has quit IRC | 08:42 | |
| *** Remco_ has quit IRC | 08:44 | |
| *** tmichael has quit IRC | 08:46 | |
| *** nacx has joined #openstack | 08:48 | |
| *** mohits has left #openstack | 08:48 | |
| *** shaon has joined #openstack | 08:52 | |
| manish | winston-d i find the file libvirt.xml.template | 08:52 |
| *** Ramonster has joined #openstack | 08:52 | |
| manish | winston-d path is /usr/lib/python2.7/dist-packages/nova/virt/libvirt.xml.template | 08:53 |
| *** katkee has quit IRC | 08:53 | |
| winston-d | good 4 u | 08:54 |
| manish | winston-d templet have interface tag | 08:56 |
| *** mohits has joined #openstack | 08:57 | |
| manish | winston-d http://paste.openstack.org/show/4192/ | 08:57 |
| manish | winston-d so why it not reflected on generated file | 08:58 |
| *** ldlework has quit IRC | 08:59 | |
| *** eglynn_ has quit IRC | 09:06 | |
| *** foexle has joined #openstack | 09:08 | |
| foexle | ahoi | 09:08 |
| *** KAM has joined #openstack | 09:10 | |
| *** mrevell has quit IRC | 09:11 | |
| *** mrevell has joined #openstack | 09:12 | |
| *** mrevell has joined #openstack | 09:12 | |
| *** KAM has left #openstack | 09:13 | |
| *** LinuxJedi has joined #openstack | 09:14 | |
| *** eglynn_ has joined #openstack | 09:18 | |
| *** uksysadmin has joined #openstack | 09:21 | |
| Madkiss | hi uksysadmin | 09:24 |
| Madkiss | did you get swift+keystone to work by any chance? ;) | 09:25 |
| uksysadmin | Madkiss, yes. | 09:27 |
| uksysadmin | Although I tried to reproduce my working yesterday on a test set up and it didn't work so going back over my notes again. | 09:28 |
| *** thickskin has left #openstack | 09:28 | |
| *** katkee has joined #openstack | 09:28 | |
| heyho | hello, I can't ping my instances, not even inside the private network..reading a few irc logs it seems that the --vlan flag in nova.conf need to be set with eth# and not br100.. should this change solve my problem? | 09:36 |
| manish | heyho try to VNC on instances | 09:38 |
| manish | heyho through private network | 09:39 |
| manish | heyho vnc://nova-compute-ip:vncport | 09:39 |
| uksysadmin | heyho, and see what the console output says for the instance to see if your interfaces have actually come up | 09:39 |
| *** eglynn_ has quit IRC | 09:39 | |
| heyho | ok thanks manish and uksysadmin I'll try that | 09:40 |
| *** darraghb has joined #openstack | 09:41 | |
| uksysadmin | On a 2 interface VM, I've got an interface that has my private network on, and an interface that serves my public network that I use to communicate with the api server and access instances | 09:41 |
| uksysadmin | in my nova.conf I have --vlan_interface=${PRIVATE_INTERFACE} | 09:41 |
| manish | uksysadmin interface file is not generated in libvirt.xml http://paste.openstack.org/show/4184/ Networking Mode is Flat Mode | 09:41 |
| uksysadmin | manish, I've not set up an openstack environment using flat mode yet | 09:43 |
| manish | uksysadmin http://paste.openstack.org/show/4186/ | 09:44 |
| manish | uksysadmin is these error due to Flat Mode | 09:44 |
| uksysadmin | sorry manish, do you have an issue? I was giving that info to heyho to see if this helps him. | 09:44 |
| uksysadmin | manish, I'm not sure - never used flat mode | 09:45 |
| *** ejat has quit IRC | 09:45 | |
| *** JesperA has joined #openstack | 09:45 | |
| manish | uksysadmin ok | 09:45 |
| *** miclorb_ has joined #openstack | 09:46 | |
| uksysadmin | manish, I would suggest viewing the console output to see if you have interfaces coming up too if you're finding that you're not getting IPs assigned to instances | 09:46 |
| uksysadmin | and view the compute and network logs | 09:46 |
| *** perestrelka has quit IRC | 09:47 | |
| *** rustam has joined #openstack | 09:48 | |
| manish | uksysadmin i am able to take VNC of Instance | 09:56 |
| manish | uksysadmin its Working Fine | 09:56 |
| uksysadmin | vnc is a direct connection - its not using networking | 09:56 |
| uksysadmin | but if you can get on, do ifconfig | 09:57 |
| manish | uksysadmin ya i tried ifconfig or lspci | 09:57 |
| uksysadmin | and is anything listed? what's the instance os? | 09:57 |
| manish | uksysadmin ethernet hardware is not present | 09:57 |
| manish | uksysadmin http://paste.openstack.org/show/4186/ | 09:58 |
| manish | uksysadmin sorry http://paste.openstack.org/show/4184/ | 09:58 |
| manish | uksysadmin no interface tag | 09:58 |
| *** Carmivore has quit IRC | 09:59 | |
| uksysadmin | what's the instance os and distro? | 09:59 |
| manish | uksysadmin debian | 09:59 |
| manish | uksysadmin libvirt.xml created before instance spaned so no interface argument passed with kvm | 09:59 |
| uksysadmin | where di you get it from? sounds like it doesn't have virtio drivers? | 10:00 |
| uksysadmin | I've not used debian under my openstack install, nor used flat mode so don't know the effects/expected outcome | 10:00 |
| *** zigo has quit IRC | 10:01 | |
| *** zigo has joined #openstack | 10:02 | |
| manish | uksysadmin yes this sys file have't any network interface | 10:03 |
| manish | uksysadmin yes this libvirt.xml file have't any network interface | 10:03 |
| manish | uksysadmin i triied ubuntu & windows as A image all have same problem | 10:04 |
| *** Carmivore has joined #openstack | 10:04 | |
| *** micadeyeye has joined #openstack | 10:04 | |
| *** micadeyeye_ has joined #openstack | 10:04 | |
| *** micadeyeye has quit IRC | 10:04 | |
| *** mohits has quit IRC | 10:06 | |
| *** micadeyeye has joined #openstack | 10:06 | |
| heyho | uksysadmin manish: i tried to use xtightvncviewer from my compute node and specify the private ip of a running instance.. but nothing happens. Also, this is a console log http://paste.openstack.org/show/3rD7IE8XszcLv4cGmxxq/ route failed :/ | 10:06 |
| *** micadeyeye_ has quit IRC | 10:06 | |
| *** Kyle__ has quit IRC | 10:10 | |
| *** ccustine has quit IRC | 10:10 | |
| *** perestrelka has joined #openstack | 10:11 | |
| *** ccustine has joined #openstack | 10:11 | |
| manish | heyho go to /var/log/libvirt/qemu/instance-name.log check on winch particular port vnc is working | 10:12 |
| *** Kyle__ has joined #openstack | 10:12 | |
| manish | heyho first try to ping your nova-compute node if its reachable then try to vnc over that | 10:13 |
| *** livemoon has quit IRC | 10:15 | |
| heyho | I can ping and ssh my nova-compute node..actually I run commands in it via ssh. Reading the log of one of my instance, it seems that the vnc uri is 0.0.0.0:0 (maybe because the instance is in the same compute node?) but at the bottom of the log I read http://paste.openstack.org/show/JDUa20mcrfDn8AnklnIh/ | 10:18 |
| *** maploin has joined #openstack | 10:21 | |
| *** eglynn_ has joined #openstack | 10:22 | |
| *** erkules|away has quit IRC | 10:25 | |
| *** erkules|away has joined #openstack | 10:25 | |
| *** erkules|away is now known as erkules | 10:25 | |
| *** wariola has quit IRC | 10:27 | |
| *** Carmivore has quit IRC | 10:27 | |
| *** zigo has quit IRC | 10:29 | |
| *** zigo-_- has joined #openstack | 10:29 | |
| *** nid0 has quit IRC | 10:30 | |
| *** nid0 has joined #openstack | 10:30 | |
| manish | heyho make fresh image and bundle it http://cssoss.wordpress.com/2011/11/28/bundling-debian-image-for-openstack/ | 10:31 |
| *** Carmivore has joined #openstack | 10:32 | |
| *** eglynn_ has quit IRC | 10:33 | |
| *** robbiew has quit IRC | 10:36 | |
| *** Remco_ has joined #openstack | 10:37 | |
| heyho | thanks manish, will try | 10:40 |
| *** robbiew has joined #openstack | 10:42 | |
| *** miclorb_ has quit IRC | 10:43 | |
| *** paulmillar has quit IRC | 10:44 | |
| uksysadmin | heyho, you using debian too? | 10:52 |
| uksysadmin | I'd keep things simple and try known distros that work with OpenStack first then try to use something that you've created yourself | 10:53 |
| *** paulmillar has joined #openstack | 10:56 | |
| heyho | uksysadmin: I'm using Ubuntu 11.10 | 11:01 |
| *** zigo-_- has quit IRC | 11:02 | |
| uksysadmin | heyho, at least yours has the interface present - its just not communicating with your environment to get an address - which is very probably down to configuration | 11:03 |
| *** eglynn_ has joined #openstack | 11:03 | |
| *** lionel has quit IRC | 11:04 | |
| *** lionel has joined #openstack | 11:04 | |
| *** flaviamissi has joined #openstack | 11:05 | |
| *** Arminder-Office has quit IRC | 11:05 | |
| *** Arminder has joined #openstack | 11:07 | |
| *** GheRivero has joined #openstack | 11:08 | |
| *** Arminder has quit IRC | 11:08 | |
| *** Arminder has joined #openstack | 11:08 | |
| *** ahasenack has joined #openstack | 11:09 | |
| *** Arminder has quit IRC | 11:12 | |
| *** mcclurmc has quit IRC | 11:14 | |
| *** mcclurmc has joined #openstack | 11:14 | |
| *** hggdh has quit IRC | 11:21 | |
| rgidron2 | hi - how do I get the ip adress of a running instance? | 11:22 |
| rgidron2 | euca-describe-instance does not show an ip-address | 11:27 |
| heyho | uksysadmin: other than not getting an address, my instance is also not finding the ramdisk (Couldn't find valid RAM disk image starting at 0) which I think is a more serious problem.. also, the vnc console shows just the seabios string and nothing else.. :/ | 11:27 |
| *** eglynn_ has quit IRC | 11:27 | |
| *** eglynn_ has joined #openstack | 11:28 | |
| uksysadmin | rgidron2, it means your instances hasn't started yet - check the state... probably says "pending" | 11:28 |
| rgidron2 | state is running | 11:28 |
| uksysadmin | rgidron2, euca-get-console of the instance id and view the console output to see why your interface hasn't come up | 11:29 |
| uksysadmin | heyho, the ubuntu cloud images don't ship with a ramdisk so you can ignore that | 11:30 |
| *** James has joined #openstack | 11:30 | |
| James | hello all | 11:31 |
| uksysadmin | heyho, ah - I thought you showed output where it was getting console output | 11:31 |
| *** James is now known as Guest29716 | 11:31 | |
| rgidron2 | uksysadmin: euca-get-console-output just hangs | 11:31 |
| Guest29716 | if have a relative easy question to ask, but i can't seem to find any answer to it in the documentation | 11:31 |
| Guest29716 | if i want to set up a server as a cloud controller, does it require the nova-compute component, or can it be omitted? | 11:32 |
| uksysadmin | heyho, I'm a bit confused - you can vnc in but your instance isn't spitting out stuff to console log? that sounds like two different instances you're looking at? | 11:32 |
| uksysadmin | omitted | 11:32 |
| *** jedi4ever has joined #openstack | 11:33 | |
| Guest29716 | currently it is installed, if i perform an uninstall it won't affect my current installation other than removing a compute node form the system right? | 11:33 |
| uksysadmin | rgidron2, check all your logs - esp the nova-api log | 11:34 |
| Kiall | and nova-compute for errors re reading the console log file... | 11:35 |
| uksysadmin | Guest29716, shouldn't do - not done it, but its not a pre-req | 11:35 |
| Guest29716 | ok | 11:35 |
| Guest29716 | thanks for the help | 11:35 |
| uksysadmin | production setups don't run compute on the controller boxes | 11:35 |
| Guest29716 | ok | 11:36 |
| *** Guest29716 has left #openstack | 11:36 | |
| heyho | uksysadmin: I used the wrong vnc port actually.. the console log spits out stuff and doing vncviewer 10.10.10.2:0 I get to a login screen! ubuntu:ubuntu doesn't seem to be the credentials. Is there a way I can pass the keypair? | 11:37 |
| *** flaviamissi has quit IRC | 11:38 | |
| rgidron2 | uksysadmin: is there a way to "kill" and instance? | 11:38 |
| *** livemoon has joined #openstack | 11:38 | |
| uksysadmin | heyho, keypairs are only for ssh | 11:40 |
| uksysadmin | rgidron2, euca-terminate-instances id | 11:40 |
| * Kiall wonders why people use the euca tools with openstack, the native tools offer more functionality :/ | 11:41 | |
| rgidron2 | uksysadmin: thanks, terminated and used cloud-run-instances - now I have an IP but instance is pending... too long, any clue? | 11:42 |
| Kiall | rgidron: have you checked the logs? The client side tools (ie euca_* and cloud_*) are designed not to reveal anything about the internal workings of the cloud... | 11:43 |
| uksysadmin | euca-get-console-output again to see if it is booting | 11:44 |
| Kiall | uksysadmin: if its in the "pending" state, nova wont even attempt to access the logfile.. | 11:44 |
| uksysadmin | Kiall, people use euca2ools because it fits better with a hyrbid cloud solution... many enterprises are using OpenStack along with EC2 | 11:45 |
| uksysadmin | sure, there are nova specifics, but there are installations that use the pretty much unofficial standard which is the EC2 api | 11:45 |
| Kiall | uksysadmin: good point .. didnt think of that to be honest! | 11:45 |
| uksysadmin | :) | 11:46 |
| rgidron2 | get-console-output returns an error, nova-api log shows errors relating to the console.fifo.out file saying it does not exist | 11:48 |
| heyho | uksysadmin: using vnc, how do I log in the instance? user:ubuntu password:ubuntu doesnt work and I'm using uec images | 11:49 |
| Kiall | rgidron2: yes, the instance changes to "active" the moment the KVM command runs, so if its not active, there will never be a logfile to check | 11:49 |
| Kiall | the only way to diagnose an instance stuck as pending is via the server side logs.. | 11:50 |
| Kiall | heyho: Can you imagine how many pawned instances would be sitting around on EC2 if there was a default password? ;) | 11:50 |
| rgidron2 | kiall; makes sense of course, server side logs are where? | 11:50 |
| Kiall | You need to supply an ssh key name with the -k option, and SSH to it using that.. | 11:50 |
| Kiall | you can then set a root password if you like.. | 11:51 |
| Kiall | rgidron2: usually /var/log/nova/* | 11:51 |
| Kiall | Depends how you installed everything though.. | 11:51 |
| rgidron2 | I followed the diablo getting started guide, have the logs all at the standard /var/log location | 11:52 |
| rgidron2 | I will try to look closer at nova-api.log | 11:52 |
| Kiall | Then thats where the logs are :) | 11:52 |
| rgidron2 | :-) | 11:53 |
| Kiall | usually, issue re pending instances have something useful in the nova-compute or nova-network logs | 11:53 |
| *** andrewsmedina has joined #openstack | 11:53 | |
| heyho | Kiall: I can't ssh in it.. I can't even ping the instance private ip inside the network; the only things that show me something are the log and vncviewer .. I may have network configuration problems.. | 11:55 |
| Kiall | heyho: can you ping/ssh to it from the controller node, or the compute node the instance is running on? | 11:56 |
| rgidron2 | kiall: ERROR: trying to add VLAN #100 to IF -:br100:- error: No such device\n' in nova-compute | 11:56 |
| heyho | no :/ | 11:56 |
| Kiall | (ie, you might be trying from you're desktop) | 11:56 |
| rgidron2 | does that mean anything to you? | 11:56 |
| Kiall | rgidron2: sounds like you need to fixup your /etc/network/interfaces | 11:56 |
| Kiall | You're using VLAN networking I assume? | 11:57 |
| Kiall | not flat or flatDHCP/ | 11:57 |
| Kiall | not flat or flatDHCP?* | 11:57 |
| bobi111 | Hi.After installing and running nova-manage db sync I get ""Lost connection to MySQL server at 'reading initial communication packet', system error: 0" | 11:57 |
| uksysadmin | heyho, you can't log in | 11:58 |
| rgidron2 | this is where my newbie face comes out screaming if it hasn't yet Kiall | 11:58 |
| heyho | Kiall: Yes, I created three private networks.. I have two tenants but three private networks | 11:58 |
| uksysadmin | not via the console | 11:58 |
| rgidron2 | I don't know, I installed from scratch using to PC's with ubuntu and 2 nics on each | 11:58 |
| heyho | uksysadmin: yes, I got it thanks | 11:58 |
| Kiall | heyho: something in your network config is off if you cant ping it on its fixed IP from both the controller for compute nodes.. | 11:59 |
| *** markvoelker1 has joined #openstack | 11:59 | |
| Kiall | I'd go through the docs on networking again to be sure you didn't miss anything.. | 11:59 |
| heyho | Kiall: yes that's what I'll do | 11:59 |
| rgidron2 | kiall: if I pasebin my /etc/network/interfaces would you be able to tell anything from it? | 11:59 |
| *** ahasenack has quit IRC | 11:59 | |
| Kiall | rgidron2: lol .. there are 3 different network modes.. flat, flat dhcp and vlan.. It's important you understand the differences and choose the one appropriate for you | 12:00 |
| *** ahasenack has joined #openstack | 12:00 | |
| Kiall | rgidron2: the docs page on "networking options" gives a somewhat cryptic explanation of the three options.. but worth a read anyhow.. | 12:01 |
| Kiall | I use the VLAN mode, But most people choose the FlatDHCP mode... | 12:01 |
| Kiall | the "Flat" mode is almost never used (that I'm aware of) | 12:01 |
| rgidron2 | http://paste.openstack.org/show/4196/ is mt interfaces file | 12:02 |
| *** mrevell has quit IRC | 12:02 | |
| Kiall | rgidron2: without knowing which network mode you need/want, there is no point checking the file.. the file would look different depending on the modes.. | 12:02 |
| *** zul has quit IRC | 12:03 | |
| *** heyho has quit IRC | 12:04 | |
| Kiall | rgidron2: a (potentially) easy way to tell, are your switches VLAN capable? | 12:04 |
| *** bobi111 has quit IRC | 12:04 | |
| *** pradeep1 has quit IRC | 12:05 | |
| rgidron2 | kiall: is this networking mode a setting to nova in any of the config files? I believe I should be using flatDHCP because I don't have a vlan enables switch in my setup here | 12:05 |
| *** ninja4u has joined #openstack | 12:05 | |
| *** msavy has joined #openstack | 12:05 | |
| Kiall | rgidron2: okay, then you have some misconfiguration.. :) | 12:06 |
| Kiall | There is an example config for FlatDHCP here.. http://docs.openstack.org/diablo/openstack-compute/admin/content/configuring-flat-dhcp-networking.html | 12:06 |
| *** ninja4u has quit IRC | 12:06 | |
| Kiall | Once you have that right, you will need to empty the "networks" and "fixed_ips" tables in MySQL, and re-add the networks using "nova-manage network add ...." | 12:06 |
| *** waytogobastard has joined #openstack | 12:07 | |
| rgidron2 | kiall: wow, nice.. I will get started with the link example you provided and probably be back for help with that mysql cleanup | 12:07 |
| rgidron2 | Thanks!! | 12:07 |
| Kiall | np, good luck ;) | 12:08 |
| *** heyho has joined #openstack | 12:08 | |
| *** ahasenack has quit IRC | 12:10 | |
| rgidron2 | kiall: interestingly enough I don't have a --network_manager entry in my nova.conf at all | 12:11 |
| *** ahasenack has joined #openstack | 12:11 | |
| rgidron2 | is that supposed to work? | 12:11 |
| Kiall | rgidron2: exactly, the default is VLAN mode :) | 12:11 |
| Kiall | yea you can leave options out and the defaults will be assumed | 12:11 |
| *** eagles051387|| has joined #openstack | 12:12 | |
| eagles051387|| | hey guys what exactly is open stack | 12:12 |
| eagles051387|| | is it its own virtualization technology or does it work in conjunction with existing virtualization technology | 12:12 |
| *** manish has left #openstack | 12:12 | |
| rgidron2 | OK so this is a demo setup for research purposes, is there an easy way to fixup my setup - for example if I added a network manager flat entry | 12:13 |
| Kiall | it sits on top of kvm/xen/lxc etc etc | 12:13 |
| rgidron2 | what does the range entry do? | 12:13 |
| eagles051387|| | Kiall: does it make setting up ipv6 guests easy to do | 12:13 |
| Kiall | rgidron2: I've honestly never used the flatDHCP manager and have no clue what the settings mean, sorry ;) | 12:13 |
| rgidron2 | np | 12:13 |
| *** manish has joined #openstack | 12:14 | |
| Kiall | eagles051387||: it supports ipv6, but I've not used it.. | 12:14 |
| eagles051387|| | ahh ok | 12:14 |
| manish | kiall hello | 12:14 |
| eagles051387|| | :-/ humm also does open stack support provisioning of guests | 12:14 |
| Kiall | ipv6 is on my list of todo's.. we have an assignment just sitting there ;) | 12:14 |
| eagles051387|| | hehe | 12:14 |
| eagles051387|| | what about provisioning of guests Kiall | 12:15 |
| Kiall | no - openstack can be considered equal to EC2 and S3.. | 12:15 |
| eagles051387|| | does openstack do that | 12:15 |
| manish | kiall i have problem with FLATMANAGER | 12:15 |
| eagles051387|| | humm ok guess on ubuntu thats where orchestra and cobbler comes in handy I'm guessing | 12:15 |
| Kiall | It provisions instances based on an disk image, nothing more | 12:15 |
| eagles051387|| | nothing with lvm | 12:15 |
| Kiall | manish: not sure why you're asking me directly ;) | 12:15 |
| Kiall | Anyway - Back to work I go.. | 12:16 |
| eagles051387|| | have fun Kiall | 12:16 |
| eagles051387|| | i need to do some testing on my desktop at home | 12:16 |
| eagles051387|| | my linux drive | 12:16 |
| *** aculich has quit IRC | 12:17 | |
| manish | any one have idea about FlatManger in my libvirt.xml no interface tag http://paste.openstack.org/show/4184/ | 12:17 |
| *** bobi111 has joined #openstack | 12:18 | |
| manish | in my libvirt.xml there is no interface tag | 12:19 |
| *** alekibango_ has quit IRC | 12:20 | |
| *** alekibango_ has joined #openstack | 12:20 | |
| *** mjfork has joined #openstack | 12:21 | |
| *** flaviamissi has joined #openstack | 12:21 | |
| *** rgidron2 has quit IRC | 12:29 | |
| *** robbiew has quit IRC | 12:31 | |
| *** hugo_kuo has joined #openstack | 12:36 | |
| *** E has joined #openstack | 12:37 | |
| *** E is now known as Guest14626 | 12:38 | |
| *** alekibango_ has quit IRC | 12:38 | |
| *** praefect has joined #openstack | 12:39 | |
| uksysadmin | so, ec2 + keystone - what do all the environment variables look like - what does a 'novarc' file look like when communicating with keystone? | 12:39 |
| *** aculich has joined #openstack | 12:40 | |
| *** hugo_kuo has quit IRC | 12:40 | |
| Kiall | uksysadmin: I have an example here: https://github.com/managedit/openstack-setup/blob/master/openrc.tmpl | 12:42 |
| Kiall | that works for both the euca and native client tools.. | 12:42 |
| uksysadmin | Kiall, awesome, thanks. | 12:42 |
| Guest14626 | someone knows about the Atlas-Load-balancer? | 12:42 |
| Kiall | uksysadmin: Well .. Mostly works.. you cant publish images via the euca/cloud/ec2 utils when keystone is used... | 12:43 |
| tightwork | anyone check out opendedup? | 12:43 |
| uksysadmin | Kiall, I'll give it a whirl | 12:43 |
| *** uncleofthestick has joined #openstack | 12:44 | |
| *** stewart has quit IRC | 12:44 | |
| *** Guest14626 has quit IRC | 12:48 | |
| *** PeteDaGuru has joined #openstack | 12:50 | |
| *** nevrax has joined #openstack | 12:51 | |
| lxu | mjforl | 12:52 |
| mjfork | yes | 12:52 |
| *** alekibango has joined #openstack | 12:53 | |
| *** mrevell has joined #openstack | 12:58 | |
| *** MrHeat has joined #openstack | 12:59 | |
| *** jedi4ever has quit IRC | 13:00 | |
| uksysadmin | does anyone know if any user authentication of any kind happens with nova? | 13:00 |
| longhudou | mjfork: | 13:01 |
| uksysadmin | I've just set my username, password, tenant, etc up as "fake", "fake", "fake" and the only thing it did was complain that my chosen keypair didn't exist (so I created one using the fake user and it accepted it) and running only after I launched the instance did it complain about no network available in the logs - the client happily thinks its ok | 13:02 |
| Kiall | uksysadmin: It certainly should complain! | 13:03 |
| Kiall | What command were you running, I'll double check if I get the same error.. | 13:03 |
| *** mrevell has joined #openstack | 13:04 | |
| uksysadmin | indeed it should - but my euca-describe-instances has: pendingfake (fake, None) | 13:04 |
| uksysadmin | actually its errorfake (fake, openstack1) where openstack1 is my host | 13:04 |
| Kiall | Okay, can you pastebin your nova.conf and your api-paste file? One of those is bound to be incorrect | 13:05 |
| uksysadmin | sure | 13:05 |
| Kiall | with invalid details, I can't even do a euca-describe-images | 13:05 |
| *** aculich has quit IRC | 13:06 | |
| *** bobi111 has quit IRC | 13:08 | |
| *** MrHeat has quit IRC | 13:09 | |
| uksysadmin | Kiall, http://paste.openstack.org/show/4198/ | 13:12 |
| Kiall | Yea.. You have auth disabled in the api-paste | 13:13 |
| *** Arminder has joined #openstack | 13:13 | |
| Kiall | pipeline = ec2faultwrap logrequest *ec2noauth* cloudrequest authorizer ec2executor | 13:13 |
| uksysadmin | haha, lol | 13:13 |
| *** zul has joined #openstack | 13:13 | |
| Kiall | uksysadmin: https://github.com/managedit/openstack-setup/blob/master/api-paste-keystone.ini.tmpl | 13:13 |
| Kiall | Just the settings at the very end should need changing for nova+keystone :) | 13:13 |
| *** aliguori has joined #openstack | 13:14 | |
| uksysadmin | I don't think a LOL describes how much I'm laughing at myself here | 13:14 |
| Kiall | ;) | 13:15 |
| *** dolphm has joined #openstack | 13:17 | |
| *** manish has quit IRC | 13:18 | |
| *** aculich has joined #openstack | 13:23 | |
| uksysadmin | I use your api-paste.ini and I get my old friend back: ImportError: No module named metadatarequesthandler | 13:24 |
| Kiall | uksysadmin: you using essex by any chance? | 13:24 |
| *** Birk_ has joined #openstack | 13:25 | |
| Kiall | The metadata service was changed a fair bit in essex, I would imagine the api paste for it changed too | 13:25 |
| uksysadmin | aye | 13:25 |
| uksysadmin | I'll try mangling the the two and remove the glaring error ;-) | 13:26 |
| Kiall | Add L25 -> 29 of your pastebin back in | 13:26 |
| Kiall | then remove L62 -> 66 from mine | 13:26 |
| Kiall | Actually - looks like more than that has changed | 13:26 |
| Kiall | Much more ;) | 13:26 |
| uksysadmin | indeed | 13:27 |
| *** BuZZ-T has quit IRC | 13:28 | |
| *** jfluhmann has quit IRC | 13:28 | |
| *** BuZZ-T has joined #openstack | 13:28 | |
| *** robbiew has joined #openstack | 13:29 | |
| *** stewart has joined #openstack | 13:29 | |
| *** yshh has joined #openstack | 13:37 | |
| *** rustam has quit IRC | 13:38 | |
| *** andreas__ has joined #openstack | 13:43 | |
| *** ahasenack has quit IRC | 13:43 | |
| *** dprince has joined #openstack | 13:44 | |
| *** 94SAAECOX has joined #openstack | 13:45 | |
| *** shaon has quit IRC | 13:50 | |
| heyho | kiall uksysadmin I reconfigured my private networks (nova-manage network...) and now i can ping the instance private ip and even ssh into the instance.. But I'm still unable to ping or ssh the floating ip I associate to the instance | 13:50 |
| Kiall | heyho: chances are you forgot to enable ping and SSH in the firewall rules.. | 13:51 |
| Kiall | the `euca-authorize` command handles that stuff.. | 13:51 |
| heyho | Kiall: I added the rules via the dashboard the tcp port 22.. as for icmp I may have made a syntax error | 13:51 |
| Kiall | imcp should be -1 / -1 usually (First box = ICMP Type, Second box = ICMP Code) | 13:52 |
| Kiall | -1 means "any" | 13:52 |
| Kiall | Try that and see if you can ping.. If you can ping, but cant SSH, the instance likely failed to boot.. Most often getting this far, but failing SSH means the instance could not reach the metadata service part of nova-api | 13:53 |
| Kiall | the console output would show if thats the case | 13:53 |
| heyho | look what i did http://i41.tinypic.com/1268k7o.png | 13:54 |
| Kiall | yea - 0 / 0 means only ping replys can go to the instance... | 13:54 |
| Kiall | if you added 8 / 0 aswell it would let ping requests through.. (or -1 / -1 for all) | 13:54 |
| *** hadrian has joined #openstack | 13:55 | |
| *** PiotrSikora has quit IRC | 13:55 | |
| heyho | Kiall.. yes thank you! I changed it from -1 -1 to 0 0 when in my previous attemps.. now ping works :) | 13:56 |
| *** PiotrSikora has joined #openstack | 13:58 | |
| *** Shentonfreude has joined #openstack | 14:00 | |
| heyho | I can also ssh.. the remaining problem now is the novnc console... If I try to connect using chrome I receive "server disconnected" error | 14:01 |
| *** hggdh has joined #openstack | 14:02 | |
| Kiall | Yea, Its not compatible with the latest versions of chrome.. | 14:04 |
| Kiall | The websockets spec is (was) too unstable and implementations kept changing.. | 14:04 |
| Madkiss | intersting. Fresh Ubuntu 11.10 installation, after trying to start nova, I see all nova components crashing after some seconds, but there is not a single line about it in any of the logs | 14:07 |
| *** paulmillar has quit IRC | 14:09 | |
| *** Leander has joined #openstack | 14:09 | |
| Leander | hello all | 14:09 |
| *** paulmillar has joined #openstack | 14:09 | |
| *** mattray has joined #openstack | 14:09 | |
| Leander | i'm having trouble terminating an instance | 14:09 |
| Leander | here's what happend, i stopped the instances | 14:10 |
| Leander | and forgot to terminate it before removing the nova-compute component from the controller | 14:10 |
| *** paulmillar has quit IRC | 14:10 | |
| Leander | the problem now is that i can't terminate that instance | 14:10 |
| Kiall | Leander: easy way, install nova-compute again for a few mins.. | 14:11 |
| Kiall | Hard way, use `virsh` to kill+remove the VM from libvirt then `rm -rf /var/lib/nova/instance/instance-ID`, then mark the instance as terminated in the DB manually... | 14:11 |
| Leander | to remove the compute node from the service list | 14:12 |
| Leander | do have to use a sql command? | 14:13 |
| Leander | ok that did it | 14:14 |
| Leander | thanks | 14:14 |
| Kiall | Yea, You have to edit the DB by hand for that.. | 14:15 |
| heyho | Kiall: you were right! IE9 worked.. now, if I want to access to the instance via vnc I have to setup a cloudpipe server? | 14:16 |
| *** Shentonfreude has left #openstack | 14:16 | |
| Kiall | huh? | 14:17 |
| Kiall | you mean via VPN? | 14:17 |
| *** rustam has joined #openstack | 14:18 | |
| Kiall | heyho: It sounds like you are using keystone, which I don't believe is compatible with cloudpipe in diablo.. (I've never tried, Only heard reports of them being incompatible) | 14:18 |
| heyho | Kiall: yes I'm using keystone..so as for now I can't access the instance via the vnc console (I dont know login and password that are promped) and I cant setup cloudpipe for imcompatibility..right? | 14:20 |
| Kiall | Oh, No .. cloudpipe has nothing to do with VNC | 14:21 |
| *** eagles051387|| has quit IRC | 14:21 | |
| Kiall | If you want to login via VNC, then you just need to assign the root (or another) user a password | 14:21 |
| Kiall | you can do that via SSH | 14:21 |
| Kiall | cloudpipe is for VPN, not VNC | 14:21 |
| heyho | Kiall: thank you for the explanation | 14:22 |
| *** Leander has quit IRC | 14:25 | |
| *** lborda has joined #openstack | 14:28 | |
| *** mrjazzcat has joined #openstack | 14:32 | |
| *** livemoon has quit IRC | 14:35 | |
| Madkiss | hm | 14:36 |
| Madkiss | can sombody supple me with an api-paste.ini for Oneiric that supports keystone? uksysadmin, maybe? | 14:36 |
| uksysadmin | if uksysadmin had one that worked I would gladly share it | 14:36 |
| uksysadmin | ;-) | 14:36 |
| Madkiss | good to hear :) | 14:36 |
| mjfork | Madkiss: paste yuor existing | 14:36 |
| mjfork | i will check it out | 14:36 |
| *** HugoKuo_ has joined #openstack | 14:37 | |
| Madkiss | well, the error message I get is ImportError: No module named nova_keystone_context | 14:37 |
| Madkiss | must be from | 14:37 |
| Madkiss | paste.filter_factory = keystone.middleware.nova_keystone_context:NovaKeystoneContext.factory | 14:37 |
| Madkiss | i guess ;) | 14:37 |
| mjfork | ok | 14:37 |
| *** berendt has joined #openstack | 14:37 | |
| mjfork | Madkiss: install from package? | 14:37 |
| *** tdi_ has joined #openstack | 14:37 | |
| Madkiss | yes. still. | 14:37 |
| *** rustam_ has joined #openstack | 14:38 | |
| Kiall | Madkiss: this works for the diablo version.. https://github.com/managedit/openstack-setup/blob/master/api-paste-keystone.ini.tmpl | 14:38 |
| berendt | is it a normal behavior that swift-proxy processes run a couple of minutes on 100% cpu usage after using swift-init proxy stop/start ? | 14:38 |
| uksysadmin | I'm using this: https://github.com/openstack/keystone/blob/milestone-proposed/doc/source/nova-api-paste.rst | 14:39 |
| uksysadmin | but fails on the paste.app_factory = nova.api.ec2.metadatarequesthandler:MetadataRequestHandler.factory | 14:39 |
| *** jeblair_ has joined #openstack | 14:39 | |
| *** jstraw has joined #openstack | 14:39 | |
| uksysadmin | there is no metadatarequesthandler class | 14:39 |
| *** mnemonic_ has joined #openstack | 14:39 | |
| Madkiss | Kiall: produces the same error message, actually ;) | 14:39 |
| Madkiss | oh, no | 14:39 |
| Madkiss | not quite. | 14:39 |
| *** nijaba_afk has joined #openstack | 14:40 | |
| *** jamespage has quit IRC | 14:40 | |
| *** markwash has quit IRC | 14:40 | |
| *** Lumiere has quit IRC | 14:40 | |
| *** jeblair has quit IRC | 14:40 | |
| *** tdi has quit IRC | 14:40 | |
| *** ches has quit IRC | 14:40 | |
| *** paltman has quit IRC | 14:40 | |
| *** mnemonic has quit IRC | 14:40 | |
| *** heydrick has quit IRC | 14:40 | |
| *** hugokuo has quit IRC | 14:40 | |
| *** ChrisAM has quit IRC | 14:40 | |
| *** nijaba has quit IRC | 14:40 | |
| *** evxd has quit IRC | 14:40 | |
| *** statik has quit IRC | 14:40 | |
| *** jstraw is now known as Lumiere | 14:40 | |
| *** mrjazzcat has quit IRC | 14:40 | |
| *** andreas__ has quit IRC | 14:40 | |
| Madkiss | root@netbook:/etc/nova# dpkg -L keystone | grep ec2 | 14:40 |
| Madkiss | root@netbook:/etc/nova# | 14:40 |
| *** andreas__ has joined #openstack | 14:40 | |
| *** rustam has quit IRC | 14:40 | |
| *** exekias has quit IRC | 14:40 | |
| Madkiss | which consequently results in ImportError: No module named ec2_token | 14:40 |
| *** exekias has joined #openstack | 14:40 | |
| *** kloeri has quit IRC | 14:41 | |
| *** markwash has joined #openstack | 14:41 | |
| Kiall | Madkiss: .. ah .. you are using the stock ubuntu packages for keystone? | 14:41 |
| mjfork | Madkiss: what version of packages | 14:41 |
| *** kloeri has joined #openstack | 14:41 | |
| Kiall | They don;t work. | 14:41 |
| Madkiss | Kiall: keystone.middleware.ec2_token is present in 2012.1 only. | 14:41 |
| *** nijaba_afk is now known as nijaba | 14:41 | |
| *** paltman has joined #openstack | 14:41 | |
| *** mrjazzcat has joined #openstack | 14:41 | |
| Madkiss | that's good to hear, which will I have to use then? | 14:41 |
| *** ChrisAM1 has joined #openstack | 14:41 | |
| *** nijaba has quit IRC | 14:41 | |
| *** nijaba has joined #openstack | 14:41 | |
| Kiall | Madkiss: yea.. the ubuntu packages for keystone and dashboard are broken, not diablo itself ;) | 14:41 |
| *** evxd has joined #openstack | 14:41 | |
| Madkiss | (they are from ubuntu oneiric-proposed) | 14:41 |
| Kiall | I've got fixed packages here: launchpad.net/~managedit/+archive/openstack | 14:41 |
| Madkiss | Kiall: r0x0r. | 14:42 |
| mjfork | Kiall: are those diabo or essex? | 14:42 |
| Kiall | diablo | 14:42 |
| mjfork | Madkiss: you want essex right? | 14:42 |
| Madkiss | no | 14:42 |
| Madkiss | I want diablo for now. | 14:42 |
| Madkiss | after all the moan caused by essex. | 14:42 |
| Madkiss | this is all making me nuts. nothing's properly documented, and findign something _always_ is diggin' in the dark ;) | 14:43 |
| *** praefect has quit IRC | 14:43 | |
| *** heydrick has joined #openstack | 14:43 | |
| Kiall | BTW - Maybe the oneiric-proposed packages have been fixed, and there is something else wrong.. I haven't looked at those, and I know the ubuntu guys are planning on fixing the packages eventually | 14:43 |
| Madkiss | there is only nova in there anyway. | 14:43 |
| uksysadmin | I was running diablo from proposed and it caused problems - in fact, I was getting that infernal metadatarequesthandler issues so I went to essex from milestones and still no joy then went to trunk and still no joy | 14:44 |
| *** dendrobates is now known as dendro-afk | 14:44 | |
| mjfork | Madkiss: so why the 2012.1 packageS? | 14:44 |
| uksysadmin | so there's a mismatch between keystone trunk and packages | 14:44 |
| Madkiss | uksysadmin: that principle just doesn't work with openstack right now apparently. | 14:44 |
| Madkiss | mjfork: i am just and simply trying to get something that works. really. I swear. | 14:44 |
| Kiall | Madkiss: .. lol clone git://github.com/managedit/openstack-setup .. edit settings file .. type ./all-in-one.sh .. enjoy ;) | 14:45 |
| Madkiss | that's gonna caused exactly the same effects once I want to update something. | 14:45 |
| Madkiss | no really. | 14:45 |
| *** RainbowDash36 has joined #openstack | 14:45 | |
| Madkiss | I know I can get stuff to work by building it by hand, i've been a debian developer for 9 years, but that's not my goal :) | 14:45 |
| *** lborda has quit IRC | 14:46 | |
| *** RainbowDash36 has quit IRC | 14:46 | |
| *** kloeri has quit IRC | 14:46 | |
| Kiall | Madkiss: sadly, there are no official openstack or ubuntu packages for diablo that work.. | 14:46 |
| *** kloeri has joined #openstack | 14:46 | |
| *** jfluhmann has joined #openstack | 14:46 | |
| Kiall | And essex is in development, I wouldn't be counting on it to be bug free and reliable ;) | 14:46 |
| Madkiss | Kiall: If yours work, i'm perfectly fine with that. I am mainly looking for stuff I can use for demonstrational purposes. | 14:46 |
| uksysadmin | Kiall, diablo isn't exactly bug free and reliable neither | 14:47 |
| Madkiss | But telling people to "just clone that git repostiry from there" during a workshop is not gonna work ;) | 14:47 |
| *** cereal_bars has joined #openstack | 14:47 | |
| Kiall | uksysadmin: that might be true of the 2011.3 tag.. but not of the stable/diablo branch that I build my packages from.. | 14:47 |
| *** ChrisAM1 is now known as ChrisAM | 14:47 | |
| Kiall | There have been hundreds of bugfixes that the ubuntu packages dont have... | 14:48 |
| uksysadmin | Madkiss, I feel your pain. Best you can do is reach out to the mailing list. | 14:48 |
| *** andreas__ has quit IRC | 14:48 | |
| uksysadmin | lots of developers are state side and will be online later on in the day | 14:48 |
| Kiall | Madkiss: the packages are in a PPA ;) | 14:48 |
| Madkiss | the oneiric-proposed-ones? | 14:48 |
| Kiall | anyway .. as I said... Sadly, there are no official openstack or ubuntu packages for diablo that work.. | 14:48 |
| mjfork | madkiss, give me a few. looks like you have mismatched packages. | 14:48 |
| mjfork | you can use the RCB pacakges, or install nova from PPA and keystone/dashbaord from source. | 14:49 |
| Kiall | well - no keystone or dashboard packages that work, and the ubuntu nova etc is minus a few hundred bug fixes* | 14:49 |
| Madkiss | Give me some minutes to sort all this stuff out | 14:50 |
| Madkiss | well. | 14:50 |
| Madkiss | Invalid command 'WSGIScriptAlias', perhaps misspelled or defined by a module not included in the server configuration | 14:50 |
| Madkiss | ;) | 14:50 |
| *** lborda has joined #openstack | 14:51 | |
| *** kloeri has quit IRC | 14:51 | |
| Madkiss | funny, dpkg --configure -a made it all good again. | 14:51 |
| Madkiss | oh well. | 14:51 |
| *** kloeri has joined #openstack | 14:51 | |
| *** mrevell has quit IRC | 14:54 | |
| *** longhudou has quit IRC | 14:54 | |
| *** rods has joined #openstack | 14:55 | |
| Madkiss | Kiall: okay, these start up like a charm at least | 14:56 |
| *** dwcramer has joined #openstack | 14:56 | |
| Madkiss | send: 'POST /v2.0/tokens HTTP/1.1\r\nHost: 10.42.0.6:5000\r\nContent-Length: 102\r\ncontent-type: application/json\r\naccept-encoding: gzip, deflate\r\nuser-agent: python-novaclient\r\n\r\n{"auth": {"tenantName": "admin", "passwordCredentials": {"username": "admin", "password": "hastexo"}}}' | 14:56 |
| Madkiss | reply: 'HTTP/1.1 500 Internal Server Error\r\n' | 14:56 |
| Madkiss | ahum. | 14:56 |
| *** kloeri has quit IRC | 14:56 | |
| *** kloeri has joined #openstack | 14:57 | |
| Madkiss | OpenStack is like a tomato made of error messages. Wherever you bite into it, random error messages will float out | 14:57 |
| mjfork | Madkiss: what is in keystone logs? | 14:57 |
| *** livemoon has joined #openstack | 14:58 | |
| Kiall | Madkiss: lol... | 14:58 |
| *** mrevell has joined #openstack | 14:59 | |
| Madkiss | oh look! after stopping keystone, deleting the old keystone.db and re-creating it, it works! | 14:59 |
| *** mrevell has joined #openstack | 14:59 | |
| Madkiss | Kiall: am I using "deprecated auth" now? | 15:00 |
| Kiall | No, not if you are using keystone | 15:00 |
| Birk_ | I'm having some problems when trying to ssh to my instances. I'm getting this error. Permission denied (publickey). My instance is running on a local network and the nova-controller is the the gateway to the external network. | 15:01 |
| Madkiss | Kiall: just wondering because keystone logs this into /var/log/keystone/keystone-legacy-auth.log ... | 15:01 |
| Kiall | that git repo I linked has working/example configs + setup scripts for everything you can use as a reference... (for diablo..) | 15:01 |
| *** robbiew has quit IRC | 15:02 | |
| Kiall | weird, my keystone-legacy-auth.log file is basically empty... | 15:02 |
| *** robbiew has joined #openstack | 15:02 | |
| *** Ryan_Lane has joined #openstack | 15:05 | |
| *** nevrax has quit IRC | 15:05 | |
| Madkiss | well, this seems to work like a charm, great! | 15:05 |
| Kiall | Madkiss: its actually quite easy to get working once you dont have to install from source, or broken packages ;) | 15:06 |
| *** PotHix has joined #openstack | 15:07 | |
| heyho | Guys, How can I force the removal of an old instance stuck in build state? I may log to mysql and do a query like "remove * from instances where keyname="oldkey" ...could it work? | 15:07 |
| heyho | I only see the instance in the dashboard.. is not even in /var/lib/nova/instances.. | 15:08 |
| mjfork | heyho: try nova-manage or delete it directly in teh DB? | 15:08 |
| Kiall | You can set terminated = 1, and terminated_date to the current time.. | 15:08 |
| mjfork | there ya go ^ | 15:08 |
| *** arun has quit IRC | 15:08 | |
| Kiall | Thats basically what nova would do if you still had nova-compute on that server | 15:09 |
| Kiall | well .. that and remove it from libvirt and /var/lib/nova/instances aswell.. | 15:09 |
| heyho | there's no instance folder in /var/lib/nova/instances .. and where do I ser terminated = 1 and terminanted_date .. I didnt understand sorry | 15:10 |
| Kiall | in the DB `instances` table.. | 15:10 |
| *** rnorwood has joined #openstack | 15:10 | |
| heyho | so, like my previous query but instead of keyname parameter using the terminated one? | 15:11 |
| Madkiss | AttributeError: 'NoneType' object has no attribute 'find' | 15:11 |
| Madkiss | hm | 15:11 |
| Kiall | heyho: yea.. basically :) | 15:12 |
| heyho | ok thanks Kiall | 15:12 |
| *** judd7 has joined #openstack | 15:12 | |
| *** jamespage has joined #openstack | 15:13 | |
| Madkiss | Kiall: http://paste.debian.net/151691/ -- this has something to do with keystone, i guess? | 15:14 |
| Kiall | Yea - I dont use swift, but it looks like the "keystone_url" variable isn't set.. | 15:15 |
| *** ben_duyujie has joined #openstack | 15:15 | |
| Madkiss | keystone_url or KEYSTONE_URL? | 15:17 |
| *** endzYme_ has joined #openstack | 15:17 | |
| Birk_ | mjfork: Hey mjfork, can you try to help me with some details ? I'm trying a diferente installation now. Im using this architecture http://imageshack.us/photo/my-images/839/architecture.png/ . The yellow circle represents the nova controller and the gateway between the local and external network. The controller machine have 2 network interfaces. | 15:18 |
| Madkiss | ah | 15:18 |
| Madkiss | nevermind | 15:18 |
| *** saju_m has quit IRC | 15:18 | |
| Madkiss | I see | 15:18 |
| Madkiss | What do service_host, service_port and service_pass actually do? They are described in every howto, but they are not actually explained | 15:19 |
| Kiall | lol | 15:19 |
| Kiall | I'm guessing thats the keystone host, port and a long life token | 15:19 |
| Kiall | But - As I said, I don't use swift.. I don't need it... | 15:20 |
| Birk_ | mjfork: i can create new instances and ping them from the controller server, but when i try to ssh the instances (all instances run just on the compute server on local network) i got this error: Permission denied (publickey). I've already tried to generate the keys again, but have no success. | 15:20 |
| Madkiss | hear, hear, swift started. | 15:21 |
| *** rods has quit IRC | 15:21 | |
| Birk_ | mjfork: it can be the network interfaces configuration ? | 15:21 |
| mjfork | Birk_: look at the console for the VMs to see if you see the SSH denial | 15:21 |
| mjfork | Birk_: wher are you trying to SSH from? try it from ther controller to the VM | 15:21 |
| Spirilis | hey so can anyone comment on NASA's Nebula project... is it based on openstack? or did openstack grow out of nebula code? | 15:22 |
| Birk_ | mjfork: yes, i'm trying from the controller to the VM, but i got that message. | 15:22 |
| Spirilis | (writing notes for my presentation this afternoon :D) | 15:22 |
| mjfork | Spirilis: i understand it as the nebula code base seeded openstack? | 15:22 |
| Spirilis | yeah that's what I'm reading... cool | 15:22 |
| *** rods has joined #openstack | 15:22 | |
| mjfork | Sp "NASA will contribute technology used to power its Nebula Cloud Platform." | 15:23 |
| uksysadmin | yeah my interpretation of Nebula too | 15:23 |
| *** livemoon has left #openstack | 15:23 | |
| Spirilis | sounds good on a note sheet :) | 15:23 |
| *** jfluhmann has quit IRC | 15:23 | |
| Madkiss | to whom are you presenting, Spirilis? | 15:25 |
| Spirilis | fellow sysadmin and IT folks, a couple apps folks too | 15:26 |
| Spirilis | my POC efforts thus far have been in a vacuum that only me and my boss know about. so here's the big "coming out" | 15:26 |
| Kiall | Spirilis: My understanding is that nova started live as nasa's nebula, and swift started life as rackspace cloud files.. and that both companies are using (or planning on using) OpenStack in production... | 15:28 |
| Spirilis | ah interesting. | 15:29 |
| Birk_ | mjfork: I've tried to catch some errors in the vm console with euca-get-console-output after the ssh attempt, but i didn't find anything. | 15:29 |
| Madkiss | hmpf. | 15:30 |
| mjfork | Birk_: can you VNC in and looka at /var/log/messages | 15:30 |
| Madkiss | http://paste.debian.net/151693/ | 15:30 |
| Madkiss | can I make swift more verbose anyhow? | 15:30 |
| mjfork | Kiall is right about Swift - "Rackspace is donating the code that powers its files and servers, the foundation for its public-cloud offerings." | 15:30 |
| mjfork | Madkiss: look like URL failed to parse and is not an object | 15:31 |
| mjfork | guessing that keystone isn't returning the storage URL? | 15:31 |
| mjfork | are you trying swift + keystone on diablo? | 15:31 |
| notmyname | Spirilis: ya, swift is running in production (and always has) at rackspace as our cloud files product. we run the latest version of swift, and have (to my knowledge) the largest swift clusters in the world | 15:31 |
| Madkiss | mjfork: yes. | 15:31 |
| mjfork | Madkiss: hmm, i was thinking the keystone middleware was broken there. was that fixed? | 15:31 |
| *** imsplitbit has joined #openstack | 15:31 | |
| Madkiss | Kiall: is it in your version? | 15:31 |
| *** endzYme has quit IRC | 15:32 | |
| Kiall | Madkiss: what now? | 15:32 |
| Madkiss | Kiall: keystone middleware for swift appears to be broken in the Diablo packages; is it fixed in yours (if you know this, by any chance) | 15:32 |
| Kiall | Its likely not, Since I dont use swift :) | 15:33 |
| Madkiss | gna | 15:33 |
| Madkiss | mjfork: How can I see what swift is actually *doing*= | 15:33 |
| Madkiss | -v doesn't make it any more verbose, actually | 15:33 |
| Spirilis | notmyname: very cool... thanks for the info! | 15:33 |
| uksysadmin | Madkiss, I've a nice working keystone + swift set up from packages | 15:35 |
| Madkiss | Oneiric default packages? | 15:35 |
| Madkiss | (could I see your proxy-server.conf please?) | 15:36 |
| Birk_ | mjfork: how can i VNC to the machine that is on the local network ? because i don't have graphic interface in the controller server. | 15:36 |
| Spirilis | does anyone know if NASA's Nebula stands separate from openstack, or is it mostly openstack at this point, or openstack with cutting edge extras that they're keeping close to the chest or whatever? | 15:36 |
| uksysadmin | no - ppa:swift-core/trunk | 15:36 |
| uksysadmin | ppa:keystone-core/trunk | 15:36 |
| mjfork | Spirilis: i don't know that i have seen an annoucement, but i would expect the Nasa is building off/using OpenSTack | 15:36 |
| mjfork | Spirilis: i am sure the ponit of open sourcing it was to save development/support costs | 15:37 |
| Spirilis | yeah, agreed | 15:37 |
| mjfork | and nasa doesn't have much reason to keep anything in house | 15:37 |
| *** Xenith has quit IRC | 15:37 | |
| uksysadmin | Madkiss, http://paste.openstack.org/show/4201/ | 15:37 |
| *** Xenith has joined #openstack | 15:37 | |
| uksysadmin | notmyname, what is the service_pass flag in swift proxy-server.conf? | 15:38 |
| notmyname | uksysadmin: in the keystone section? I have no idea. that's the realm of keystone | 15:39 |
| *** dillon-w has joined #openstack | 15:39 | |
| *** Xenith has quit IRC | 15:40 | |
| uksysadmin | np... just I didn't set a password if that's what it is so assuming its for "future use" | 15:40 |
| uksysadmin | yet all examples of this have this set | 15:40 |
| *** Xenith has joined #openstack | 15:40 | |
| btorch | uksysadmin: I never used that setting on my tests (keystone services access password) ... although I kept my distance from "egg:keystone#swiftauth" and " egg:keystone#tokenauth" and instead used " egg:swiftkeystone2#keystone2" | 15:43 |
| *** zzed has joined #openstack | 15:43 | |
| uksysadmin | btorch, what's the difference? what does keystone2 provide that tokenauth doesn't? | 15:43 |
| btorch | uksysadmin: Last time I checked tokenauth had a pretty bad bug and should not be used | 15:44 |
| dillon-w | i've got some questions about multizone setup, anyone care to share some experience? | 15:44 |
| uksysadmin | btorch, thanks - and I was sitting here quietly smug that I have a nice swift + keystone test setup! ;-) | 15:44 |
| btorch | uksysadmin: if I remember correctly it allowed a valid token to talk to any account | 15:44 |
| Madkiss | mjfork: do you have any idea what is supposedly "broken"? | 15:45 |
| uksysadmin | ah - that would make sense and highlight the big error message on their doc page regarding that - but didn't know why | 15:45 |
| uksysadmin | (did I say big - there's a mention of it) | 15:45 |
| *** ben_duyujie has quit IRC | 15:46 | |
| mjfork | i know the original, rleeased PPA packages where broken | 15:46 |
| mjfork | then i was thinking that swift + keystone was still broken, check the answers section of launchpad | 15:46 |
| *** mgoldmann has quit IRC | 15:47 | |
| *** tmichael has joined #openstack | 15:48 | |
| notmyname | the proper middleware to use with swift for keystone is swift_auth (at least it used to be) | 15:48 |
| notmyname | token auth has/had the issue that btorch mentioned | 15:49 |
| *** Xenith has quit IRC | 15:49 | |
| btorch | uksysadmin: it could have changed ... last time a checked was last year but either way I remember sometime ago some someone telling me to use swiftauth instead but never got that working (didn't feel like wasting time either) so I used chmouel_'s keystone2 filter | 15:49 |
| *** Xenith has joined #openstack | 15:50 | |
| Madkiss | uksysadmin: you are running today's trunk of keystoneà according to adam_g, something is broken there ... | 15:51 |
| *** robbiew has quit IRC | 15:51 | |
| *** krish has joined #openstack | 15:54 | |
| *** ben_duyujie has joined #openstack | 15:55 | |
| *** rustam_ has quit IRC | 15:55 | |
| *** dragondm has joined #openstack | 15:55 | |
| *** adjohn has joined #openstack | 15:55 | |
| *** rustam has joined #openstack | 15:56 | |
| *** mrevell has quit IRC | 15:56 | |
| *** ppradhan has joined #openstack | 15:57 | |
| Xenith | Hmm. Is the keystone package still broken in ubuntu 11.10? | 15:59 |
| *** mrevell has joined #openstack | 16:01 | |
| *** antenagora has quit IRC | 16:03 | |
| *** pbacterio has joined #openstack | 16:05 | |
| *** ldlework has joined #openstack | 16:06 | |
| *** panterax has joined #openstack | 16:07 | |
| *** joesavak has joined #openstack | 16:07 | |
| *** pbacterio has quit IRC | 16:08 | |
| *** rods has quit IRC | 16:10 | |
| *** reed_ has quit IRC | 16:11 | |
| *** rods has joined #openstack | 16:11 | |
| *** andreas__ has joined #openstack | 16:12 | |
| *** reidrac has quit IRC | 16:12 | |
| *** guigui has quit IRC | 16:13 | |
| sante_ | Hi all. I'm trying to build an openstack lab with XenServer and Essex-2. I did some test on Xenserver6 and Xenserver 5.6. On both version I'm unable to resize a virtual machine more than once. On second resize I receive errors on both hypervisor's version | 16:13 |
| *** deshantm has joined #openstack | 16:13 | |
| *** freeflyi1g has quit IRC | 16:14 | |
| *** rods has quit IRC | 16:14 | |
| *** robbiew has joined #openstack | 16:14 | |
| *** rods has joined #openstack | 16:15 | |
| *** freeflying has joined #openstack | 16:15 | |
| sante_ | to let you figure out the type of problem, on 5.6, on second resize, when the vm has been copyed on destination host (xensource.log) I receive an error like this: http://paste.openstack.org/show/4108/ | 16:15 |
| sante_ | is there some xen people that can address me to a solution? | 16:16 |
| deshantm | sante_: you want to try to ask in ##xen and #xen-api | 16:18 |
| *** mindpixel has quit IRC | 16:18 | |
| dillon-w | if i'd like to setup two zones, one parent + one child, what is least keystone/glance service i should have? | 16:19 |
| *** cp16net has joined #openstack | 16:22 | |
| *** Hakon|mbp has quit IRC | 16:23 | |
| *** Hakon|mbp has joined #openstack | 16:23 | |
| uksysadmin | Madkiss, I did an update earlier on - seems fine but I've not really done anything apart from curl some containers and view them, etc. | 16:29 |
| *** localadmin__ has joined #openstack | 16:30 | |
| *** bepernoot has quit IRC | 16:30 | |
| localadmin__ | Has anyone come across this nova-network error: ERROR: trying to add VLAN #100 to IF -:br100:- error: No such device | 16:32 |
| *** Remco_ has quit IRC | 16:32 | |
| *** waytogobastard has quit IRC | 16:34 | |
| *** yshh has quit IRC | 16:35 | |
| *** ben_duyujie has quit IRC | 16:36 | |
| *** ben_duyujie has joined #openstack | 16:36 | |
| *** koolhead17 has joined #openstack | 16:37 | |
| uksysadmin | localadmin__, I have done numerous times and I've forgotten how to get around it. | 16:37 |
| uksysadmin | essentially your bridging hasn't worked | 16:37 |
| *** jeblair_ is now known as jeblair | 16:38 | |
| *** nebuchadnezzar has joined #openstack | 16:39 | |
| nebuchadnezzar | hello | 16:39 |
| *** spiffxp has joined #openstack | 16:39 | |
| localadmin__ | uksysadmin: any clue as to what I should do? i did apt-get on vlan and it said I was up to notch | 16:40 |
| localadmin__ | how does one debug bridging issues? | 16:40 |
| *** dnjaramba has joined #openstack | 16:41 | |
| *** spackest has joined #openstack | 16:41 | |
| *** rods has quit IRC | 16:42 | |
| *** pixelbeat has joined #openstack | 16:43 | |
| *** rods has joined #openstack | 16:46 | |
| *** robbiew has quit IRC | 16:47 | |
| *** foexle has quit IRC | 16:49 | |
| *** maploin has quit IRC | 16:50 | |
| *** popux has quit IRC | 16:50 | |
| *** krow has joined #openstack | 16:50 | |
| *** zul has quit IRC | 16:51 | |
| uksysadmin | brctl show | 16:52 |
| uksysadmin | but this is controlled by nova-network so check out /var/log/nova-network.log | 16:52 |
| *** krow has quit IRC | 16:52 | |
| *** uksysadmin has quit IRC | 16:53 | |
| *** daysmen_ has joined #openstack | 16:54 | |
| *** zzed has quit IRC | 16:55 | |
| *** dolphm has quit IRC | 16:58 | |
| *** cereal_bars has quit IRC | 16:58 | |
| *** primeministerp has joined #openstack | 16:58 | |
| *** rnorwood1 has joined #openstack | 17:01 | |
| Birk_ | If i'm using two network interfaces in the nova controller I have to do anything different in the nova.conf file or i need just to configure the local network properly ? | 17:01 |
| *** adjohn has quit IRC | 17:01 | |
| *** 94SAAECOX has quit IRC | 17:03 | |
| *** heyho has quit IRC | 17:03 | |
| *** devananda has joined #openstack | 17:03 | |
| *** rnorwood has quit IRC | 17:03 | |
| *** rocambole has quit IRC | 17:03 | |
| *** ben_duyujie has quit IRC | 17:04 | |
| *** codatory has joined #openstack | 17:06 | |
| *** hggdh has quit IRC | 17:06 | |
| *** ben_duyujie has joined #openstack | 17:07 | |
| *** heckj has joined #openstack | 17:07 | |
| *** dwcramer has quit IRC | 17:08 | |
| *** hugokuo has joined #openstack | 17:09 | |
| *** cp16net has quit IRC | 17:10 | |
| hugokuo | anyone around ? | 17:10 |
| hugokuo | I'm confusing about nova-volume concept ..... | 17:10 |
| *** freeflying has quit IRC | 17:11 | |
| *** cp16net has joined #openstack | 17:11 | |
| hugokuo | a question , does the nova-volumes group must be located on same host with nova-volume service ? as I know , not necessary though | 17:11 |
| *** Ramonster has quit IRC | 17:11 | |
| *** sleepsonthefloo has joined #openstack | 17:11 | |
| mjfork | localadmin__: run "mysql -uroot -pnova -e "select bridge_interface from networks;" nova" and i bet the bridge_itnerface is br100 instead of ethX | 17:12 |
| mjfork | Birk_: you need to specify public ineteface | 17:13 |
| *** ben_duyujie has quit IRC | 17:13 | |
| Birk_ | mjfork: the public interface is the interface that connects to the external network or with the local ? | 17:14 |
| mjfork | Birk_: are you usign vlan | 17:14 |
| Birk_ | mjfork: I'm using flatDHCP | 17:15 |
| *** codatory has left #openstack | 17:15 | |
| dillon-w | hugokuo : working so late? | 17:15 |
| Birk_ | mjfork: in my nova.conf the pub. interface is eth0. Mine eth0 is connecting with the external network | 17:15 |
| hugokuo | just can not sleep well if I can not figure out this issue .... | 17:15 |
| mjfork | Birk_: what problem do you have | 17:16 |
| Birk_ | the flat_network is configured eth1. eth1 is connecting with the lan | 17:16 |
| hugokuo | that makes me so confuse ……. | 17:16 |
| mjfork | HugoKuo: are you using standard nova-voume driver against LVM? | 17:16 |
| hugokuo | as default …. I guess it's standard nova-volume manager | 17:16 |
| Birk_ | mjfork: My VM is running in a compute node that is located ins the local network. I'm trying to ssh this VM but i got Permission denied (publickey) | 17:17 |
| mjfork | HugoKuo then yes, run nova-volume on every machine you want to export storage from LVM | 17:17 |
| Birk_ | I've executed the sames steps as in my other configurations that are working | 17:17 |
| mjfork | Birk_: are you SSHing from outside the cloud? have your tiredit from compute node? | 17:17 |
| *** freeflying has joined #openstack | 17:17 | |
| hugokuo | mjfork , wow …..seems my consideration totally wrong .... | 17:18 |
| Birk_ | mjfork: I'm trying to ssh from the gateway (nova-controller) | 17:18 |
| Birk_ | mjfork: i'll try from the compute node in the lan now | 17:18 |
| *** hky has quit IRC | 17:18 | |
| mjfork | HugoKuo the nova-compute then mounts the iscsi export by discovering it | 17:18 |
| mjfork | HugoKuo doesn't work through nova-volume | 17:18 |
| hugokuo | So do ya means that nova-volume worker must run with all VGs those I want to export ? | 17:18 |
| hugokuo | a question about the flag --iscsi_ip_perfix …. | 17:19 |
| Birk_ | mjfork: from the compute node i got the same error. Permission denied (publickey). | 17:19 |
| mjfork | Birk_: how did you launch | 17:20 |
| hugokuo | what's the real usage ? I thought it's for specifying the address which has the sharing VG ….. am I right ? | 17:20 |
| *** maplebed has joined #openstack | 17:20 | |
| Birk_ | mjfork: do you men how did i created the VM ? | 17:20 |
| mjfork | HugoKuo my understanding is that iscsi prefix just scopes the search to the right range, some expoerts are on multiple IPs, you want the one for your network | 17:21 |
| mjfork | Birk_: yes | 17:21 |
| Birk_ | mjfork: with this: euca-run-instances ami-00000002 -k nuvemcpca -t m1.tiny. | 17:21 |
| Birk_ | mjfork: I'm trying to ssh with ssh -i nuvemcpca.priv ubuntu@10.0.0.2 | 17:22 |
| *** reed_ has joined #openstack | 17:22 | |
| *** dwcramer has joined #openstack | 17:22 | |
| mjfork | Birk_: do you have verbose on in the conf? | 17:22 |
| mjfork | HugoKuo don't need to specify the whole IP address, many machines can run nova-volume | 17:23 |
| dillon-w | hugokuo : just FYI, i encountered an issue when i didn't set iscsi_ip_prefix correctly. | 17:23 |
| mjfork | HugoKuo are you on diablo or exsex ? | 17:23 |
| *** dillon-w has quit IRC | 17:24 | |
| hugokuo | mjfork , nova essex trunk …. | 17:24 |
| hugokuo | dillon-w , thx | 17:24 |
| Birk_ | mjfork: yes, in nova.conf i have verbose=true | 17:24 |
| *** marcuz has quit IRC | 17:25 | |
| mjfork | HugoKuo i *think* the behavior may have changed in essex, will want to do some checking | 17:25 |
| mjfork | Birk_: can you recycle your logs and do a new provisioning and captuer the nova-compute log? | 17:25 |
| mjfork | paste bin that | 17:25 |
| Birk_ | ok | 17:25 |
| hugokuo | my nova-volume always detect local VG ….. if I separate the VG from nova-volume server :< ….. it totally failed by can't find VG named nova-volumes | 17:26 |
| *** Kyle__ has quit IRC | 17:26 | |
| mjfork | yeah, nova-volume runs on many services | 17:26 |
| mjfork | many SERVERS that is | 17:26 |
| mjfork | HugoKuo each server with a VG to export must run nova-volume | 17:27 |
| hugokuo | Really !!? | 17:28 |
| hugokuo | there's a description in flags doc | 17:28 |
| hugokuo | [no]use_local_volumes: if True, will not discover local volumes | 17:28 |
| hugokuo | (default: 'true') | 17:28 |
| hugokuo | if True will not discover local volumes….. | 17:29 |
| *** rnorwood1 has quit IRC | 17:29 | |
| *** Leander has joined #openstack | 17:29 | |
| Leander | hello once again | 17:29 |
| Leander | i'm having some trouble with accessing virtual machine instances on the network | 17:30 |
| *** rnorwood has joined #openstack | 17:30 | |
| *** maplebed has quit IRC | 17:30 | |
| Leander | i have two nodes, a controller and a compute | 17:30 |
| Leander | which reside on separate servers | 17:30 |
| mjfork | HugoKuo looking at the code, that just appears to use a local path to the volume if it can | 17:30 |
| mjfork | HugoKuo other wise it does discovery | 17:30 |
| *** LeanReox has joined #openstack | 17:31 | |
| *** llang629 has joined #openstack | 17:31 | |
| Leander | the instances are loaded and according to euca-describe-instances they appear to be running. I've checked the logs and can't find anything wrong. They have the appropriate permissions. | 17:31 |
| mjfork | HugoKuo so basically, if that flag is enabeld and the VM and the sotrage being mounted are on teh same machine, then use a local path | 17:31 |
| Leander | but i can't ping them nor ssh into themm | 17:31 |
| LeanReox | Hi guys, trying to curl to create a security group | 17:31 |
| LeanReox | getting bad request 400 | 17:32 |
| LeanReox | curl -X POST -d "{"security_group": {"name": "testSG", "description": "testSG"}}" "http://172.16.x.y:8774/v1.1/5/os-security-groups" -H "X-Auth-Token: 3dd60c26-5f1d-438a-9e26-a2ff86570196" | 17:32 |
| LeanReox | thats the curl | 17:32 |
| mjfork | Leander: what is the error message | 17:32 |
| LeanReox | is the JSON bad ? | 17:32 |
| hugokuo | mkfork , got it …. I should dig more from codes…. | 17:32 |
| mjfork | LeanReox: did you check the answers section of launch ad? | 17:32 |
| Leander | mjfork: there is no error message | 17:32 |
| mjfork | LeanReox: also pastebin the nova-api.log | 17:32 |
| mjfork | leander what does SSH do? | 17:32 |
| hugokuo | mjfork , appreciate your help , but I'm home now …. will check it tmr | 17:32 |
| LeanReox | mjfork , not seeing anything like it | 17:33 |
| Leander | mjfork: ssh: connect to host 10.1.1.3 port 22: No route to host | 17:33 |
| LeanReox | but cant be missing the header "application type json" ? | 17:33 |
| mjfork | Leander: did you euca-authorize | 17:33 |
| mjfork | Leanrox make sure verbose is on in the conf and pastebin the nova-api log after running | 17:34 |
| Birk_ | mjfork: follow http://pastebin.com/RPLFt9vZ | 17:34 |
| Leander | mjfork: how can i check that? | 17:34 |
| *** rods has quit IRC | 17:34 | |
| Leander | euca-describe-group? | 17:34 |
| *** warik has joined #openstack | 17:34 | |
| mjfork | Leander: http://docs.openstack.org/diablo/openstack-compute/admin/content/enabling-ping-and-ssh-on-vms.html | 17:34 |
| *** rnorwood has quit IRC | 17:34 | |
| *** rnorwood has joined #openstack | 17:34 | |
| Leander | mjfork: both rules already exist | 17:35 |
| mjfork | leander what image? is ssh running? | 17:35 |
| mjfork | Leander: can you VNC iN? | 17:35 |
| Birk_ | mjfork: I'm having the same problem with the new instance | 17:35 |
| Leander | mjfork: i haven't configured vnc and the image is ttylinux-uec-amd64-12.1_2.6.35-22_1 | 17:36 |
| LeanReox | mjfork im getting that that resource is specting 3 parameters and im sending 2 ... is pretty straighforward, | 17:36 |
| LeanReox | going to check a little bit more the curl with debugginf | 17:36 |
| *** andrewbogott has joined #openstack | 17:38 | |
| *** andrewbogott has quit IRC | 17:39 | |
| *** Rajaram has quit IRC | 17:39 | |
| *** andrewbogott has joined #openstack | 17:39 | |
| Leander | mjfork: when i set up a single node everything worked fine, after i separated nova-compute into a separate server i couldn't connect to my instances anymore | 17:39 |
| *** freeflying has quit IRC | 17:40 | |
| Birk_ | Leander: I'll try to help you. Do you have nova-controller in one machine and nova-compute in the other right ? | 17:41 |
| Leander | Birk_, that is correct | 17:41 |
| Birk_ | Leander: Ok, for me worked this way: I copied the nova.conf file from the controller and put it on the compute server. | 17:42 |
| mjfork | Birk_: paste bin the logs | 17:42 |
| Birk_ | Leander: i set up my network interface bridge to promiscuous mode | 17:43 |
| Leander | Birk_: i use the same config file for both the controller and the compute | 17:43 |
| Birk_ | Leander: on the controller node | 17:43 |
| mjfork | leander sounds like your networkign is hosed. if you only have 1 nic, enable promisc as birk said | 17:43 |
| mjfork | leanrox patebin the log for nova-api | 17:43 |
| Birk_ | Leander: you can do it with this command: sudo ifconfig br100 promisc | 17:44 |
| Leander | mjfork, Birk_, does that apply to a the VLAN configuration as well? | 17:44 |
| mjfork | Leander: i would suspect | 17:44 |
| mjfork | never tried it though | 17:44 |
| Birk_ | Leander: same as mjfork | 17:45 |
| Leander | now i need to restart the instance right? | 17:45 |
| Birk_ | mjfork: what log's you want ? | 17:45 |
| Leander | or nova-network? or both? | 17:45 |
| Birk_ | Leander: try to restart nova-network, nova-compute and nova-api. I always restart all of them. | 17:46 |
| Leander | with or without stopping the instance first? | 17:47 |
| mjfork | leander not sure a restart is required for promisc | 17:47 |
| mjfork | but provision new isntance | 17:47 |
| Birk_ | Leander: can you create another instance ? i would try that. | 17:47 |
| Birk_ | Leander: if you just changed the promisc you dont need to restart anything, like mjfork said | 17:47 |
| mjfork | Birk_: nvoa-compute | 17:47 |
| Birk_ | follow http://pastebin.com/RPLFt9vZ | 17:48 |
| Birk_ | mjfork: follow http://pastebin.com/RPLFt9vZ | 17:48 |
| Birk_ | mjfork: nothing changed on log after trying to connect via ssh | 17:48 |
| *** warik has quit IRC | 17:48 | |
| *** warik has joined #openstack | 17:49 | |
| *** rods has joined #openstack | 17:49 | |
| *** krish has quit IRC | 17:49 | |
| mjfork | Birk_: i was confirming key injection was ok, and it didn't error out from what i see] | 17:50 |
| mjfork | Birk_: since you are gettinga key error, it has to be there. | 17:50 |
| mjfork | Birk_: can you connect to the isntance via VNC? | 17:50 |
| mjfork | run virsh dumpxml <instance id> | grep vnc on the compute node | 17:50 |
| *** gyee has joined #openstack | 17:51 | |
| *** koolhead17 has left #openstack | 17:51 | |
| Leander | mjfork: still nothing | 17:52 |
| Leander | restarted all services | 17:52 |
| Leander | and launched new image | 17:52 |
| mjfork | what IP does your image have? | 17:52 |
| *** jdurgin has joined #openstack | 17:52 | |
| Leander | 10.1.1.3 | 17:52 |
| mjfork | Leander: can you SSH to it from teh compute node or from the controller node? | 17:52 |
| *** nati2 has joined #openstack | 17:53 | |
| Birk_ | mjfork: i don't have graphic interface on my servers, so i'm installing in another machine, then i will integrate this machine to the local network an then try to vnc | 17:53 |
| *** rods has quit IRC | 17:53 | |
| Leander | mjfork: not working | 17:54 |
| *** rods has joined #openstack | 17:54 | |
| *** dcramer_ has joined #openstack | 17:54 | |
| Leander | mjfork: on the controller it says:" ssh: connect to host 10.1.1.3 port 22: No route to host" and on the compute it just hangs | 17:55 |
| Birk_ | Leander: can you ping your instances ? | 17:55 |
| Leander | Birk_: neither ping nor ssh work | 17:55 |
| mjfork | leander flat dhcp? | 17:56 |
| Leander | mjfork: i'm using the VLAN configuration | 17:56 |
| *** rods has joined #openstack | 17:56 | |
| mjfork | Birk_: if you have VNC port, just VNC remotely | 17:56 |
| *** rods has quit IRC | 17:58 | |
| mjfork | leander, try it from outside your env | 17:58 |
| mjfork | i.e. outside the public interface | 17:58 |
| *** dwcramer has quit IRC | 17:58 | |
| *** rods has joined #openstack | 17:59 | |
| *** rods has quit IRC | 17:59 | |
| *** rods has joined #openstack | 18:01 | |
| *** cp16net has quit IRC | 18:02 | |
| *** imsplitbit has quit IRC | 18:02 | |
| *** cp16net has joined #openstack | 18:02 | |
| *** pdtpatrick_ has joined #openstack | 18:02 | |
| *** dolphm has joined #openstack | 18:03 | |
| *** sleepsonthefloo_ has joined #openstack | 18:03 | |
| *** reed__ has joined #openstack | 18:03 | |
| *** bepernoot has joined #openstack | 18:03 | |
| *** sleepsonthefloo has quit IRC | 18:04 | |
| *** sleepsonthefloo_ is now known as sleepsonthefloo | 18:04 | |
| *** reed_ has quit IRC | 18:05 | |
| *** adjohn has joined #openstack | 18:06 | |
| *** zzed has joined #openstack | 18:06 | |
| *** MarkAtwood has quit IRC | 18:07 | |
| Leander | mjfork: the network admin came by and it might be a problem with the network switch | 18:08 |
| Leander | it might be blocking the vlans | 18:08 |
| mjfork | Leander: that would do it :-) | 18:08 |
| *** darraghb has quit IRC | 18:08 | |
| Leander | mjfork: hope that's the case. Will be back tomorrow if it's not! | 18:11 |
| Leander | thanks for help so far | 18:11 |
| *** rustam has quit IRC | 18:12 | |
| *** Leander has left #openstack | 18:12 | |
| *** aculich has quit IRC | 18:12 | |
| *** MarkAtwood has joined #openstack | 18:12 | |
| *** rods has quit IRC | 18:14 | |
| *** uncleofthestick has quit IRC | 18:15 | |
| *** rods has joined #openstack | 18:18 | |
| *** bengrue has joined #openstack | 18:18 | |
| *** andreas__ is now known as ahasenack | 18:19 | |
| *** hugokuo has left #openstack | 18:20 | |
| *** nacx has quit IRC | 18:28 | |
| *** katkee has quit IRC | 18:29 | |
| *** tmichael has quit IRC | 18:29 | |
| *** jog0 has joined #openstack | 18:33 | |
| *** jog0 has left #openstack | 18:35 | |
| *** NOKAH has joined #openstack | 18:36 | |
| *** spackest has quit IRC | 18:36 | |
| *** spackest has joined #openstack | 18:36 | |
| *** judd7_ has joined #openstack | 18:36 | |
| *** statik has joined #openstack | 18:37 | |
| *** judd7 has quit IRC | 18:38 | |
| *** judd7_ is now known as judd7 | 18:38 | |
| *** Hakon|mbp has quit IRC | 18:40 | |
| *** rnorwood has quit IRC | 18:42 | |
| *** aculich has joined #openstack | 18:44 | |
| *** Carmivore has quit IRC | 18:45 | |
| uvirtbot | New bug: #914418 in nova "Nova notifications now seem to be using the default rabbit exchange " [Undecided,New] https://launchpad.net/bugs/914418 | 18:46 |
| warik | Hi, can anyone confirm me that is possible to use cc-host (configured with nova-network FlatDHCP) to give ups to instances launched on another nova-compute server? | 18:47 |
| *** Unicron-yeah has joined #openstack | 18:48 | |
| *** oubiwann has joined #openstack | 18:50 | |
| *** Unicron-yeah has quit IRC | 18:50 | |
| *** jakedahn has joined #openstack | 18:53 | |
| *** Carmivore has joined #openstack | 18:53 | |
| *** aculich has quit IRC | 18:55 | |
| *** oubiwann1 has joined #openstack | 18:55 | |
| *** oubiwann has quit IRC | 18:55 | |
| *** jog0 has joined #openstack | 18:56 | |
| mjfork | warik: not following, "give ups"? | 18:56 |
| warik | oops, I meant "IPs" | 18:56 |
| *** jakedahn has quit IRC | 18:57 | |
| mjfork | warik: yes, that is the basic use case | 18:57 |
| *** rnorwood has joined #openstack | 18:58 | |
| warik | ok, because in my case, the second nova-compute host the instance but I don't any dhcp request on the cc-host | 18:58 |
| warik | my instance got spawn, i can see the ip selected by the schedule server | 18:59 |
| warik | on the web interface but the nova-network never give away the ip, well it does but after 30 sec I can no longer ping my instance | 19:00 |
| warik | mjfork: is that sounds familiar to you ? | 19:00 |
| *** davlap has joined #openstack | 19:00 | |
| *** reed__ has quit IRC | 19:00 | |
| *** aculich has joined #openstack | 19:00 | |
| *** lionel has quit IRC | 19:00 | |
| *** lionel_ has joined #openstack | 19:01 | |
| *** reed has joined #openstack | 19:01 | |
| mjfork | warik, some what. so after 30 secs what happens? | 19:03 |
| *** MarkAtwood has quit IRC | 19:04 | |
| warik | i can ping my instance for about 30/35 sec and then the instance is no reachable | 19:04 |
| warik | but no error in the logs | 19:04 |
| mjfork | ok | 19:05 |
| mjfork | i know others have seen this | 19:05 |
| mjfork | trying to recal what it is | 19:05 |
| mjfork | could it be firewall rules? did you euca authorize? | 19:05 |
| spackest | I got debian squeeze installed and am trying to get openstack up and running, but when I try apt-get install -y glance it doesn't find the package | 19:05 |
| mjfork | spackest: add a repostirioy? | 19:05 |
| warik | I have the log from the cc: http://pastebin.com/KY5vbvWN and the compute: http://pastebin.com/3LBceMJR | 19:05 |
| spackest | when I try add-apt-repository ppa:glance-core/trunk it doesn't find the command | 19:06 |
| spackest | yeah, guess I'm asking which repo to add :) | 19:06 |
| mjfork | spackest: pase the error | 19:06 |
| warik | I add the euca-authorize for the ICMP and the SSH | 19:06 |
| uvirtbot | New bug: #914430 in openstack-ci "Config files needed on Jenkins build slaves for glance functional tests" [Medium,In progress] https://launchpad.net/bugs/914430 | 19:06 |
| uvirtbot | New bug: #914431 in openstack-ci "remove automatic comment text from reviews" [Low,Triaged] https://launchpad.net/bugs/914431 | 19:06 |
| spackest | -su: add-apt-repository: command not found | 19:06 |
| mjfork | spackest: ok, do you want diablo | 19:06 |
| spackest | that's what I was trying | 19:06 |
| mjfork | try sudo apt-get install python-software-properties | 19:06 |
| spackest | I am root and on debian so no sudo | 19:07 |
| mjfork | apt-get install python-software-properties | 19:07 |
| spackest | already installed | 19:07 |
| warik | spackest: and apt-get update :) | 19:07 |
| spackest | cat /etc/debian_version - 6.0.3 | 19:08 |
| *** tjikkun has quit IRC | 19:08 | |
| spackest | and even after the update I can't find the add-apt-repository command | 19:09 |
| spackest | am I just missing something obvious? | 19:09 |
| *** cakiiooo has joined #openstack | 19:10 | |
| warik | that should be it | 19:10 |
| *** dnjaramba has quit IRC | 19:10 | |
| warik | apt-get install python-software-properties then add-apt-repo then apt-get update | 19:10 |
| mjfork | warik: need nova-noetwork log | 19:10 |
| mjfork | spackest: if still no luck ,sounds like an OS issue, i would ask on their channel | 19:11 |
| *** tjikkun has joined #openstack | 19:11 | |
| *** tjikkun has joined #openstack | 19:11 | |
| *** dolphm has quit IRC | 19:11 | |
| warik | mjfork: the nova-network logs are part of the CC log. (sorry it's messy..) | 19:11 |
| mjfork | warik are you trying to access a private ip or floating ip? | 19:12 |
| spackest | is there a magic line I can put into /etc/apt/sources.list to point to a debian repo with openstack packages? | 19:12 |
| warik | private | 19:12 |
| warik | the 10.0.0.0/8 ramge | 19:12 |
| warik | range* | 19:12 |
| *** funzo has joined #openstack | 19:12 | |
| mjfork | no floating I{ | 19:13 |
| warik | spackest: idk, I always use the add-apt-repo | 19:13 |
| mjfork | IP | 19:13 |
| mjfork | do you have multihost? | 19:13 |
| mjfork | ie.. does nova-network run on every node | 19:13 |
| warik | nope, I have only one nove-network, which sits on the cc host | 19:13 |
| mjfork | you don't have the multihost flag set right? | 19:14 |
| warik | no | 19:14 |
| *** rnorwood has quit IRC | 19:14 | |
| mjfork | nova-network was enver started on the 2nd node? | 19:15 |
| warik | never installed, never run | 19:15 |
| mjfork | hmm, what happens after 30 secs I wonder. | 19:16 |
| mjfork | Kiall: you around? seems like we have answered this before. | 19:16 |
| Kiall | huh? | 19:16 |
| mjfork | Kiall: do you recall an issue where a new VM is accessible for 30-35 secs then you can no longer ping/ssh | 19:17 |
| *** rnorwood has joined #openstack | 19:17 | |
| Kiall | Yea - nova-api is inaccessible, there is an iptables NAT rule listed in the docs you need to use on the compute-only nodes to make it work right.. | 19:17 |
| spackest | warik: so add-apt-repo works for you just fine on stock debian? | 19:17 |
| warik | spackest: correct | 19:18 |
| spackest | 6.0.3? | 19:18 |
| *** cakiiooo has quit IRC | 19:18 | |
| Kiall | spackest: the apt-add-repository command is in the python-software-properties packages | 19:18 |
| warik | 6.0.1 | 19:18 |
| Kiall | Oh, I thought you said the command wasnt found.. ignore me ;) | 19:18 |
| warik | Kiall: is it this iptables rule ? 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 | 19:20 |
| Kiall | Yup - thats the one | 19:20 |
| warik | hmm, I added this one to my nova-compute already | 19:20 |
| *** bepernoot has quit IRC | 19:21 | |
| warik | the thing is, I don't get any error message in the logs.. | 19:21 |
| Kiall | Ah.. Okay, in that case there are some nova flags that set the local ranges to prevent the requests from being SNAT'ed before arriving at nova-api | 19:21 |
| spackest | http://pastebin.com/nqhqCekQ | 19:22 |
| *** zzed has quit IRC | 19:22 | |
| Kiall | warik: https://github.com/managedit/openstack-setup/blob/master/nova.conf.tmpl L27 -> L29 i think | 19:22 |
| Kiall | warik: not add-apt-... its apt-add-... | 19:23 |
| warik | Kiall: here are my config files : http://pastebin.com/DcHd3Xrp | 19:23 |
| warik | Kiall: yes, wrong spelling | 19:23 |
| warik | :S | 19:23 |
| Kiall | sorry - spackest: not add-apt-... its apt-add-... | 19:23 |
| warik | spackest: my bad.. | 19:24 |
| spackest | -su: apt-add-repository: command not found | 19:24 |
| Kiall | warik: Add the flags from L27 -> L29 from https://github.com/managedit/openstack-setup/blob/master/nova.conf.tmpl | 19:24 |
| russellb | i thought it was add-apt-repository ... always thought it was weird | 19:24 |
| Kiall | Usage: apt-add-repository <sourceline> | 19:24 |
| warik | Kiall: to both cc and compute ? | 19:25 |
| spackest | equally not available :) -su: add-apt-repository: command not found | 19:25 |
| Kiall | yea - the same config file should work on all nodes... | 19:25 |
| *** dpippenger has joined #openstack | 19:25 | |
| warik | ok | 19:26 |
| warik | ty | 19:26 |
| Kiall | spackest: weird, you said this was debian? | 19:26 |
| *** cp16net_ has joined #openstack | 19:26 | |
| Kiall | Maybe they dont have it ;) | 19:26 |
| *** PaulM has joined #openstack | 19:26 | |
| Kiall | spackest: the launchpad page for it also gives the apt lines you can use manually | 19:26 |
| *** funzo has quit IRC | 19:26 | |
| *** bpg has quit IRC | 19:27 | |
| spackest | Kiall: yeah, debain squeeze from today | 19:27 |
| Kiall | well .. I guess its an ubuntu addition so.. | 19:27 |
| *** cp16net has quit IRC | 19:29 | |
| *** cp16net_ is now known as cp16net | 19:29 | |
| *** Eyk^off is now known as Eyk | 19:30 | |
| spackest | Kiall: sorry, which launchpad page? | 19:31 |
| mjfork | warik: can you uyse VNC to get console on the guest and see if it is timing out (or use teh get console feature of dashbaord) | 19:31 |
| uvirtbot | New bug: #914440 in glance "No way to track show image bandwidth usage for reporting" [Undecided,In progress] https://launchpad.net/bugs/914440 | 19:31 |
| *** funzo has joined #openstack | 19:31 | |
| Kiall | spackest: the page for whatever PPA you are trying to add.. | 19:32 |
| *** bepernoot has joined #openstack | 19:33 | |
| warik | mjfork: I don't have VNC access, but when I try to see the log from Horizon, I get that in the log http://pastebin.com/RyWPvRPD | 19:34 |
| warik | Remote error: TypeError get_console_output() got an unexpected keyword argument 'tail_length' | 19:35 |
| mjfork | warik: cany ou see if VNC is open? virsh list then virsh dumpxml <id> | grep vnc | 19:35 |
| Kiall | warik: sounds like you are using the essex dashboard with a diablo nova... | 19:35 |
| Kiall | I think that was something added in essex (the tail_length param) | 19:35 |
| warik | Kiall: it's no good to use essex ? | 19:36 |
| *** mattray has quit IRC | 19:36 | |
| warik | because the instances that I launched on the cc host are doing alright | 19:36 |
| *** jakedahn has joined #openstack | 19:36 | |
| warik | only the remote ones are failing | 19:36 |
| Kiall | I *think* the ability to tail the logfile was added in essex, which if it was, would explain that error | 19:36 |
| Kiall | oh really? local ones work? | 19:37 |
| warik | yes! | 19:37 |
| mjfork | warik: even for get console? or just networking? | 19:37 |
| mjfork | Kiall's comment was just on wrong dasbhoard | 19:37 |
| warik | the local instances are fully operational, both ping/ssh + console | 19:38 |
| Kiall | yea.. there is no reference to "tail_length" anywhere in the diablo nova code.. and a few to it in the essex code... | 19:39 |
| warik | interesting, fyi, I am using the same version on both cc and compute | 19:40 |
| Kiall | yea.. something weird is going on there :) | 19:40 |
| warik | Kiall: I added the dmz flags and restart the services, no change | 19:41 |
| Kiall | warik: humm, can you pastebin nova-api.log ? | 19:42 |
| warik | Kiall: http://pastebin.com/KY5vbvWN it's a full log of the cc host (all the CC nova services are into this log, network, api, schedule) | 19:43 |
| mjfork | warik what did you isntall from? | 19:44 |
| warik | github | 19:44 |
| Kiall | yea, I don't see any errors in those logs.. | 19:45 |
| warik | master branch | 19:45 |
| Kiall | on both nodes? | 19:45 |
| warik | yep that's my problem! | 19:45 |
| warik | on both node | 19:45 |
| warik | no errors | 19:45 |
| warik | http://pastebin.com/3LBceMJR | 19:45 |
| warik | compute node ^ | 19:45 |
| Kiall | K - so everything is essex? no diablo anywhere? | 19:45 |
| warik | I don't really see any dhcp request either | 19:45 |
| *** pixelbeat has quit IRC | 19:48 | |
| warik | yes, except keystone | 19:48 |
| Kiall | weird - have you checked if VNC is listening like mjfork asked? | 19:48 |
| mjfork | warik: so wait, how can you ping for 30 secs if there is no DHCP request? | 19:48 |
| Kiall | would be worth connecting directly to that and seeing WTH is up.. | 19:48 |
| warik | I know! it's crazy | 19:48 |
| Kiall | actually .. yea .. that makes no sense -_- | 19:48 |
| warik | I guess, it's because the schedule request this ip and somehow it's gets reachable for a sec and then it's gone | 19:50 |
| mjfork | IP is prob dnsmasq'edo n the host ? | 19:50 |
| mjfork | abut then when assigned to the guest, its not reachable? | 19:50 |
| *** jedi4ever has joined #openstack | 19:50 | |
| *** mdomsch has joined #openstack | 19:52 | |
| warik | Kiall: virsh dumpxml instance-00000022 |grep vnc | 19:52 |
| warik | <graphics type='vnc' port='5900' autoport='yes' listen='0.0.0.0' keymap='en-us'/> | 19:52 |
| Kiall | warik: okay, so grab a VNC client and try to connect to "$compute_node:0" | 19:53 |
| warik | mjfork: I'm not following you.. | 19:53 |
| *** redconnection has joined #openstack | 19:53 | |
| mjfork | warik: n/m, lets get on VNC | 19:53 |
| warik | which packages should I install for ubuntu 11.10? | 19:54 |
| Kiall | any normal VNC client will do... | 19:54 |
| Kiall | (BTW you need a GUI for this.. so install it on your desktop) | 19:55 |
| warik | ok | 19:55 |
| Ryan_Lane | I need to move my rabbit queue from one node to another. is there any recommended way of doing so? | 19:56 |
| *** jedi4ever has quit IRC | 19:56 | |
| Ryan_Lane | rabbit basically says you can't move queues. I need to change hostname. | 19:56 |
| *** tr3buchet has quit IRC | 19:57 | |
| warik | Kiall: btw, I am running this test env. on virtual box ( idk if that's a clue or not) | 19:57 |
| Ryan_Lane | if I bring up a fresh queue, with no data, how will this affect my environment? | 19:57 |
| *** tr3buchet has joined #openstack | 19:57 | |
| Kiall | Ryan_Lane: you can wipe rabbitmq with no backups and reinstall fresh if you like.. no damage will be done, other than in-transit messages will be lost.. but thats no harm unless its a production environment | 19:58 |
| Ryan_Lane | it's a production environmnt | 19:58 |
| Ryan_Lane | I can disable the nova servies | 19:58 |
| Ryan_Lane | *services | 19:58 |
| Ryan_Lane | if that helps | 19:58 |
| Ryan_Lane | is shutting down the api server for a short period of time a reasonable way of killing in-transit messages? | 19:59 |
| Ryan_Lane | I can take an api outage, but I'd prefer not to corrupt things | 19:59 |
| *** freeflying has joined #openstack | 19:59 | |
| *** carlp has joined #openstack | 20:00 | |
| Kiall | Shutting down the api server will stop the majority of messages alright, a few seconds after you do that there should be nothing but helthcheck style messages bouncing around.. | 20:00 |
| Ryan_Lane | great | 20:00 |
| warik | Kiall / mjfork : I can't access the instance with VNC on the compute node, but I can with an instance on the CC node | 20:00 |
| Ryan_Lane | Kiall: thanks | 20:00 |
| *** mrmartin has joined #openstack | 20:01 | |
| *** adjohn has quit IRC | 20:02 | |
| Kiall | warik: Something is up with the instance, is the kvm process definitely running? | 20:02 |
| Spirilis | whew, presentation went alright, had our unix sysadmins glassy-eyed (lol) and one of them fell asleep, the cluster crashed when we stopped rabbitmq momentarily and started back up (haven't been able to start any new instances since, nova-network isn't playing nice anymore) | 20:05 |
| *** cereal_bars has joined #openstack | 20:05 | |
| Spirilis | otherwise, I think I got the point across as to what it is, got people thinking about use cases :) | 20:06 |
| *** aculich has quit IRC | 20:07 | |
| *** jfluhmann has joined #openstack | 20:08 | |
| mjfork | Spirilis: nice job! | 20:09 |
| Spirilis | lol looks like nova-network and nova-compute both bomb with some kombu error and broken pipe | 20:09 |
| *** GheRivero_ has joined #openstack | 20:09 | |
| Spirilis | the messages all say that the AMQP connection came back (nova-compute.log, nova-network.log) but yet it takes a dump | 20:09 |
| Spirilis | restarted all the services, going to try booting a VM now... | 20:10 |
| Spirilis | ok now it's in Build state, that's further than I was getting before. hrmm | 20:10 |
| Spirilis | seems like it doesn't gracefully handle RabbitMQ dieing and coming back | 20:10 |
| *** willaerk has joined #openstack | 20:11 | |
| *** rnorwood has quit IRC | 20:11 | |
| *** hermy|away is now known as hermy | 20:12 | |
| warik | Kiall: yes libvirtd is running | 20:12 |
| *** mdomsch has quit IRC | 20:12 | |
| Kiall | but is the kvm process running? | 20:12 |
| warik | kvm is not available on this server, it running qemu | 20:13 |
| warik | it's* | 20:13 |
| Kiall | ok, is qemu running then ;) | 20:13 |
| warik | :) | 20:13 |
| warik | I can see the qemu process for the instance but no qemu process itself | 20:14 |
| warik | is it normal ? | 20:14 |
| *** adjohn has joined #openstack | 20:14 | |
| Kiall | yea.. | 20:14 |
| Kiall | Okay, if qemu is running and you cant VNC to it directly (ie not with the dashboard, with a real/normal VNC client) then you have qemu issues ;) | 20:15 |
| warik | damn, ok | 20:16 |
| Kiall | Well.. that or a firewall blocking you from connecting via VNC | 20:17 |
| Spirilis | FWIW, http://paste.openstack.org/show/4215/ ... there's an example of the rabbitmq error from nova-network's perspective | 20:17 |
| warik | Kiall: i will look into that | 20:17 |
| Spirilis | It obviously reconnects to the rabbitmq instance, but when it tries to publish a reply "2012-01-10 14:55:08,803 ERROR nova.rpc [-] Failed to publish message [Errno 32] Broken pipe" | 20:17 |
| warik | Kiall / mjfork : thank you guys for the time! | 20:18 |
| Kiall | Spirilis: diablo or essex? I think RMQ reconnection was fixed in essex | 20:18 |
| benner | for only running vm (compute node) only nova-compute is nessesary? | 20:18 |
| *** maplebed has joined #openstack | 20:18 | |
| Spirilis | diablo | 20:18 |
| Spirilis | Kiall: ahh sweet | 20:18 |
| Spirilis | good to know that this is 1. a known bug and 2. fixed :) | 20:18 |
| Kiall | Actually, I think the fix might have been backported to diablo to... | 20:18 |
| * Kiall tests | 20:19 | |
| Spirilis | aye, I may have to update my copy of the codebase | 20:19 |
| Spirilis | mine was pulled from GIT on 12/21 | 20:20 |
| *** krow has joined #openstack | 20:24 | |
| *** aculich has joined #openstack | 20:25 | |
| *** MarkAtwood has joined #openstack | 20:25 | |
| Spirilis | yep there were definitely some changes to the codebase since 12/21 | 20:26 |
| Spirilis | just did a diff comparison | 20:26 |
| *** rnorwood has joined #openstack | 20:27 | |
| *** bepernoot has quit IRC | 20:30 | |
| *** aarcane_ has quit IRC | 20:32 | |
| *** funzo has quit IRC | 20:37 | |
| *** jog0 has left #openstack | 20:38 | |
| *** ohnoimdead has joined #openstack | 20:42 | |
| *** funzo has joined #openstack | 20:43 | |
| *** tmichael has joined #openstack | 20:44 | |
| *** LeanReox has quit IRC | 20:46 | |
| *** katkee has joined #openstack | 20:47 | |
| Spirilis | so it looks like restarting rabbitmq doesn't tank anything now... | 20:47 |
| Spirilis | but I ran into another error, this time in nova-network | 20:48 |
| Spirilis | er | 20:48 |
| Spirilis | nova-compute | 20:48 |
| Spirilis | (nova.compute.manager): TRACE: 'ip_address': mapping['ips'][0]['ip'], | 20:48 |
| Spirilis | (nova.compute.manager): TRACE: IndexError: list index out of range | 20:48 |
| Spirilis | ohh actually, weird, nova-api kicked off another rabbitmq reconnect error | 20:49 |
| *** cp16net has quit IRC | 20:50 | |
| *** heckj has quit IRC | 20:50 | |
| *** Gordonz has quit IRC | 20:50 | |
| *** ppradhan has quit IRC | 20:50 | |
| *** dolphm has joined #openstack | 20:50 | |
| Tobbe | mjfork: So I installed devstack as per your suggestion, and I've now logged in to the dashboard. All images that came with devstack were 64bit images, so I downloaded a 32bit ubuntu image and added that to Glance. But how do I launch it? | 20:50 |
| Tobbe | Going to the "instances" page it tells me there are no instances but that I can create a new one from the "images" page. Going to the "images" page I see no option to start/create anything. Just "edit" and "delete" | 20:51 |
| Tobbe | what am I missing? | 20:51 |
| Tobbe | Anyone else knows? | 20:54 |
| mjfork | Tobbe: flip over to user side | 20:54 |
| Tobbe | ahh, there it is :) | 20:55 |
| *** zzed has joined #openstack | 20:57 | |
| Tobbe | ohh, it got an IP (10.0.0.2) | 20:58 |
| Tobbe | "State" is still "Build" | 20:58 |
| *** uksysadmin has joined #openstack | 20:58 | |
| *** rnorwood has quit IRC | 20:59 | |
| Tobbe | woot, state changed to "active" :) | 21:00 |
| Tobbe | trying my luck with the vnc console... | 21:00 |
| spackest | well, I worked through some things and got glance up and running it appears, but now when I try to install the nova components, I get this -> http://pastebin.com/FTP0sND2 | 21:02 |
| *** bepernoot has joined #openstack | 21:02 | |
| *** dolphm has quit IRC | 21:02 | |
| spackest | on debian squeeze | 21:02 |
| mjfork | spackest: can you use Ubuntu ? | 21:03 |
| *** primeministerp has quit IRC | 21:03 | |
| spackest | I started on ubuntu 11.10 but had troubles, was going to go to 11.04, but debian is an approved OS for us, so I was trying it out | 21:03 |
| spackest | I have had several issues with debian and I'm not to the point where I was on 11.10 :) | 21:04 |
| *** flaviamissi has quit IRC | 21:04 | |
| *** Kiall2 has joined #openstack | 21:04 | |
| *** funzo has quit IRC | 21:04 | |
| Tobbe | well, the vnc console says "Connected (unencrypted) to: QEMU (instance-00000001)" But I don't see any output, and I can't interact with it. At least not yet. I'm a little worried about the mention of QEMU. IIRC someone told me that qemu doesn't work inside a VirtualBox machine | 21:04 |
| spackest | 11.10 I got all the way to starting an instance, and then things blew up on me | 21:04 |
| mjfork | spackest: what issues on ubunut? using devstck? | 21:05 |
| spackest | I need multi-node | 21:05 |
| *** Kiall has quit IRC | 21:05 | |
| spackest | an infinite loop on networking, just a sec, looks like a known bug, I just need to find it | 21:05 |
| *** Kiall2 is now known as Kiall | 21:05 | |
| mjfork | spackest: diablo release, multi node? what netowrking? | 21:06 |
| spackest | yes, diable release, multi node | 21:06 |
| Spirilis | Ok well, now that I've updated nova,glance,keystone,horizon,openstackx .... The rabbitmq reconnect error has changed disposition; nova-network and nova-compute no longer dump, but nova-api does | 21:06 |
| mjfork | spackest: what networking? | 21:06 |
| *** judd7 has quit IRC | 21:07 | |
| spackest | based on the networking here http://docs.openstack.org/diablo/openstack-compute/starter/content/Networking_Configuration-d1e1073.html using the extra config from the comment | 21:07 |
| Tobbe | I get this if I click on the "log" button: http://paste2.org/p/1861335 can anyone spot anything wrong? | 21:08 |
| mjfork | spackest: pretty straight fowrad, what repository did you use? | 21:08 |
| Tobbe | where does the nova logs end up when doing a devstack install? | 21:08 |
| spackest | this is what I hit https://bugs.launchpad.net/ubuntu/+source/linux/+bug/856417 | 21:08 |
| Tobbe | /var/log/nova does not exist | 21:08 |
| spackest | the box actually locks up when I try to start an instance | 21:09 |
| spackest | I asked about it a couple times in here but didn't get a resposne | 21:09 |
| spackest | response | 21:09 |
| spackest | wasn't sure how to manually configure eth1 | 21:09 |
| *** funzo has joined #openstack | 21:09 | |
| *** bepernoot has quit IRC | 21:10 | |
| mjfork | spackest: hmm | 21:10 |
| spackest | mjfork: exactly :) | 21:12 |
| *** BasTichelaar has quit IRC | 21:12 | |
| *** localhost has joined #openstack | 21:12 | |
| mjfork | spackest: on the phone..one sec. | 21:12 |
| Spirilis | Here's my latest rabbitmq restart error-- http://paste.openstack.org/show/4217/ ... Note that rabbitmq was restarted at 16:07:38, so it had been running for at least 20 seconds before nova-api dumped that broken pipe error. | 21:12 |
| *** JesperA has quit IRC | 21:14 | |
| Spirilis | then doing another hit, nova-api works correctly, but nova-network takes a dump | 21:17 |
| Spirilis | I'd say the rabbitmq problem's not fixed in latest git copy of stable/diablo | 21:18 |
| *** Kiall_ has joined #openstack | 21:23 | |
| *** tmichael has quit IRC | 21:23 | |
| *** DigitalFlux has joined #openstack | 21:24 | |
| *** DigitalFlux has joined #openstack | 21:24 | |
| *** tmichael has joined #openstack | 21:26 | |
| uvirtbot | New bug: #914476 in nova "New Python-Novaclient Extension mechanism doesn't work w/ 2.6" [Medium,In progress] https://launchpad.net/bugs/914476 | 21:36 |
| *** andrewbogott_ has joined #openstack | 21:36 | |
| *** andrewbogott_ has quit IRC | 21:36 | |
| *** andrewbogott_ has joined #openstack | 21:36 | |
| *** rods has quit IRC | 21:39 | |
| *** rods has joined #openstack | 21:39 | |
| *** andrewbogott has quit IRC | 21:39 | |
| *** andrewbogott_ has quit IRC | 21:41 | |
| *** Carmivore has quit IRC | 21:43 | |
| *** jaypipes has quit IRC | 21:43 | |
| *** Trebortech has joined #openstack | 21:43 | |
| *** andrewbogott has joined #openstack | 21:45 | |
| Trebortech | Anyone have an idea on how to add an additional virtual nic to a running server image? | 21:46 |
| *** maplebed has quit IRC | 21:46 | |
| Spirilis | wow weird, found a bug or quirk in the meta-data retrieval.... | 21:46 |
| *** dprince has quit IRC | 21:47 | |
| Spirilis | VMs on 2 out of my 3 nova-compute nodes have no trouble hitting http://169.254.169.254/latest/meta-data | 21:47 |
| Spirilis | a VM on the 3rd node does... and nova-api on the first node sees the request as coming from that VM's hypervisor's IP, not the VM IP like it sees with the other 2 | 21:47 |
| *** Carmivore has joined #openstack | 21:48 | |
| *** jaypipes has joined #openstack | 21:49 | |
| Spirilis | huh... nat is definitely misbehaving on that one node | 21:50 |
| uvirtbot | New bug: #914484 in nova "Boot from ISO is different in XS/KVM" [High,Triaged] https://launchpad.net/bugs/914484 | 21:51 |
| mjfork | spackest: i apologize, i have to run. call ran over. | 21:54 |
| *** uksysadmin has quit IRC | 21:58 | |
| *** spiffxp has quit IRC | 21:59 | |
| Spirilis | humm ok, so with nova-network running on every node and multi_host enabled, it appears outbound 169.254.169.254 connections get masqueraded so the nova-api host thinks they're coming from the hypervisor rather than the VM | 22:00 |
| *** sdake has quit IRC | 22:00 | |
| Spirilis | if you assign a floating IP to the VM, then, somehow it works and the nova-api server sees the 192.168.* private IP | 22:00 |
| Trebortech | Can someone direct me to good network documentation for compute? | 22:01 |
| *** hingo has quit IRC | 22:02 | |
| spackest | mjfork: ok, I downloaded ubuntu 11.04 and will give it a god and then try to get it approved by our org somehow :) | 22:04 |
| *** nati2 has quit IRC | 22:06 | |
| *** mcclurmc has quit IRC | 22:06 | |
| WormMan | VZW bought the MCI group I used to work for, we did back end networking for things like smog testing | 22:06 |
| WormMan | sigh | 22:06 |
| *** mcclurmc has joined #openstack | 22:07 | |
| * WormMan tries again, with the correct window | 22:07 | |
| *** leifmadsen has joined #openstack | 22:10 | |
| *** pixelbeat has joined #openstack | 22:13 | |
| *** sean_roberts has joined #openstack | 22:19 | |
| *** seanrob has joined #openstack | 22:19 | |
| *** seanrob has quit IRC | 22:19 | |
| *** sdake has joined #openstack | 22:19 | |
| *** ckleban has joined #openstack | 22:19 | |
| *** flaviamissi has joined #openstack | 22:24 | |
| *** dcramer_ has quit IRC | 22:25 | |
| *** tmichael has quit IRC | 22:27 | |
| *** tmichael has joined #openstack | 22:29 | |
| *** GheRivero_ has quit IRC | 22:30 | |
| *** andrewbogott has quit IRC | 22:31 | |
| *** jmoore has joined #openstack | 22:32 | |
| *** sean_roberts has quit IRC | 22:34 | |
| jmoore | quick question that hopefully someone in here can answer. does the openstack interface for nova-api have functions for stopping and starting an instance? like the ec2 StartInstances and StopInstances? | 22:34 |
| jmoore | i spend some time looking through but didn't see anything. | 22:35 |
| jmoore | if there isn't, are there plans to add some functions to do that? | 22:35 |
| *** krow has quit IRC | 22:36 | |
| *** redconnection has quit IRC | 22:41 | |
| *** ahasenack has quit IRC | 22:43 | |
| *** andrewsmedina has quit IRC | 22:45 | |
| *** panterax_ has joined #openstack | 22:45 | |
| *** andrewbogott has joined #openstack | 22:46 | |
| *** zzed has quit IRC | 22:46 | |
| *** panterax has quit IRC | 22:48 | |
| *** RobertLaptop has joined #openstack | 22:49 | |
| *** RobertLaptop has left #openstack | 22:49 | |
| *** lborda has quit IRC | 22:50 | |
| spackest | hmm, looks like nova-manage recently changed? the relevant commands here http://docs.openstack.org/diablo/openstack-compute/starter/content/Nova_conf-d2s367.html are failing | 22:50 |
| spackest | sudo nova-manage network create private 192.168.4.0/24 1 256 -> ValueError: invalid literal for int() with base 10: '192.168.4.0/24' | 22:51 |
| Madkiss | hm | 22:52 |
| *** miclorb_ has joined #openstack | 22:52 | |
| spackest | sudo nova-manage floating create --ip_range=10.88.39.96/27 -> TypeError: create() takes exactly 3 arguments (2 given) | 22:52 |
| *** jfluhmann has quit IRC | 22:52 | |
| *** joesavak has quit IRC | 22:53 | |
| *** Remco_ has joined #openstack | 22:53 | |
| *** mnabil has joined #openstack | 22:58 | |
| *** leifmadsen has quit IRC | 22:58 | |
| *** chrismsnz has joined #openstack | 22:58 | |
| spackest | I followed the advice here https://answers.launchpad.net/nova/+question/178297 and dropped the word private, which seemed to fix things? | 22:59 |
| *** mrmartin has left #openstack | 22:59 | |
| spackest | the command worked as documented in 11.10, but I guess not 11.04? | 22:59 |
| *** Remco_ has quit IRC | 23:00 | |
| spackest | according to http://nova.openstack.org/runnova/nova.manage.html nova-manage floating create --ip_range=10.88.39.96/27 should have a host in there somewhere | 23:02 |
| spackest | this went through sudo nova-manage floating create 10.88.39.90 10.88.39.96/27 but I don't know if it's right | 23:03 |
| Madkiss | openstackx - Blah python module that adds blah and bleh | 23:03 |
| Madkiss | wtf. | 23:03 |
| *** leifmadsen has joined #openstack | 23:03 | |
| *** Remco_ has joined #openstack | 23:05 | |
| *** cereal_bars has quit IRC | 23:09 | |
| *** pradeep1 has joined #openstack | 23:10 | |
| *** jfluhmann has joined #openstack | 23:11 | |
| *** rnorwood has joined #openstack | 23:12 | |
| *** Damianz has quit IRC | 23:14 | |
| *** lts has quit IRC | 23:15 | |
| *** Eyk is now known as Eyk^off | 23:15 | |
| *** Remco_ has quit IRC | 23:15 | |
| *** andrewsmedina has joined #openstack | 23:17 | |
| *** markvoelker1 has quit IRC | 23:20 | |
| *** DanF has quit IRC | 23:21 | |
| *** DanF has joined #openstack | 23:21 | |
| *** katkee has quit IRC | 23:22 | |
| *** spiffxp has joined #openstack | 23:22 | |
| *** AlanClark has joined #openstack | 23:31 | |
| *** _adjohn has joined #openstack | 23:41 | |
| *** jfluhmann has quit IRC | 23:41 | |
| *** rnorwood has quit IRC | 23:42 | |
| *** deshantm_ has joined #openstack | 23:43 | |
| *** adjohn has quit IRC | 23:44 | |
| *** _adjohn is now known as adjohn | 23:44 | |
| *** nati2 has joined #openstack | 23:44 | |
| Spirilis | heh found a workaround for my metadata-from-vms-on-multi_host-enabled-networks problem | 23:45 |
| Spirilis | http://forums.openstack.org/viewtopic.php?f=16&t=562&p=2212#p2212 | 23:45 |
| Spirilis | just run nova-api on all the compute nodes, along with nova-network, and set the --ec2_dmz_host on those nodes to point to its own IP rather than the IP of the "head" api node | 23:46 |
| *** deshantm has quit IRC | 23:46 | |
| *** Trebortech has quit IRC | 23:47 | |
| *** rods has quit IRC | 23:51 | |
| *** cereal_bars has joined #openstack | 23:53 | |
| *** rods has joined #openstack | 23:55 | |
| *** PotHix has quit IRC | 23:57 | |
| *** ldlework has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
