| uvirtbot | New bug: #904529 in keystone "Schema of ec2tokens is incompatible with XML" [Undecided,New] https://launchpad.net/bugs/904529 | 00:01 |
|---|---|---|
| *** adjohn has quit IRC | 00:01 | |
| *** _adjohn is now known as adjohn | 00:01 | |
| jkyle | how do you start up a vpn with nova-manage, diablo version? | 00:05 |
| jkyle | the docs say 'nova-manage vpn run <projectname>'. My nova-manage rejects that and says the proper syntax is nova-manage vpn run --project=<proj> --user=<user> | 00:07 |
| *** sdake has quit IRC | 00:16 | |
| *** hugo_kuo has quit IRC | 00:22 | |
| *** katkee has joined #openstack | 00:23 | |
| *** HowardRoark has joined #openstack | 00:25 | |
| *** v0id has quit IRC | 00:26 | |
| *** dragondm has quit IRC | 00:27 | |
| *** nati2 has quit IRC | 00:28 | |
| *** nati2 has joined #openstack | 00:28 | |
| *** dolphm has joined #openstack | 00:29 | |
| *** koolhead17 has quit IRC | 00:33 | |
| *** dnjaramba_ has joined #openstack | 00:33 | |
| *** dnjaramba has quit IRC | 00:33 | |
| *** wilmoore_ has joined #openstack | 00:38 | |
| *** wilmoore has quit IRC | 00:40 | |
| *** PiotrSikora has quit IRC | 00:44 | |
| *** PiotrSikora has joined #openstack | 00:44 | |
| *** po has quit IRC | 00:49 | |
| *** MarcMorata has quit IRC | 00:50 | |
| *** cole has quit IRC | 00:51 | |
| *** anotherjesse has quit IRC | 00:55 | |
| *** zul has quit IRC | 00:57 | |
| *** dolphm has quit IRC | 00:59 | |
| *** katkee has quit IRC | 00:59 | |
| *** lloydde has quit IRC | 01:00 | |
| *** wilmoore_ has quit IRC | 01:06 | |
| comstud | misheska: openstack xenserver code retries if it notices the DomID change | 01:06 |
| *** sdake has joined #openstack | 01:07 | |
| misheska | Thanks! | 01:07 |
| misheska | I'm not seeing this behavior though | 01:07 |
| comstud | misheska: we're probably moving away from xenstore for communication... into something that can be used commonly between all hypervisors... but at the moment that's the only way to configure xenserver guests | 01:08 |
| misheska | Thanks for the update - I'll do some more digging - I hope it's just a config problem on my end | 01:08 |
| *** pixelbeat has joined #openstack | 01:09 | |
| comstud | misheska: Ok. yeah.. essentially there's a get_agent_version() piece of code in nova.. and that's where the waiting happens | 01:09 |
| comstud | misheska: We want for agent reply.. in the middle of waiting, if we notice domID change, we start writing to the new domID | 01:09 |
| misheska | Thus you shouldn't need the a "guest agent ready" command in the protocol, right? | 01:10 |
| *** matiu_ has joined #openstack | 01:10 | |
| comstud | s/want/wait/ | 01:10 |
| comstud | correct | 01:10 |
| comstud | That was never implemented.. although that might have been handy ;) | 01:10 |
| *** tungvs has joined #openstack | 01:11 | |
| misheska | Thank you very much - at least I know now how it is supposed to work - much appreciated | 01:11 |
| misheska | comstud: one other question, you say it starts writing to the new domID | 01:16 |
| *** tryggvil_ has quit IRC | 01:16 | |
| misheska | What if the password request was on the old one, does it get rewritten to the new one? | 01:16 |
| misheska | If it doesn't that kinda explains the behaviour I'm seeing | 01:17 |
| comstud | no, it doesn't, because we get and wait for agent version first | 01:18 |
| *** jj0hns0n has joined #openstack | 01:18 | |
| misheska | Ah, so the use case I'm encountering is a Windows guest that Syspreps - thus it reboots | 01:19 |
| comstud | yes | 01:20 |
| comstud | that's the same that we use | 01:20 |
| comstud | the code flow in nova is this: | 01:20 |
| comstud | power on the VM | 01:20 |
| comstud | send guest agent 'version' command | 01:20 |
| comstud | wait for a period for response | 01:21 |
| comstud | if no response, try again... with current DomID | 01:21 |
| comstud | that 'try again' happens for a while | 01:21 |
| comstud | only after that do we try password, etc | 01:22 |
| *** matiu_ has quit IRC | 01:22 | |
| comstud | so essentially we keep retrying with the correct DomID until we get a response | 01:23 |
| misheska | Ah OK, so if this is working, I should be enabling the guest agent service AFTER Sysprep completes? I've tried it both ways though, with the guest agent on the Windows image set to auto-start, then also setting it to start post-Sysprep | 01:23 |
| comstud | even if DomID changes | 01:23 |
| *** andrewsben has quit IRC | 01:23 | |
| comstud | misheska: Good question.. I'm not sure which way ours is configured | 01:23 |
| comstud | i would think that either one should work | 01:23 |
| comstud | maybe post-sysprep is better | 01:24 |
| *** dotdevops has quit IRC | 01:24 | |
| misheska | It doesn't if you enable it first only because to prevent the Administrator password prompt from appearing you have to get Sysprep to set it - so it undoes the change the agent made - I verified this | 01:24 |
| comstud | misheska: In your logs on the nova compute node, do you see a message like "domid changed from <x> to <y>" ? | 01:24 |
| *** andreask has left #openstack | 01:25 | |
| misheska | No, but we're on cactus code - that's probably part of the problem | 01:25 |
| misheska | I'm working on getting us upgraded to diablo | 01:25 |
| comstud | Oooh | 01:25 |
| comstud | Yeah, that very well could be it | 01:25 |
| comstud | cactus might not have caught the DomID change. | 01:25 |
| misheska | I'll compare them - I bet you added this in diablo - I have to check | 01:25 |
| *** jj0hns0n has quit IRC | 01:25 | |
| comstud | We definitely made improvements in diablo | 01:25 |
| misheska | The guest agent code wasn't even made public until last December, I think, which would be in the diablo timeframe | 01:26 |
| *** matiu_ has joined #openstack | 01:27 | |
| misheska | Thanks, that gives me enough to go on - much appreciated | 01:27 |
| *** matiu_ has quit IRC | 01:27 | |
| *** nati2 has quit IRC | 01:28 | |
| *** nati2 has joined #openstack | 01:28 | |
| *** nati2 has quit IRC | 01:30 | |
| *** tmichael has quit IRC | 01:31 | |
| *** nati2 has joined #openstack | 01:31 | |
| *** nati2_ has joined #openstack | 01:38 | |
| *** nati2 has quit IRC | 01:38 | |
| *** stewart has quit IRC | 01:39 | |
| *** stewart has joined #openstack | 01:40 | |
| *** lzyeval has joined #openstack | 01:40 | |
| comstud | misheska: No prob | 01:43 |
| hugokuo | morning | 01:43 |
| misheska | comstud And yes I just verified that get_agent_version only exists in diablo | 01:44 |
| misheska | Totally explains what I was seeing - thank you! A big help to get confirmation on this without having to install diablo first ;-) | 01:44 |
| comstud | misheska: Yeah, that makes sense. I thought we were talking about diablo at first ;) | 01:44 |
| *** tmichael has joined #openstack | 01:53 | |
| *** stewart has quit IRC | 01:54 | |
| *** nati2_ has quit IRC | 01:54 | |
| *** nati2 has joined #openstack | 01:54 | |
| *** pixelbeat has quit IRC | 01:56 | |
| *** stewart has joined #openstack | 01:57 | |
| *** RicardoSSP has quit IRC | 02:01 | |
| *** jkyle has quit IRC | 02:02 | |
| *** adjohn has quit IRC | 02:03 | |
| *** misheska has quit IRC | 02:05 | |
| *** rods has quit IRC | 02:05 | |
| *** rustam has quit IRC | 02:06 | |
| *** MarkAtwood has quit IRC | 02:08 | |
| *** ben_duyujie has joined #openstack | 02:12 | |
| *** osier has joined #openstack | 02:12 | |
| *** lorin1 has joined #openstack | 02:13 | |
| *** Ryan_Lane|away is now known as Ryan_Lane | 02:16 | |
| *** zul has joined #openstack | 02:18 | |
| *** rsampaio has joined #openstack | 02:18 | |
| *** jakedahn has quit IRC | 02:22 | |
| *** jdurgin has quit IRC | 02:22 | |
| *** matiu_ has joined #openstack | 02:27 | |
| *** matiu_ has joined #openstack | 02:27 | |
| *** lorin1 has quit IRC | 02:30 | |
| *** dpippenger has quit IRC | 02:30 | |
| uvirtbot | New bug: #904559 in nova "PEP8 errors too verbose" [Undecided,In progress] https://launchpad.net/bugs/904559 | 02:31 |
| *** wilmoore has joined #openstack | 02:31 | |
| *** longhudou has joined #openstack | 02:33 | |
| *** sdake has quit IRC | 02:34 | |
| longhudou | 2011-12-15 00:58:24,437 - DataSourceEc2.py[WARNING]: waiting for metadata service at http://169.254.169.254/2009-04-04/meta-data/instance-id | 02:34 |
| longhudou | 2011-12-15 00:58:24,438 - DataSourceEc2.py[WARNING]: 00:58:24 [ 1/30]: url error [[Errno 111] Connection refused] | 02:34 |
| longhudou | anybody who can help | 02:34 |
| longhudou | this is in the console.log | 02:36 |
| mjfork | longhudou: how many nodes in your install? guessing 2 | 02:36 |
| mjfork | or 2+ | 02:36 |
| longhudou | 3 | 02:36 |
| longhudou | one controller node | 02:36 |
| mjfork | 2 computes? | 02:36 |
| longhudou | two compute node | 02:36 |
| longhudou | yes? | 02:37 |
| mjfork | is nova-network running on just the controller? | 02:37 |
| longhudou | yes | 02:37 |
| mjfork | ok | 02:37 |
| longhudou | nova-network is also running on the compute node | 02:38 |
| longhudou | is that right | 02:38 |
| longhudou | ? | 02:38 |
| mjfork | unless you are going for multihost, no. | 02:38 |
| mjfork | i would shut down nova-network on the compute nodes | 02:38 |
| mjfork | run it only on the controller node | 02:38 |
| *** 64MAABA57 has quit IRC | 02:38 | |
| longhudou | I have tried that ,the same error | 02:38 |
| mjfork | ok, shut it down on controoler nodes | 02:39 |
| mjfork | err | 02:39 |
| mjfork | shut it down on compute nodes | 02:39 |
| longhudou | ok | 02:39 |
| mjfork | after that, go into mysql on controlller node and pastebin the output of SELECT * FROM networks; | 02:39 |
| longhudou | SELECT * FROM networks;you mean that | 02:40 |
| longhudou | ERROR 1046 (3D000): No database selected | 02:41 |
| uvirtbot | New bug: #904560 in nova "InvalidInstanceState 'Cannot wrapped_f while the instance is in this state.'" [Critical,New] https://launchpad.net/bugs/904560 | 02:41 |
| longhudou | mjfork:ERROR 1046 (3D000): No database selected | 02:42 |
| mjfork | mysql nova | 02:42 |
| mjfork | or USE nova; in mysql | 02:42 |
| longhudou | Database changed | 02:43 |
| longhudou | after USE nova; | 02:43 |
| mjfork | now SELECT * FROM networks; | 02:43 |
| longhudou | http://paste.openstack.org/show/3813/ | 02:44 |
| longhudou | this is the result | 02:44 |
| mjfork | nova-cc your controller? | 02:45 |
| longhudou | yes | 02:45 |
| *** HowardRo_ has joined #openstack | 02:46 | |
| *** HowardRoark has quit IRC | 02:46 | |
| mjfork | ok, you have have to restart the compute nodes (make sure nova-network is disabled) | 02:46 |
| *** livemoon has joined #openstack | 02:46 | |
| mjfork | what is EC2_URL on the compute nodes? | 02:46 |
| longhudou | --ec2_url=http://192.168.1.91:8773/services/Cloud | 02:47 |
| longhudou | 91 is controller node ip_addr | 02:48 |
| mjfork | ok - try restarting the compute node to clear the neworking chagnes from nova-network | 02:48 |
| longhudou | http://paste.openstack.org/show/3814/ | 02:49 |
| longhudou | this is the euca-describe-availability-zone verbose | 02:49 |
| longhudou | on controller node | 02:49 |
| longhudou | is that ok? | 02:49 |
| mjfork | nova-mangage service disable nova-network nova-compute1 | 02:50 |
| mjfork | nova-mangage service disable nova-network nova-compute2 | 02:50 |
| mjfork | i think that | 02:50 |
| uvirtbot | New bug: #904561 in nova "--just-pep8 prompts for virtual env creation" [Undecided,In progress] https://launchpad.net/bugs/904561 | 02:50 |
| *** flamboz has quit IRC | 02:53 | |
| mjfork | longhudou: did that stop the service? | 02:53 |
| mjfork | or disable it i should say | 02:53 |
| *** pradeep has joined #openstack | 02:55 | |
| longhudou | 2011-12-15 10:59:14,276 nova: No service for nova-network, nova-compute2 | 02:59 |
| longhudou | mjfork: | 02:59 |
| mjfork | reverse the last 2 args | 02:59 |
| mjfork | nova-mangage service disable nova-compute1 nova-network | 02:59 |
| longhudou | that is right | 03:00 |
| longhudou | and then? | 03:00 |
| *** livemoon has quit IRC | 03:01 | |
| mjfork | run it for both nodes? | 03:01 |
| longhudou | yes | 03:01 |
| mjfork | does euca describe now show those nova-networks disabled? | 03:02 |
| longhudou | yes,I can not see nova-network on compute node | 03:02 |
| longhudou | now | 03:02 |
| *** livemoon has joined #openstack | 03:03 | |
| mjfork | ok, then restart your nodes | 03:04 |
| mjfork | compute nodes | 03:04 |
| longhudou | ok | 03:07 |
| longhudou | I have restart them both | 03:07 |
| *** vladimir3p has quit IRC | 03:07 | |
| longhudou | mjfork | 03:09 |
| mjfork | yes | 03:09 |
| mjfork | i would | 03:09 |
| longhudou | ? | 03:11 |
| mjfork | oh, you did restart them | 03:11 |
| mjfork | restart your guest, see if the error goes away | 03:12 |
| longhudou | another problem come:from the compute node ,I can not ping the instance on the controller node | 03:12 |
| longhudou | seems that I don't have br0? | 03:13 |
| longhudou | on the compute node | 03:13 |
| mjfork | what is in the console log of the guets | 03:13 |
| *** krow has joined #openstack | 03:13 | |
| *** mgius has quit IRC | 03:14 | |
| longhudou | http://paste.openstack.org/show/3815/ | 03:15 |
| longhudou | I think it is right | 03:15 |
| mjfork | yes, looksg ood. | 03:15 |
| mjfork | what is yoru networking? flat or flatdhcp | 03:15 |
| *** MarcMorata has joined #openstack | 03:15 | |
| longhudou | flatdhcp | 03:15 |
| longhudou | http://paste.openstack.org/show/3816/ | 03:16 |
| longhudou | after"brctl show " | 03:16 |
| mjfork | can you VNC into your guest? | 03:17 |
| longhudou | the first is on the controller node | 03:17 |
| mjfork | run ipconfig and see if the IP is set | 03:17 |
| longhudou | and the second is on the compute node | 03:17 |
| mjfork | can you use VNC to access your guest and see if the IP is set? | 03:18 |
| longhudou | you mean the fixed Ip ? | 03:18 |
| *** osier has quit IRC | 03:18 | |
| *** lmh_ has quit IRC | 03:18 | |
| longhudou | or floating IP | 03:18 |
| *** osier has joined #openstack | 03:18 | |
| mjfork | any IP | 03:19 |
| longhudou | I can not use vnc to access it | 03:19 |
| mjfork | why is that? | 03:19 |
| longhudou | even the controller node, | 03:19 |
| mjfork | virsh list | 03:19 |
| mjfork | then virsh dumpxml <instance name> | grep vnc | 03:20 |
| longhudou | 1 instance-00000039 running | 03:20 |
| mjfork | virsh dumpxml instance-00000039 | grep vnc | 03:20 |
| longhudou | <graphics type='vnc' port='5900' autoport='yes' listen='0.0.0.0' keymap='en-us'/> | 03:21 |
| mjfork | point VNC to port 5900 on that node | 03:21 |
| longhudou | sorry,dont know how to do that | 03:22 |
| longhudou | what do you mean? | 03:22 |
| mjfork | do you have a VNC viewer isntalled? | 03:23 |
| *** lmh has joined #openstack | 03:23 | |
| longhudou | I think yes | 03:23 |
| longhudou | when I use vncviewer in the command line | 03:24 |
| longhudou | VNC Viewer Free Edition 4.1.1 for X - built Apr 9 2010 15:52:37 | 03:24 |
| longhudou | it shows that | 03:24 |
| mjfork | connect to the IP address of of the machine running that guest | 03:25 |
| longhudou | vncviewer 10.0.0.2:5900 | 03:26 |
| longhudou | ? | 03:26 |
| *** mchenetz has quit IRC | 03:27 | |
| mjfork | sure, try that. | 03:27 |
| longhudou | http://paste.openstack.org/show/3817/ | 03:28 |
| longhudou | the result is like that | 03:28 |
| mjfork | do you have X running? | 03:28 |
| *** smashthestack has joined #openstack | 03:29 | |
| longhudou | I am not sure | 03:29 |
| longhudou | ? | 03:29 |
| longhudou | what is that | 03:29 |
| longhudou | just a rookie,sorry to say that | 03:29 |
| mjfork | ok, lets take a different approach | 03:30 |
| mjfork | make sure debug/verbose is on in nova.conf on all your nodes | 03:30 |
| longhudou | --verbose ,set it true? | 03:31 |
| mjfork | yes, on all 3 nodes | 03:32 |
| longhudou | ok | 03:33 |
| longhudou | I restart all the services | 03:33 |
| longhudou | on 3 nodes | 03:33 |
| mjfork | including nova-network on nova-cc? | 03:33 |
| longhudou | yes | 03:34 |
| mjfork | ok | 03:34 |
| mjfork | then provision a new server and pastebin nova-network.log from nova-cc and nova-compute.log from the host that gets the new VM | 03:34 |
| longhudou | a new server ? what is that , you mean an additional compute node | 03:35 |
| mjfork | no., sorry. a new guest | 03:37 |
| longhudou | I just delete all the nova-network.log,and then start a new guest ,,,,,,,,,,there is no nova-network.log on the controller node | 03:41 |
| longhudou | why? | 03:41 |
| *** mdomsch has joined #openstack | 03:43 | |
| mjfork | may need to restart service | 03:43 |
| mjfork | after deleting log file | 03:43 |
| *** pradeep has quit IRC | 03:47 | |
| longhudou | http://paste.openstack.org/show/3818/ | 03:48 |
| longhudou | the first before****************************is nova-network.log | 03:48 |
| longhudou | and the latter is nova-compute.log | 03:49 |
| mjfork | your glance server is not listening. | 03:49 |
| mjfork | 192.168.1.20:9292 | 03:49 |
| mjfork | can you telnet to that from the compute node | 03:50 |
| longhudou | 20? | 03:50 |
| longhudou | my controller node is 91 | 03:50 |
| mjfork | glance, where is glance installed? | 03:51 |
| longhudou | on the controller node | 03:51 |
| mjfork | ok, then fix the glance server in nova.conf | 03:52 |
| longhudou | when I use "telnet 192.168.1.91 9292" from the compute node | 03:52 |
| longhudou | Trying 192.168.1.91... Connected to 192.168.1.91. Escape character is '^]'. | 03:52 |
| longhudou | it shows that | 03:52 |
| mjfork | your compute node is looking @ .20 | 03:53 |
| mjfork | glance_host : 192.168.1.20 in the log you pasted | 03:53 |
| longhudou | let me have a look | 03:53 |
| longhudou | 2011-12-15 11:45:07,238 nova: glance_host : 192.168.1.20 | 03:54 |
| longhudou | from that? | 03:54 |
| longhudou | http://paste.openstack.org/show/3819/ | 03:55 |
| longhudou | this is the nova.conf on the compute node | 03:55 |
| mjfork | you have a - not an = | 03:56 |
| *** woleium has quit IRC | 03:57 | |
| mjfork | --glance_host-192.168.1.91 should be --glance_host=192.168.1.91 | 03:57 |
| *** map_nw has quit IRC | 03:59 | |
| longhudou | ok, i see ,now let me have a try for another time | 03:59 |
| *** map_nw has joined #openstack | 03:59 | |
| ipl31 | can someone tell me if it possible to have nova-network not change the IP of br100 in flatdhcp mode? | 04:00 |
| *** jakedahn has joined #openstack | 04:00 | |
| *** pradeep1 has joined #openstack | 04:03 | |
| *** andrewbogott has quit IRC | 04:03 | |
| longhudou | http://paste.openstack.org/show/3820/ | 04:03 |
| longhudou | mjfork | 04:03 |
| longhudou | this is the nova-compute.log on the compute node | 04:04 |
| mjfork | looks like perhaps your iamge was not loaded in glance properly | 04:05 |
| longhudou | seems that something is not transferred from the controller node(/var/lib/nova/instances/_base) | 04:05 |
| *** smashthestack has quit IRC | 04:07 | |
| *** cmasseraf has joined #openstack | 04:08 | |
| mjfork | what is the output of ls -l /var/lib/nova/instances/_base | 04:08 |
| longhudou | -rw-r--r-- 1 nova nogroup 0 2011-12-15 11:46 0000000d -rw-r--r-- 1 libvirt-qemu kvm 1476395008 2011-12-15 12:02 0000000e_sm | 04:09 |
| mjfork | delete the file "1" | 04:10 |
| mjfork | that is 0 bytes in length | 04:10 |
| mjfork | try again | 04:10 |
| *** bcwaldon has joined #openstack | 04:11 | |
| *** cfant has quit IRC | 04:12 | |
| longhudou | seems running from the controller node,but when use"virsh list" it is not running | 04:14 |
| mjfork | look in nova--compute.log | 04:15 |
| mjfork | should be an error | 04:15 |
| longhudou | http://paste.openstack.org/show/3821/ | 04:18 |
| longhudou | the nova-compute.log on the compute node | 04:18 |
| mjfork | looks OK | 04:19 |
| mjfork | what does virsh list have to say | 04:19 |
| *** woleium has joined #openstack | 04:20 | |
| longhudou | Id Name State ---------------------------------- 1 instance-00000039 running | 04:20 |
| longhudou | it is the instance running on the controller node | 04:20 |
| mjfork | ? | 04:20 |
| mjfork | what directories are in /var/lib/nova/instances/ | 04:21 |
| longhudou | now I have an instance running successfully on the controller node | 04:21 |
| mjfork | ok | 04:21 |
| longhudou | you mean on the compute node? | 04:22 |
| mjfork | yes, where you sent me the log that says instance instance-0000005e: is running | 04:22 |
| vlan-os | does nova network act as a gateway between different projects in case of vlanmanager ? | 04:23 |
| longhudou | on the compute node | 04:24 |
| longhudou | of course | 04:24 |
| longhudou | http://paste.openstack.org/show/3823/ | 04:26 |
| longhudou | console.log for instance-0000005e | 04:26 |
| mjfork | but it doesn't appear in virsh list? | 04:27 |
| longhudou | yes | 04:27 |
| mjfork | yes it does appear in virsh list? | 04:28 |
| longhudou | only the instance running on the controller node appear in virsh list | 04:28 |
| mjfork | then you need to run virs hlist on the compute node | 04:28 |
| longhudou | Id Name State ---------------------------------- 2 instance-0000005e running | 04:29 |
| longhudou | ok ,I find it | 04:29 |
| *** PeteDaGuru has quit IRC | 04:30 | |
| longhudou | but from the console.log ,there is still some error ,is that right? | 04:31 |
| mjfork | is nova-network running on this host? | 04:31 |
| mjfork | the comptue host this vm is running oni | 04:31 |
| longhudou | A moment ago,you got me to disable nova-network on the nova-network | 04:31 |
| *** cmasseraf has quit IRC | 04:31 | |
| longhudou | and I disabled that | 04:32 |
| longhudou | I am a little confused | 04:32 |
| mjfork | ok, just verifying that it is indeed not running here | 04:33 |
| mjfork | it should not be | 04:33 |
| mjfork | often "waiting for metadata service at http://169.254.169.254/2009-04-04/meta-data/instance-id" means that nova-network is running on tha tnode | 04:33 |
| longhudou | ok | 04:34 |
| longhudou | from console.log,can you see any error? | 04:35 |
| longhudou | http://paste.openstack.org/show/3823/ | 04:35 |
| mjfork | yes,the "waiting for metadata service at http://169.254.169.254/2009-04-04/meta-data/instance-id" followed by timeout errors | 04:36 |
| *** rsampaio has quit IRC | 04:37 | |
| longhudou | last time it is errno 111 | 04:38 |
| mjfork | that was nova-compute log wtih glance | 04:38 |
| longhudou | this time it is time out | 04:39 |
| mjfork | yes, the guest started, cannot hit the 169.254 addr | 04:40 |
| mjfork | on the compute node, what does "curl http://169.254.169.254" show | 04:41 |
| longhudou | <html><body><h1>It works!</h1> <p>This is the default web page for this server.</p> <p>The web server software is running but no content has been added, yet.</p> </body></html> | 04:42 |
| mjfork | that looks wrong | 04:42 |
| mjfork | did you ensure nova-network is not running and restart the node? | 04:42 |
| longhudou | http://paste.openstack.org/show/3824/ | 04:44 |
| longhudou | euca-describe-availability-zones verbose | 04:44 |
| longhudou | on the controller node | 04:44 |
| longhudou | maybe I should restart controller node? | 04:44 |
| mjfork | you restarted this compute node correct/ | 04:44 |
| longhudou | I use "shutdown -r now" | 04:45 |
| mjfork | what does "ip addr show | grep 169" show? | 04:45 |
| mjfork | ok | 04:45 |
| mjfork | let it reboot | 04:45 |
| longhudou | inet 169.254.169.254/32 scope link lo | 04:46 |
| mjfork | that should not be there | 04:47 |
| longhudou | ? | 04:47 |
| mjfork | well, wait | 04:48 |
| mjfork | was that on the comptue node? | 04:48 |
| longhudou | ok | 04:48 |
| longhudou | yes | 04:48 |
| mjfork | yes, if that was on teh compute node, it should node be | 04:48 |
| mjfork | if nova-network is disabled, a reboot should have cleared tha tout | 04:49 |
| *** krow has quit IRC | 04:49 | |
| longhudou | let me restart all 3 nodes | 04:49 |
| longhudou | do you think it is ok? | 04:49 |
| adam_g | anyone familiar with the new configuration module, specifically wrt glance? | 04:49 |
| mjfork | longhudou: you probably don't have to restart controller | 04:50 |
| longhudou | ok | 04:50 |
| longhudou | still there" inet 169.254.169.254/32 scope link lo" | 04:52 |
| mjfork | is there a nova-network log on that box? | 04:52 |
| longhudou | oh my god,there is nova-network.log | 04:53 |
| longhudou | maybe I should use apt-get auto remove to remove nova-network | 04:54 |
| mjfork | did you disable the serivce? | 04:55 |
| mjfork | i don't think you can remove it | 04:55 |
| longhudou | I disable the service on the controller node | 04:55 |
| longhudou | use the command you give | 04:55 |
| mjfork | update-rc.d -f nova-network remove | 04:56 |
| mjfork | maybe liek that? | 04:56 |
| mjfork | not an ubuntu user | 04:56 |
| *** jkyle has joined #openstack | 04:56 | |
| *** krow has joined #openstack | 04:56 | |
| longhudou | when I use"nova-manage service list" | 04:58 |
| longhudou | http://paste.openstack.org/show/3825/ | 04:58 |
| longhudou | on the compute node | 04:58 |
| mjfork | i am talking a system levle service | 04:58 |
| mjfork | not the openstack service | 04:58 |
| mjfork | openstack doesn't control if nova-network starts on the compute node | 04:58 |
| longhudou | I just run"update-rc.d -f nova-network remove" on the compute node | 04:59 |
| mjfork | ok | 05:02 |
| mjfork | try restring now | 05:02 |
| longhudou | the process nova-network is still alive | 05:02 |
| mjfork | after running nova-neotwork remove and restarting? | 05:03 |
| longhudou | nova 1711 1657 2 13:01 ? 00:00:01 /usr/bin/python /usr/bin/nova-network --flagfile=/etc/nova/nova.conf | 05:03 |
| longhudou | maybe it is caused of nova.conf? | 05:03 |
| *** llang629 has joined #openstack | 05:03 | |
| *** llang629 has left #openstack | 05:03 | |
| mjfork | no, its in /etc/init.d | 05:03 |
| mjfork | try | 05:05 |
| mjfork | chkconfig nova-network off | 05:05 |
| longhudou | and then | 05:07 |
| longhudou | restart the compute node? | 05:07 |
| longhudou | or just do chkconfig nova-network off | 05:08 |
| *** mestery has quit IRC | 05:08 | |
| longhudou | ok ,I have done"chkconfig nova-network off" | 05:08 |
| *** mestery has joined #openstack | 05:08 | |
| mjfork | ok | 05:09 |
| mjfork | reboot node | 05:09 |
| *** sdake has joined #openstack | 05:09 | |
| *** HowardRoark has joined #openstack | 05:12 | |
| *** HowardRo_ has quit IRC | 05:12 | |
| *** wilmoore has quit IRC | 05:14 | |
| longhudou | nova-network.log is still there | 05:16 |
| longhudou | mjfork | 05:17 |
| mjfork | is it running? | 05:17 |
| mjfork | ps auwxw | grep nova-nework | 05:17 |
| mjfork | ps auxww | grep nova-network | 05:17 |
| longhudou | nova 1605 0.0 0.0 3108 1316 ? Ss 13:15 0:00 su -c nova-network --flagfile=/etc/nova/nova.conf nova | 05:18 |
| longhudou | nova 1647 1.9 0.1 37948 22696 ? S 13:15 0:02 /usr/bin/python /usr/bin/nova-network --flagfile=/etc/nova/nova.conf | 05:18 |
| *** livemoon_ has joined #openstack | 05:18 | |
| mjfork | what does "chkconfig nova-newtork" say? | 05:20 |
| longhudou | I just use"apt-get purge nova-network" | 05:21 |
| longhudou | wait for a moment | 05:21 |
| mjfork | ok | 05:21 |
| longhudou | I will restart | 05:21 |
| *** livemoon has quit IRC | 05:21 | |
| *** jakedahn has quit IRC | 05:22 | |
| *** jakedahn has joined #openstack | 05:22 | |
| *** livemoon_ has quit IRC | 05:23 | |
| mjfork | k | 05:23 |
| longhudou | mjfork,this time ,nova-network disappears | 05:23 |
| mjfork | ok | 05:23 |
| *** rsampaio has joined #openstack | 05:25 | |
| longhudou | mjfork | 05:27 |
| *** jakkudan_ has joined #openstack | 05:27 | |
| longhudou | another problem | 05:27 |
| longhudou | I wonder could you help | 05:27 |
| mjfork | ok | 05:27 |
| longhudou | ? | 05:27 |
| mjfork | what is it? | 05:28 |
| longhudou | now I can ping and ssh the instance running on the compute node from the controller node | 05:28 |
| longhudou | but can not ping and ssh it from the compute node | 05:28 |
| *** krow has quit IRC | 05:29 | |
| mjfork | thats not surprising - flat DHCP right? | 05:30 |
| longhudou | yes | 05:30 |
| mjfork | the IP is actually bound to the controller node since it has nova-network running | 05:31 |
| mjfork | do you need to be able to SSH to it from compute node? | 05:31 |
| longhudou | yes | 05:32 |
| mjfork | why? | 05:32 |
| longhudou | can I? | 05:32 |
| mjfork | all of you access should originate outside of the OpenStack infrastrecutre | 05:32 |
| longhudou | you mean from controller node | 05:33 |
| longhudou | to communicate with openstack | 05:33 |
| longhudou | ? | 05:33 |
| mjfork | why do you need to SSH to your running guest from a compute node | 05:33 |
| mjfork | vs from your workstation | 05:33 |
| longhudou | the instance is running on the compute node , I should ping the vm running on my physical machine, | 05:34 |
| longhudou | is that right | 05:34 |
| mjfork | not sure what you are trying to do | 05:35 |
| longhudou | I should be able to | 05:35 |
| mjfork | are you trying to access the running VM ? | 05:35 |
| longhudou | yes | 05:35 |
| mjfork | and you can SSH to it from the controoller? | 05:35 |
| longhudou | yes | 05:36 |
| *** smashthestack has joined #openstack | 05:36 | |
| mjfork | then i woudl say you are all set. | 05:37 |
| mjfork | how do you access yuor controller? via SSH? | 05:37 |
| longhudou | yes | 05:37 |
| longhudou | I have not setup dashboard | 05:37 |
| longhudou | yet | 05:37 |
| mjfork | can you SSH to your guest IP from your workstatioN? | 05:38 |
| longhudou | what do you mean by "workstation"? sorry a little confused | 05:38 |
| mjfork | from where do you SSH to conttroler? | 05:39 |
| *** tungvs has quit IRC | 05:39 | |
| longhudou | an PC running windows7 | 05:39 |
| mjfork | ok, and what IP is that controller? | 05:39 |
| longhudou | 192.1681.91 | 05:40 |
| longhudou | 192.168.1.91 | 05:40 |
| mjfork | ok ,and what IP did you geuest VM get that you SSH to? | 05:40 |
| longhudou | 10.0.0.* | 05:41 |
| mjfork | what is the specific IP? | 05:41 |
| mjfork | you did not SSH to 10.0.0* and acces sthe guest right? | 05:41 |
| longhudou | 10.0.0.6 | 05:41 |
| mjfork | ok, can you SSH to that from your windows 7 machien? | 05:42 |
| longhudou | I cannot | 05:43 |
| mjfork | ok, then your routing is messed up I suspect | 05:43 |
| mjfork | if you can SSH to it from the controller | 05:43 |
| longhudou | I use ssh -i test3.pem ubuntu@10.0.0.6 from the controller node | 05:43 |
| longhudou | and I can ping and ssh 10.0.0.6,I mean this | 05:44 |
| mjfork | yep, but not from your windows 7 machien? | 05:44 |
| longhudou | yes,from my windows7 machine,I just try to ssh to 10.0.0.6 use putty(I wonder whether you know),but failed | 05:45 |
| *** nerens has joined #openstack | 05:45 | |
| mjfork | yeah, so my guess is your PC has no idea that 10.0.0.6 is available on your compute node | 05:46 |
| mjfork | err, on your controller node | 05:46 |
| *** rsampaio has quit IRC | 05:46 | |
| longhudou | on controller node? | 05:47 |
| mjfork | right | 05:47 |
| longhudou | I should assign an 192 address to the vm? | 05:49 |
| longhudou | or something like floating IP | 05:49 |
| mjfork | yes, that would help | 05:49 |
| *** tungvs has joined #openstack | 05:50 | |
| longhudou | can I assign an 10.0.0.* address to the compute node, so that from the compute node I can ping and ssh the instance | 05:51 |
| longhudou | ? | 05:51 |
| mjfork | it should work, but i don't see what that gets you | 05:51 |
| mjfork | just SSH to controller | 05:51 |
| longhudou | you mean then from the controller ping and ssh the instance(whether on the controller node or compute node),am I right? | 05:54 |
| *** livemoon has joined #openstack | 05:54 | |
| *** swills has quit IRC | 05:54 | |
| mjfork | yes, also SSH to controller node to access guests | 05:54 |
| *** livemoon has quit IRC | 05:54 | |
| *** swills has joined #openstack | 05:54 | |
| *** swills has joined #openstack | 05:54 | |
| longhudou | ok,I got it | 05:55 |
| mjfork | yes, always SSH to the controller node to access guests that is | 05:55 |
| longhudou | ok , thanks a lot | 05:56 |
| mjfork | np. glad to be of help. | 05:56 |
| *** livemoon has joined #openstack | 05:56 | |
| longhudou | I saw someone said you should have nova-network on each compute node | 05:56 |
| longhudou | maybe it is wrong | 05:56 |
| mjfork | that is for mutlihsot setup | 05:57 |
| mjfork | which can be useful if have a specific rason | 05:58 |
| mjfork | reason | 05:58 |
| longhudou | multihost is multi NIC card on one machine? | 05:58 |
| longhudou | or? | 05:59 |
| mjfork | no | 05:59 |
| mjfork | multiple hosts runing nova-nework | 05:59 |
| *** HowardRoark has quit IRC | 05:59 | |
| *** redconnection has quit IRC | 06:00 | |
| *** livemoon has left #openstack | 06:00 | |
| *** livemoon has joined #openstack | 06:01 | |
| *** mdomsch has quit IRC | 06:02 | |
| longhudou | mjfork: ok,got it,maybe will ask for your help in the future | 06:04 |
| *** pradeep1 has quit IRC | 06:07 | |
| *** mjfork has quit IRC | 06:07 | |
| *** jedi4ever has joined #openstack | 06:08 | |
| *** egant has joined #openstack | 06:09 | |
| *** smashthestack has quit IRC | 06:18 | |
| *** stewart has quit IRC | 06:19 | |
| *** bryguy has quit IRC | 06:23 | |
| *** bryguy has joined #openstack | 06:24 | |
| *** laclasse has quit IRC | 06:24 | |
| *** ldlework has quit IRC | 06:25 | |
| adam_g | d | 06:25 |
| *** adjohn has joined #openstack | 06:26 | |
| *** hadrian has quit IRC | 06:27 | |
| *** stewart has joined #openstack | 06:31 | |
| *** plu has quit IRC | 06:33 | |
| *** bepernoot has joined #openstack | 06:36 | |
| *** vlan-os has quit IRC | 06:37 | |
| *** jkyle has quit IRC | 06:39 | |
| *** laclasse has joined #openstack | 06:40 | |
| Madkiss | hi adjohn | 06:42 |
| Madkiss | err | 06:42 |
| Madkiss | hi adam_g | 06:42 |
| *** bepernoot has quit IRC | 06:51 | |
| *** rocambol1 has joined #openstack | 06:54 | |
| *** mindpixel has joined #openstack | 07:08 | |
| *** GheRivero has quit IRC | 07:15 | |
| *** osier_ has joined #openstack | 07:18 | |
| *** mgoldmann has joined #openstack | 07:20 | |
| *** osier has quit IRC | 07:20 | |
| *** guigui1 has joined #openstack | 07:21 | |
| *** ejat has joined #openstack | 07:24 | |
| *** ejat has joined #openstack | 07:24 | |
| *** koolhead17 has joined #openstack | 07:29 | |
| *** plu has joined #openstack | 07:29 | |
| *** TheOsprey has joined #openstack | 07:31 | |
| *** wilmoore has joined #openstack | 07:34 | |
| *** marcuz has joined #openstack | 07:36 | |
| *** dnjaramba_ has quit IRC | 07:43 | |
| *** dnjaramba has joined #openstack | 07:48 | |
| *** bepernoot has joined #openstack | 07:51 | |
| *** adjohn has quit IRC | 07:54 | |
| *** woleium has quit IRC | 07:55 | |
| *** nati2 has quit IRC | 08:00 | |
| *** MarcMorata has quit IRC | 08:04 | |
| *** Ramonster has joined #openstack | 08:10 | |
| *** reidrac has joined #openstack | 08:16 | |
| *** nRy has quit IRC | 08:17 | |
| *** nati2 has joined #openstack | 08:18 | |
| *** katkee has joined #openstack | 08:20 | |
| *** mikemowgli has quit IRC | 08:22 | |
| *** map_nw has quit IRC | 08:23 | |
| *** map_nw has joined #openstack | 08:23 | |
| *** slop has quit IRC | 08:24 | |
| *** nacx has joined #openstack | 08:26 | |
| *** nati2 has quit IRC | 08:28 | |
| *** vmf has joined #openstack | 08:33 | |
| *** koolhead17 has quit IRC | 08:33 | |
| *** foexle has joined #openstack | 08:34 | |
| foexle | ahoi | 08:34 |
| *** pixelbeat has joined #openstack | 08:34 | |
| *** mikemowgli has joined #openstack | 08:37 | |
| *** binbash_ has quit IRC | 08:40 | |
| *** longhudou has quit IRC | 08:40 | |
| *** binbash_ has joined #openstack | 08:41 | |
| *** alekibango has joined #openstack | 08:42 | |
| *** bepernoot has quit IRC | 08:42 | |
| *** bepernoot has joined #openstack | 08:42 | |
| *** vmf has quit IRC | 08:46 | |
| *** nijaba has quit IRC | 08:49 | |
| *** nijaba has joined #openstack | 08:50 | |
| *** nijaba has joined #openstack | 08:50 | |
| *** dev_sa has joined #openstack | 08:52 | |
| *** dev_sa has left #openstack | 08:52 | |
| *** zul has quit IRC | 08:52 | |
| *** osier_ has quit IRC | 08:54 | |
| *** andreask has joined #openstack | 08:56 | |
| *** katkee has quit IRC | 08:56 | |
| *** tungvs has quit IRC | 09:03 | |
| *** tungvs_ has joined #openstack | 09:03 | |
| *** lzyeval has quit IRC | 09:03 | |
| *** uksysadmin has joined #openstack | 09:04 | |
| *** dxd828 has joined #openstack | 09:04 | |
| *** osier has joined #openstack | 09:06 | |
| *** siwos has joined #openstack | 09:09 | |
| *** eqx311 has quit IRC | 09:12 | |
| *** map_nw_ has joined #openstack | 09:12 | |
| *** map_nw has quit IRC | 09:12 | |
| *** eqx311 has joined #openstack | 09:12 | |
| *** wariola has joined #openstack | 09:12 | |
| *** javiF has joined #openstack | 09:23 | |
| *** bepernoot has quit IRC | 09:30 | |
| *** Oneiroi has joined #openstack | 09:32 | |
| *** Oneiroi has quit IRC | 09:32 | |
| *** Oneiroi has joined #openstack | 09:32 | |
| *** Oneiroi has quit IRC | 09:33 | |
| *** Oneiroi has joined #openstack | 09:33 | |
| *** arBmind has joined #openstack | 09:37 | |
| *** wariola has quit IRC | 09:38 | |
| *** tryggvil_ has joined #openstack | 09:40 | |
| *** Oneiroi has quit IRC | 09:41 | |
| uksysadmin | hello all | 09:46 |
| uksysadmin | I tried to get Win2k3 SP2 ISO to boot under KVM last night to play with creating new images... but the ISO isn't bootable. What's the best workaround - copying the contents and creating your own bootable ISO? | 09:47 |
| *** ben_duyujie has left #openstack | 09:47 | |
| *** rustam has joined #openstack | 09:48 | |
| *** darraghb has joined #openstack | 09:49 | |
| *** Oneiroi has joined #openstack | 09:49 | |
| *** guigui1 has quit IRC | 09:59 | |
| *** vdo has joined #openstack | 10:03 | |
| *** nerens has quit IRC | 10:09 | |
| *** perestrelka has quit IRC | 10:13 | |
| *** livemoon has left #openstack | 10:15 | |
| *** perestrelka has joined #openstack | 10:17 | |
| *** dev_sa has joined #openstack | 10:17 | |
| *** guigui has joined #openstack | 10:18 | |
| *** dev_sa has left #openstack | 10:18 | |
| *** smashthestack has joined #openstack | 10:19 | |
| smashthestack | I was able to run the instances | 10:19 |
| smashthestack | but i cannot ping or ssh it | 10:19 |
| smashthestack | why? | 10:19 |
| smashthestack | its saying destination host unreachable | 10:20 |
| smashthestack | also is it possible to create a server with one network interface? | 10:20 |
| *** daysmen has joined #openstack | 10:26 | |
| *** zul has joined #openstack | 10:27 | |
| uksysadmin | smashthestack, have you created an appropriate security group that allows ssh and icmp? | 10:28 |
| smashthestack | yes | 10:28 |
| smashthestack | i followed your tutorial for virtual machines | 10:28 |
| smashthestack | it worked | 10:29 |
| smashthestack | but when it comes to physical machines | 10:29 |
| smashthestack | i used docs.openstack.org | 10:29 |
| smashthestack | but i got this pbm | 10:29 |
| uksysadmin | still using VLAN network? | 10:29 |
| smashthestack | no physical LAN | 10:29 |
| uksysadmin | yeah, but with VLAN tagging? Does your switch support it? | 10:29 |
| uksysadmin | my tutorial is quite "lazy" in that I know vbox doesn't care about VLANs... so it ignores them and allows the packets through regardless - like an unmanaged switch | 10:30 |
| uksysadmin | but if you have a managed switch then the ports connected to it must understand 802.11q and know to pass through VLAN 100 (default) | 10:31 |
| smashthestack | i ll check it out | 10:31 |
| *** pixelbeat has quit IRC | 10:31 | |
| smashthestack | i didnt think about it | 10:31 |
| smashthestack | Thanks anyway! | 10:33 |
| *** guigui has quit IRC | 10:33 | |
| *** wilmoore has quit IRC | 10:40 | |
| *** smashthestack has quit IRC | 10:41 | |
| *** jedi4ever has quit IRC | 10:45 | |
| *** map_nw has joined #openstack | 10:50 | |
| *** map_nw_ has quit IRC | 10:54 | |
| *** katkee has joined #openstack | 10:57 | |
| *** mindpixel has quit IRC | 11:01 | |
| *** Hakon|mbp has joined #openstack | 11:05 | |
| *** yshh has joined #openstack | 11:06 | |
| *** pixelbeat has joined #openstack | 11:10 | |
| *** katkee has quit IRC | 11:20 | |
| *** katkee has joined #openstack | 11:21 | |
| *** bepernoot has joined #openstack | 11:24 | |
| *** ejat has quit IRC | 11:26 | |
| *** ton_katsu has joined #openstack | 11:28 | |
| ton_katsu | hello | 11:32 |
| *** mcclurmc has quit IRC | 11:34 | |
| ton_katsu | i'm tring multi zone. diablo version. | 11:35 |
| ton_katsu | I can not start instance. | 11:35 |
| *** ejat has joined #openstack | 11:36 | |
| *** ejat has joined #openstack | 11:36 | |
| *** mcclurmc has joined #openstack | 11:36 | |
| *** livemoon has joined #openstack | 11:37 | |
| *** anpl has joined #openstack | 11:37 | |
| anpl | Hi | 11:37 |
| anpl | I have installed Diablo 4 on SL 6.1 | 11:37 |
| anpl | I used GridDynamics Yum repo | 11:38 |
| anpl | everything went fine | 11:38 |
| ton_katsu | anpl: multi zone ok? | 11:38 |
| anpl | but when I try to launch an instance I get the following error | 11:38 |
| anpl | Unable to launch instance: 'Fault' object has no attribute 'explanation' | 11:39 |
| anpl | nit multizone | 11:39 |
| anpl | I did a single node installation | 11:39 |
| *** rods has joined #openstack | 11:41 | |
| ton_katsu | anpl: keystone use ? | 11:44 |
| anpl | yes | 11:44 |
| anpl | I used keystone | 11:44 |
| *** pradeep has joined #openstack | 11:45 | |
| anpl | before I managed to launch an instance using same steps( 2 months back) using same repo | 11:45 |
| anpl | I have tried from scratch so many times | 11:45 |
| anpl | and I get the same error | 11:45 |
| anpl | do you need me share the error from the nova-api.log | 11:46 |
| anpl | one more error I see is when I try to allocate floating IP | 11:46 |
| anpl | it says the tenant has no free floating IP | 11:46 |
| anpl | I used nova-manage network create novanetwork 10.11.12.0/24 1 64 bridge_interface=eth0 | 11:49 |
| anpl | do we supply the tenant ID here? | 11:49 |
| *** marcuz has quit IRC | 11:51 | |
| ton_katsu | euca-describe-instances is ok? | 11:51 |
| *** tungvs_ has quit IRC | 12:00 | |
| *** pradeep has quit IRC | 12:00 | |
| *** ugarit has joined #openstack | 12:05 | |
| ugarit | do vmware's vmdk images "run" without problems on #openstack ? | 12:08 |
| *** ldlework has joined #openstack | 12:12 | |
| *** GeoDud has quit IRC | 12:19 | |
| *** jakkudan_ has quit IRC | 12:24 | |
| *** pixelbeat has quit IRC | 12:27 | |
| *** livemoon has left #openstack | 12:28 | |
| *** ugarit has quit IRC | 12:29 | |
| *** livemoon has joined #openstack | 12:29 | |
| Madkiss | /window 8 | 12:29 |
| Madkiss | oops | 12:29 |
| *** rsampaio has joined #openstack | 12:31 | |
| siwos | ugarit: even if not, you can easily convert them to qcow2 which is a native format for openstack | 12:32 |
| *** ben_duyujie has joined #openstack | 12:35 | |
| uvirtbot | New bug: #904723 in swift "additional fsync operation needed after file upload" [Undecided,New] https://launchpad.net/bugs/904723 | 12:41 |
| *** koolhead11 has joined #openstack | 12:41 | |
| *** hadrian has joined #openstack | 12:44 | |
| foexle | hey guys ... do any have an issue if any instance boot and comes to cloud-init they will be hangs up | 12:44 |
| *** livemoon has left #openstack | 12:44 | |
| *** yshh has quit IRC | 12:44 | |
| foexle | the instance won't to try to fetch metadata :( | 12:44 |
| foexle | no errors reported and i'm using default ubuntu uec image | 12:45 |
| *** po has joined #openstack | 12:45 | |
| uvirtbot | New bug: #904724 in nova "nova rebuild won't work" [Undecided,New] https://launchpad.net/bugs/904724 | 12:46 |
| *** itz_ has joined #openstack | 12:46 | |
| koolhead11 | foexle: and what is your nw like | 12:46 |
| foexle | nnw? | 12:47 |
| itz_ | Hi: installed nova-compute on RH6.1. The instance is created but fails to run.I see the error ignoring error injecting data into image 8 (cannot find any suitable libguestfs supermin or ordinary appliance on LIBGUESTFS_PATH (search path: /usr/lib64/guestfs)) in the log | 12:48 |
| koolhead11 | foexle network i meant | 12:49 |
| koolhead11 | can you pastebin error u see | 12:49 |
| * koolhead11 scratches his head | 12:50 | |
| koolhead11 | foexle: is the error says connecting to metadata server again and again? | 12:54 |
| foexle | nono | 12:56 |
| foexle | the boot process hangs up | 12:56 |
| *** saurabh has joined #openstack | 12:56 | |
| foexle | before they try to get metadata | 12:56 |
| saurabh | hello | 12:56 |
| koolhead11 | foexle: sounds cool | 12:56 |
| *** mjfork has joined #openstack | 12:56 | |
| *** ejat has quit IRC | 12:56 | |
| saurabh | I need to know about using ACLs | 12:57 |
| *** mjfork has quit IRC | 12:57 | |
| foexle | koolhead11: :/ yeah | 12:57 |
| koolhead11 | foexle: interesting. paste me the image url. i will download and try to run in my end | 12:57 |
| koolhead11 | !acl | 12:58 |
| openstack | koolhead11: Error: "acl" is not a valid command. | 12:58 |
| saurabh | I am using 'X-Container-Read: .r:*, .r: -test:tester3', need to know will this deny tester3 to download any object for the container in swift-? | 12:58 |
| foexle | koolhead11: the image runs ... it's the same image i used the last weeks | 12:59 |
| koolhead11 | foexle: so did you do any change in your infta <OS> | 13:00 |
| foexle | no | 13:01 |
| foexle | but i think it's maybe a networking problem | 13:02 |
| *** livemoon has joined #openstack | 13:03 | |
| foexle | -.- | 13:03 |
| foexle | solved | 13:03 |
| *** ejat has joined #openstack | 13:04 | |
| foexle | man ... i see don't see the forest for the trees -.- | 13:05 |
| *** pradeep1 has joined #openstack | 13:06 | |
| *** egant has quit IRC | 13:08 | |
| *** egant has joined #openstack | 13:10 | |
| *** daysmen has quit IRC | 13:10 | |
| *** rustam has quit IRC | 13:11 | |
| koolhead11 | foexle: i thought so | 13:11 |
| koolhead11 | :P | 13:11 |
| *** ejat has quit IRC | 13:13 | |
| *** ejat has joined #openstack | 13:13 | |
| *** ejat has joined #openstack | 13:13 | |
| *** mjfork has joined #openstack | 13:13 | |
| *** praefect has joined #openstack | 13:14 | |
| *** Nac_ has joined #openstack | 13:15 | |
| *** stuntmachine has joined #openstack | 13:16 | |
| Nac_ | Hello everybody | 13:17 |
| *** gohko_nao has joined #openstack | 13:18 | |
| rwmjones | itz_: which version of nova-compute is this? AFAIK there is no libguestfs in nova | 13:18 |
| Nac_ | I have a little cluster made by 6 nodes and they are allway cosuming a lot CPU and net bandwidth...... also when the cluster is empty and nobody is doing anything | 13:19 |
| Nac_ | why? | 13:19 |
| *** javiF has quit IRC | 13:19 | |
| Nac_ | the procces is "swift-container" | 13:19 |
| Nac_ | and it is consume a lot of CPU .... ALWAYS.....in ALL the nodes.....why? | 13:21 |
| *** rustam has joined #openstack | 13:21 | |
| *** rsampaio has quit IRC | 13:21 | |
| *** fujin has quit IRC | 13:23 | |
| *** jingizu_ has quit IRC | 13:23 | |
| *** scottsanchez has quit IRC | 13:23 | |
| *** ben_duyujie has left #openstack | 13:24 | |
| *** PotHix has joined #openstack | 13:24 | |
| *** pixelbeat has joined #openstack | 13:30 | |
| *** fujin has joined #openstack | 13:32 | |
| *** cmasseraf has joined #openstack | 13:33 | |
| *** hugokuo has quit IRC | 13:33 | |
| *** saurabh has left #openstack | 13:34 | |
| *** fghaas has joined #openstack | 13:36 | |
| *** jingizu_ has joined #openstack | 13:36 | |
| *** egant has quit IRC | 13:38 | |
| mjfork | what is the right way to prevent the nova-scheduler service from starting on Ubuntu? | 13:39 |
| foexle | mjfork: update-rc.d -f nova-scheduler remove | 13:40 |
| mjfork | thanks foexle | 13:42 |
| *** ninkotech has quit IRC | 13:42 | |
| *** vlan-os has joined #openstack | 13:45 | |
| *** arBmind_ has joined #openstack | 13:46 | |
| *** mjfork has quit IRC | 13:47 | |
| *** mjfork has joined #openstack | 13:48 | |
| *** mattray has joined #openstack | 13:48 | |
| *** arBmind has quit IRC | 13:48 | |
| *** arBmind_ is now known as arBmind | 13:48 | |
| *** Xenith has quit IRC | 13:52 | |
| *** Xenith has joined #openstack | 13:53 | |
| *** PeteDaGuru has joined #openstack | 13:55 | |
| vlan-os | I know that network host acts as the gateway for traffic from vm's in a project to the outside world but does it also act as gateway for vm's in different project (subnets) to talk to each other ? | 13:59 |
| itz_ | rwmjones: openstack-nova-compute-2011.3-b1188 | 14:00 |
| mjfork | vlan-os: VLANs or ? | 14:00 |
| *** PeteDaGuru has quit IRC | 14:00 | |
| itz_ | rwmjones: https://review.openstack.org/#change,1994 | 14:01 |
| vlan-os | mjfork: yes 1:1 mapping of vlans and subnet | 14:01 |
| *** PeteDaGuru has joined #openstack | 14:01 | |
| mjfork | yes, i would expect that nova-network would have to act as gateway since none of the machines could see each other | 14:02 |
| *** scottsanchez has joined #openstack | 14:03 | |
| vlan-os | mjfork: but that implies that vm's in one project be able to reach vm's on another project via the network host -< this part is giving me sleepless nights | 14:03 |
| rwmjones | itz_: oh I see you're using padraig's patch | 14:05 |
| rwmjones | itz_: what host platform is this? | 14:05 |
| rwmjones | and what version of libguestfs have you got installed | 14:05 |
| *** lorin1 has joined #openstack | 14:05 | |
| rwmjones | and can you run libguestfs-test-tool | 14:05 |
| mjfork | checking | 14:07 |
| *** cloudgeek has joined #openstack | 14:07 | |
| vlan-os | mjfork: thank you sir ! | 14:07 |
| mjfork | vlan-os: have you tested it today? | 14:08 |
| vlan-os | mjfork : using diablo release on ubuntu - yes | 14:08 |
| mjfork | and your machines on different VLANs can talk? | 14:08 |
| foexle | vlan-os: searching by google openstack shared networks | 14:08 |
| *** osier has quit IRC | 14:09 | |
| vlan-os | mjfork: yes, coz different vlan's map to different subnets and network host has a bridge interface that acts as a gateway for each subnet. | 14:10 |
| *** Razique has joined #openstack | 14:10 | |
| mjfork | so yes, you tested it and verified that they can talk cross VLANs through gateway? | 14:11 |
| vlan-os | mjfork: yes | 14:11 |
| mjfork | ok, so i would propose that on the mailing list and ask for solutions | 14:12 |
| mjfork | i would expect it is some set of iptables rules | 14:12 |
| *** dprince has joined #openstack | 14:12 | |
| rwmjones | pixelbeat: ^^ itz_ is using your patch | 14:13 |
| kodapa | Hello, I'm trying to implement this: http://blog.defunct.ca/2011/09/19/using-nova-networks-multi_host-to-remove-spof/ | 14:15 |
| vlan-os | mjfork: ok..but i the behavior that i'm seeing shouldnt be happening coorect ? | 14:15 |
| kodapa | Networking is working, except for cloud-init metadata stuff | 14:15 |
| *** ejat has quit IRC | 14:16 | |
| kodapa | on the nodes running only nova-compute and nova-network | 14:16 |
| kodapa | I added --ec2_dmz_host=10.24.1.10 and --multi_host to nova.conf | 14:16 |
| kodapa | Any ideas? | 14:16 |
| mjfork | vlan-os: well, my guess is if your routing is setup to allow it nova-network doesn't do anythign to break it. | 14:17 |
| *** joesavak has joined #openstack | 14:17 | |
| *** mrjazzcat has joined #openstack | 14:17 | |
| vlan-os | mjfork: yes, ip forwarding is enabled and the routes are populated to direct traffic to the correct bridge interface for each subnet on the network hos | 14:18 |
| mjfork | vlan-os: i would propose your setup on the general mailing list, the current behavior, and your desired behavior and ask how best to get there. | 14:19 |
| *** jsavak has joined #openstack | 14:20 | |
| uvirtbot | New bug: #904767 in nova "nova-manage vpn spawn won't work" [Undecided,New] https://launchpad.net/bugs/904767 | 14:21 |
| *** Shentonfreude has joined #openstack | 14:22 | |
| *** mrjazzcat has left #openstack | 14:23 | |
| *** joesavak has quit IRC | 14:23 | |
| *** cloudgeek has quit IRC | 14:25 | |
| *** freeflying has quit IRC | 14:28 | |
| *** freeflying has joined #openstack | 14:28 | |
| itz_ | rwmjones: RH6.1 | 14:31 |
| *** lborda has joined #openstack | 14:32 | |
| mjfork | kodapa start nova-api on each hsot | 14:32 |
| Razique | hey guys :) | 14:32 |
| foexle | bon soir monsieur Razique :) | 14:33 |
| *** Shentonfreude has left #openstack | 14:33 | |
| kodapa | mjfork: hmm i tried that before | 14:33 |
| kodapa | mjfork: I seem to have the same problem as this guy: https://answers.launchpad.net/nova/+question/181258 | 14:33 |
| kodapa | No answers there :( | 14:33 |
| foexle | kodapa: this guy its me :D | 14:34 |
| mjfork | so, you started nova-api on each compute host running nova-network? | 14:35 |
| foexle | kodapa: do you have your api on an other host as your compute ? | 14:35 |
| rwmjones | itz_: that's RHEL 6.1 with libguestfs 1.7.17-17? | 14:35 |
| *** alekibango has quit IRC | 14:35 | |
| kodapa | foexle: yes | 14:36 |
| kodapa | foexle: i've got two nodes | 14:36 |
| kodapa | foexle: one running nova-api, nova-scheduler, nova-compute, nova-network | 14:36 |
| kodapa | foexle: and another running nova-compute and nova-network | 14:36 |
| kodapa | foexle: running multi host mode and vlan networking | 14:36 |
| mjfork | the 2nd node should be running nova-pi | 14:36 |
| mjfork | api | 14:36 |
| kodapa | mjfork: should I remove --ec2_dmz_host=10.24.2.10 in nova.conf? | 14:37 |
| kodapa | That's what a guide said fix the metadata stuff | 14:37 |
| foexle | kodapa: ah ok :) .... and if you spawn an instance on host 1 you get a connection to metadata and on the second not right ? | 14:37 |
| kodapa | foexle: excactly | 14:37 |
| Razique | hey mr foexle | 14:38 |
| foexle | kodapa: ok so that is a problem in nova network .... you have no route to your api | 14:38 |
| *** katkee has quit IRC | 14:38 | |
| Razique | sup ? | 14:38 |
| kodapa | foexle: hm | 14:38 |
| foexle | karmabot: you have 3 options | 14:38 |
| karmabot | What? | 14:38 |
| kodapa | mjfork: it's working now after installing nova-api again on second node | 14:39 |
| kodapa | mjfork: strange because I had it before | 14:39 |
| *** katkee has joined #openstack | 14:39 | |
| kodapa | foexle: is there a way to fix it? | 14:39 |
| foexle | kodapa: so the route are not correct in nova | 14:39 |
| foexle | kodapa: you have a route to your api => if you have set the correct dmz host and souce_ip | 14:39 |
| *** huslage has joined #openstack | 14:39 | |
| foexle | kodapa: but the masqarading of your internatl ip | 14:39 |
| kodapa | souce_ip? source_ip? | 14:39 |
| foexle | its the problem | 14:40 |
| foexle | kodapa: yes | 14:40 |
| mjfork | kodapa so it works? what are your ec2_dmz_host and ec2_url | 14:40 |
| *** JesperA has joined #openstack | 14:40 | |
| *** cloudgeek has joined #openstack | 14:40 | |
| foexle | kodapa: you can take your api-server in the same net (floating or fixed) | 14:41 |
| foexle | kodapa: but you have running a compute and network on your api host | 14:41 |
| foexle | kodapa: so you can change your dmz-host to the bridge ip of host 1 | 14:42 |
| foexle | on host2 | 14:42 |
| foexle | and your source ip (second node) ip of the bridge | 14:42 |
| kodapa | mjfork: No it didn't work, it just spawned on first node :P | 14:42 |
| kodapa | mjfork: I'm blind, sorry xD | 14:42 |
| kodapa | foexle: http://pastebin.com/3fQtT19n | 14:42 |
| *** jedi4ever has joined #openstack | 14:42 | |
| foexle | but thats a verry messi way | 14:42 |
| kodapa | foexle: that's my nova.conf | 14:42 |
| foexle | which fixed network you are using ? | 14:43 |
| *** cmasseraf has quit IRC | 14:43 | |
| Nac_ | Hi | 14:43 |
| kodapa | foexle: I'm running vlan networking | 14:43 |
| foexle | yeah i mean wich ip range | 14:44 |
| Nac_ | How can I see a list of user on Swift? | 14:44 |
| kodapa | foexle: different for each project | 14:44 |
| kodapa | foexle: 10.24.2.0/24 is the net between nova boxes | 14:44 |
| kodapa | foexle: 10.24.2.10 is box 1 and 10.24.2.10 is box 2 | 14:44 |
| kodapa | ooops | 14:44 |
| kodapa | 10.24.2.11 is box 2 | 14:45 |
| foexle | yeah thats was my problem :) .... my workaround solution => auto-floating ip | 14:45 |
| kodapa | That's not what we want | 14:45 |
| kodapa | We need VLAN | 14:45 |
| kodapa | :/ | 14:45 |
| foexle | yeah it's only to startup | 14:45 |
| foexle | to communicato to get metadata | 14:45 |
| kodapa | Yes but this wont work for us | 14:46 |
| kodapa | Customers should be able to launch own instances | 14:46 |
| foexle | after that you can disattach the pub ip | 14:46 |
| kodapa | aha now I understand what you mean | 14:46 |
| vlan-os | mjfork: https://answers.launchpad.net/nova/+question/181994 | 14:47 |
| foexle | kodapa: so openstack have many ha-options and tricks, but the main thing it is not :( unfortunately | 14:47 |
| *** joesavak has joined #openstack | 14:47 | |
| mjfork | you could pop over to openstack-dev and ask vishy - he wrote the multihost articles | 14:47 |
| *** ton_katsu has quit IRC | 14:48 | |
| *** zigo has joined #openstack | 14:48 | |
| foexle | so if you need to communicate over the backnet | 14:48 |
| foexle | should you take a look to quantum | 14:48 |
| foexle | but i think it's not production ready | 14:48 |
| foexle | Razique: whats up man ? :) | 14:49 |
| *** jsavak has quit IRC | 14:51 | |
| kodapa | foexle: I'm getting TRACE: RemoteError: FloatingIpNotFoundForAddress Floating ip not found for address | 14:52 |
| kodapa | foexle: with auto floating ip | 14:52 |
| kodapa | :/ | 14:52 |
| *** itz_ has quit IRC | 14:52 | |
| foexle | kodapa: sec | 14:52 |
| mjfork | is auto floating IP an option? | 14:52 |
| *** zz_bonzay is now known as bonzay | 14:52 | |
| foexle | kodapa: https://github.com/openstack/nova/commit/80d73a2df2c33212217c59783f62d74673776ebb change this 4 lines | 14:53 |
| foexle | and restart nova-network | 14:53 |
| foexle | after that its running | 14:53 |
| kodapa | foexle: I'm running ubuntu 11.10 packages :/ | 14:53 |
| *** bonzay is now known as zz_bonzay | 14:54 | |
| *** robbiew has joined #openstack | 14:54 | |
| foexle | me too | 14:54 |
| foexle | its the same | 14:55 |
| foexle | nova diablo 2011.3 | 14:55 |
| *** hugokuo has joined #openstack | 14:55 | |
| *** mattray has quit IRC | 14:55 | |
| foexle | you can locate this file in /usr/shared/pyshared/nova/network | 14:55 |
| *** mchenetz has joined #openstack | 14:57 | |
| kodapa | foexle: then what happens if theres an update? | 14:57 |
| kodapa | foexle: will the changes be overwritten? | 14:57 |
| foexle | if they not in the next update yes :( | 14:58 |
| foexle | but i've backuped manage.py | 14:58 |
| kodapa | foexle: please tell me if you find another solution for this | 15:00 |
| foexle | kodapa: you too :) | 15:00 |
| kodapa | foexle: I have a deadline on the openstack cloud setup in january so there's still some time to fix things | 15:00 |
| kodapa | However people want to test it like, yesterday | 15:00 |
| foexle | kodapa: yeah i have the same problem :) | 15:01 |
| *** rsampaio has joined #openstack | 15:01 | |
| foexle | kodapa: but my first openstack cloud setup is with 18 servers and shared services like loadbalancer and db cluster -.- .... many problems :) but i think the essex version will be better for ha | 15:02 |
| *** CaptTofu has joined #openstack | 15:03 | |
| *** deshantm has joined #openstack | 15:03 | |
| CaptTofu | morning! | 15:04 |
| kodapa | foexle: yeah hopefully | 15:04 |
| JesperA | evening | 15:04 |
| kodapa | foexle: I have customers queued already for the new cloud | 15:04 |
| foexle | dito :> | 15:04 |
| kodapa | foexle: our old cloud is full and we're out of rack space and the DC is so damn slow to setup new racks | 15:04 |
| kodapa | foexle: we've been waiting over a year for a patch cable :) | 15:04 |
| foexle | :) | 15:04 |
| kodapa | foexle: our new cloud is in another DC ;D | 15:05 |
| foexle | :> | 15:05 |
| gnu111 | anyone has good instructions for installing dashboard + keystone in RHEL diablo? | 15:05 |
| kodapa | gnu111: it's pain in the ass | 15:05 |
| *** jsavak has joined #openstack | 15:06 | |
| kodapa | gnu111: I've spent so much time trying to get it to work | 15:06 |
| gnu111 | kodapa: :( yup that's what I am discovering. | 15:06 |
| foexle | kodapa: hahaha sign :) | 15:06 |
| gnu111 | kodapa: are you using LDAP as well? | 15:06 |
| kodapa | gnu111: no | 15:06 |
| otaku2 | wtfwtf | 15:07 |
| uksysadmin | otaku2, yes you say that alot with OpenStack | 15:07 |
| * uksysadmin ducks and covers | 15:07 | |
| gnu111 | kodapa: I might try to install it with the python instructions instead of the griddynamics rpms. | 15:08 |
| mjfork | gnu111: i have it working, with instructions | 15:09 |
| kodapa | otaku2: why wtf? | 15:09 |
| kodapa | gnu111: is there any advantage of using ldap? | 15:09 |
| gnu111 | kodapa: I have already existing users in LDAP so trying to leverage that. | 15:09 |
| kodapa | AHa | 15:09 |
| kodapa | Anyone else, is there any reason to use ldap with openstack? | 15:10 |
| *** joesavak has quit IRC | 15:10 | |
| mjfork | gnu111: what are you trying to get with dashboard/keystone? | 15:10 |
| *** joesavak has joined #openstack | 15:10 | |
| *** jsavak has quit IRC | 15:10 | |
| gnu111 | mjfork: I just did yum install and bit confused on how to proceed to the next step. I can manage most of the stuff from command line but for others I need a web interface and also need LDAP with keystone working. | 15:11 |
| *** hugokuo has quit IRC | 15:11 | |
| *** siwos has quit IRC | 15:12 | |
| *** cp16net has joined #openstack | 15:12 | |
| mjfork | gnu111: i have not done LDAP, but i can tell you i had no luck with yum packages of Dashobard/Keystone abouta month back, I installed from github source. | 15:12 |
| *** rods has quit IRC | 15:12 | |
| mjfork | gnu111: have you see http://mirantis.blogspot.com/2011/08/ldap-identity-store-for-openstack.html | 15:13 |
| gnu111 | mjfork: ok. that's what I was going to try next. | 15:13 |
| mjfork | gnu111: let me point you to some directions | 15:13 |
| gnu111 | mjfork: thanks! | 15:13 |
| mjfork | gnu111: http://forums.openstack.org/viewtopic.php?f=23&t=482&p=1745&hilit=github+horizon+keystone#p1745 | 15:14 |
| gnu111 | mjfork: ok, that looks helpful, I will try that. | 15:15 |
| *** imsplitbit has quit IRC | 15:16 | |
| mjfork | gnu111: feel free to reach out with questions, I have diablo versions of dashboard + keystone + nova + glance all running (nova on RHEL, rest on Ubunut) | 15:16 |
| uvirtbot | New bug: #904792 in openstack-manuals "Cleanup on nova-manage commands" [Undecided,New] https://launchpad.net/bugs/904792 | 15:18 |
| *** crescendo_ has quit IRC | 15:21 | |
| *** dnjaramba has quit IRC | 15:22 | |
| *** MarkAtwood has joined #openstack | 15:25 | |
| *** javiF has joined #openstack | 15:26 | |
| *** wilmoore has joined #openstack | 15:28 | |
| gnu111 | mjfork: how did you install python-pip in RHEL? it is not on the yum repo. | 15:29 |
| *** dubsquared has joined #openstack | 15:29 | |
| *** wilmoore has quit IRC | 15:30 | |
| mjfork | i did those in Ubuntu | 15:30 |
| gnu111 | mjfork: i have easy_install. is that the same thing? | 15:31 |
| mjfork | what version of RHEL? | 15:32 |
| gnu111 | mjfork: RHEL 6.2 | 15:33 |
| *** vladimir3p has joined #openstack | 15:34 | |
| mjfork | http://pkgs.org/centos-6-rhel-6/epel-i386/python-pip-0.8-1.el6.noarch.rpm.html | 15:35 |
| mjfork | could try that | 15:35 |
| reidrac | you may want to upgrade pip anyway, 0.8 it's too old | 15:35 |
| *** yshh has joined #openstack | 15:36 | |
| reidrac | pip --upgrade install pip; or something like that | 15:36 |
| mjfork | yeah, i see i am on 1.0.2 | 15:36 |
| reidrac | I've found 0.8 to be unusable, like downloading wrong versions! | 15:36 |
| mjfork | gnu111: is it erroring out on pip? if you isntall python-setuptools and run easy_install virtualenv and python tools/install_venv.py does it work? | 15:37 |
| *** AlanClark has joined #openstack | 15:38 | |
| *** deshantm_ has joined #openstack | 15:38 | |
| *** lborda has quit IRC | 15:39 | |
| *** jsavak has joined #openstack | 15:39 | |
| gnu111 | mjfork: i will try the easy_install path. how do you use pip? I installed the rpm but pip , python-pip seems to not available as a command. | 15:40 |
| *** livemoon has left #openstack | 15:41 | |
| mjfork | i ran these against ubuntu | 15:41 |
| *** JesperA has quit IRC | 15:41 | |
| mjfork | all i can say is try it on RHEL and see if it fails | 15:41 |
| gnu111 | mjfork: it installed it here: /usr/bin/pip-python | 15:41 |
| *** datajerk has quit IRC | 15:42 | |
| *** deshantm has quit IRC | 15:42 | |
| mjfork | ok, good. | 15:42 |
| *** rnorwood has joined #openstack | 15:42 | |
| *** joesavak has quit IRC | 15:43 | |
| gnu111 | mjfork: the pip install part failed. http://paste.openstack.org/show/3833/ | 15:45 |
| mjfork | looks like network conection failed? | 15:47 |
| mjfork | re run it? | 15:47 |
| *** Trebortech has joined #openstack | 15:48 | |
| gnu111 | It seems to be having issues with Downloading/unpacking coverage. re running it now. | 15:48 |
| *** lborda has joined #openstack | 15:48 | |
| gnu111 | ok. seems to be moving along. | 15:48 |
| *** deshantm_ is now known as deshantm | 15:50 | |
| Spirilis | hmm | 15:51 |
| Spirilis | anyone know what to do with this kind of error- | 15:51 |
| Spirilis | ImproperlyConfigured: Error importing middleware django_openstack.middleware.keystone: "No module named openstack" | 15:51 |
| Spirilis | found in Apache's error.log? trying to get horizon working... | 15:51 |
| Spirilis | using the git repos referenced in devstack, but, doing it by hand | 15:52 |
| *** imsplitbit has joined #openstack | 15:52 | |
| *** katkee has quit IRC | 15:52 | |
| *** jsavak has quit IRC | 15:52 | |
| otaku2 | 2011-12-15 10:52:44,259 nova: No module named MySQLdb | 15:53 |
| *** jsavak has joined #openstack | 15:53 | |
| otaku2 | this is on an openstack-nova diablo installed from rpm's - what am I missing? | 15:54 |
| *** nati2 has joined #openstack | 15:54 | |
| *** cmasseraf has joined #openstack | 15:54 | |
| mjfork | otaku2: do you have MySQL-python package installed? | 15:54 |
| otaku2 | nope. apparently it was not in the dependencies. | 15:55 |
| otaku2 | thanks. | 15:55 |
| *** rnirmal has joined #openstack | 15:56 | |
| *** adjohn has joined #openstack | 15:57 | |
| mjfork | Spirilis: are you installing from latest branch? | 15:57 |
| Trebortech | I just created a Centos Image following the instructions on the OpenStack site. When I boot the image I get a "No bootable device" error message. The image boots find in kvm locally. Any idea? | 15:58 |
| *** jsh has joined #openstack | 15:58 | |
| *** HowardRoark has joined #openstack | 15:58 | |
| gnu111 | mjfork: ok. the pip step is done. I needed to also install python-devel sqlite-devel and python-ldap. | 15:58 |
| *** huslage has quit IRC | 15:59 | |
| mjfork | gnu111: ok, good. glad you were able to work through it. | 16:00 |
| *** stuntmachine has quit IRC | 16:00 | |
| *** rocambol1 has quit IRC | 16:01 | |
| *** cp16net has quit IRC | 16:02 | |
| *** cp16net has joined #openstack | 16:02 | |
| Spirilis | mjfork: I'm not sure, using stable/diablo when I do the git checkout? | 16:02 |
| Spirilis | tbh I'm not that familiar with git yet :) | 16:02 |
| mjfork | ok, then no, not latest code. so if you want to get dashbaord up use this link: http://forums.openstack.org/viewtopic.php?f=23&t=482&p=1745&hilit=github+horizon+keystone#p1745 | 16:03 |
| *** jedi4ever has quit IRC | 16:03 | |
| mjfork | same process gnu111 is going through now | 16:03 |
| *** cp16net has quit IRC | 16:03 | |
| *** cp16net has joined #openstack | 16:03 | |
| *** cmasseraf has quit IRC | 16:03 | |
| *** datajerk has joined #openstack | 16:03 | |
| Spirilis | ah k, I did install keystone from the git repo I think, let me see what's different about the dashboard | 16:04 |
| *** TheOsprey has quit IRC | 16:04 | |
| Spirilis | thx | 16:04 |
| *** rnorwood has quit IRC | 16:04 | |
| *** cmasseraf has joined #openstack | 16:04 | |
| *** dragondm has joined #openstack | 16:05 | |
| Spirilis | ah definitely missing some of the pip reqs for keystone | 16:05 |
| *** rnorwood has joined #openstack | 16:06 | |
| *** jsavak has quit IRC | 16:07 | |
| gnu111 | mjfork: the part regarding integrating keystone into nova. "link keystone into the nova root dir" "ln -s keystone/keystone nova/keystone" where is this keystone/keystone dir? | 16:08 |
| mjfork | gnu111: where do you see that? | 16:09 |
| gnu111 | https://github.com/openstack/keystone/blob/master/README.md | 16:09 |
| mjfork | i didn't have do that | 16:10 |
| *** freeflyi1g has joined #openstack | 16:10 | |
| mjfork | refer back to the forum post for stable/diablo | 16:11 |
| gnu111 | mjfork: how does nova know about keystone? is there a nova.conf flag? | 16:11 |
| *** whit has joined #openstack | 16:11 | |
| mjfork | no, look in /etc/nova/api-paste.ini | 16:12 |
| *** whit has quit IRC | 16:12 | |
| Spirilis | werd, doing the install_venv.py for horizon now. yet something else I forgot ;) | 16:12 |
| uvirtbot | New bug: #904815 in keystone "sampledata does not work with real LDAP backend" [Undecided,New] https://launchpad.net/bugs/904815 | 16:13 |
| *** whit has joined #openstack | 16:13 | |
| mjfork | gnu111: is there a keystone entry in the pipeline | 16:13 |
| *** freeflying has quit IRC | 16:13 | |
| gnu111 | mjfork: no there isn't one. | 16:14 |
| *** whit has quit IRC | 16:14 | |
| *** maplebed has quit IRC | 16:16 | |
| *** fghaas has quit IRC | 16:16 | |
| mjfork | let me paste the conf | 16:16 |
| *** perestrelka has quit IRC | 16:16 | |
| Tristani3D | mjfork: got some time to do a little troubleshooting with me? Im trying to figure out why my storage is still red in dashboard | 16:17 |
| mjfork | gnu111: http://paste.openstack.org/show/3837/ | 16:18 |
| mjfork | Tristani3D: i cn try | 16:19 |
| Tristani3D | mjfork: ok where to start | 16:19 |
| mjfork | what "Storage"? | 16:19 |
| mjfork | nova-volume? swift? | 16:19 |
| *** cmasseraf has quit IRC | 16:20 | |
| mjfork | gnu111: that is my api-paste.ini | 16:20 |
| Tristani3D | mjfork: swift | 16:20 |
| *** cmasseraf has joined #openstack | 16:20 | |
| gnu111 | mjfork: ok. got it. how did you generate the admin token? | 16:20 |
| *** cmasseraf has quit IRC | 16:21 | |
| Spirilis | huh, the python tools/install_venv.py keeps bombing with some timeout doing a download (on eventlet, last attempt I tried) | 16:21 |
| Spirilis | oh | 16:22 |
| *** dxd828 has quit IRC | 16:22 | |
| Spirilis | betcha the supposed firewall code upgrade our network team has scheduled for today is at fault :-) | 16:22 |
| *** perestrelka has joined #openstack | 16:23 | |
| *** RobertLaptop has joined #openstack | 16:23 | |
| gnu111 | Spirilis: I am having similar issues: http://paste.openstack.org/show/3838/ | 16:23 |
| *** RobertLaptop has left #openstack | 16:23 | |
| *** reidrac has quit IRC | 16:23 | |
| Tristani3D | mjfork: where to start | 16:24 |
| *** jkyle has joined #openstack | 16:24 | |
| mjfork | gnu111: http://paste.openstack.org/show/3839/ run those commands, subsistiting your IPs | 16:25 |
| otaku2 | and do you really need swift? it seems like such a pain to set up for just a fancy httpd server with overcomplicated storage buckets. What's the point? | 16:25 |
| mjfork | Tristani3D: i haven't not done swift inside dahsbaord, but lets start with the endpoint, what does keystone-manage endpointTemplates list | 16:25 |
| uvirtbot | New bug: #904820 in openstack-ci "Gerrit should set bugs In Progress with assignee when a patchset is proposed" [Wishlist,In progress] https://launchpad.net/bugs/904820 | 16:26 |
| notmyname | otaku2: I think swift is a bit more than "a fancy httpd server with overcomplicated storage buckets" ;-) | 16:27 |
| Tristani3D | mjfork: http://pastebin.com/5C75pB0T | 16:27 |
| otaku2 | notmyname: please, enlighten me. what is the point of it then? | 16:27 |
| notmyname | otaku2: storing unstructured data that can grow without bound | 16:28 |
| notmyname | otaku2: common use cases include storing backups and storing static web content | 16:28 |
| notmyname | otaku2: http://www.buildcloudstorage.com/2011/12/if-youve-been-following-my-previous.html | 16:28 |
| *** nati2 has quit IRC | 16:29 | |
| mjfork | Tristani3D: select * from endpointTemplates in the sqliite db | 16:29 |
| mjfork | most interested in admin url | 16:29 |
| notmyname | otaku2: swift isn't designed to give you a few GB or TB of hard drive space. it's designed to allow you to store TB to PB of data cheaply and reliably | 16:30 |
| Spirilis | gnu111: lol ah k, I'm not going insane then | 16:30 |
| *** GheRivero has joined #openstack | 16:30 | |
| *** bepernoot has quit IRC | 16:30 | |
| Tristani3D | mjfork: there is no sqlite? | 16:30 |
| Tristani3D | only mysql? | 16:30 |
| *** dev_sa has joined #openstack | 16:31 | |
| otaku2 | notmyname: but something still has to know what api call to hit, it's not like I can just wget my.jpg right? | 16:31 |
| notmyname | otaku2: the API is all over HTTP, so ya, you actually can wget a file :-) | 16:31 |
| otaku2 | I guess I have a lot more RTFM'ing to do. I'm from the old pre-s3 cdn mindset | 16:32 |
| *** gyee has quit IRC | 16:33 | |
| mjfork | Tristani3D: on the Keystone node, use sqllite: sqlite3 <path to keystone.db> | 16:34 |
| otaku2 | I also don't understand what good it does for running VM's. A VM needs hard drive images that are constantly mounted. Is there a driver that mounts a root drive via s3? Sounds scary. | 16:35 |
| gnu111 | Spirilis: finally got it done after 10 timeouts. | 16:36 |
| Tristani3D | mjfork: like this? sql_connection = mysql://keystone:BSMZsjMwC4saw2V31RuH@192.168.124.81/keystone | 16:36 |
| *** lorin1 has quit IRC | 16:36 | |
| mjfork | oh, didn't realize you setup mysql for keystone, then use Mysql | 16:36 |
| *** betodalas has joined #openstack | 16:36 | |
| mjfork | eitehr way, want to the contents of endpointTemplates table | 16:36 |
| *** lorin1 has joined #openstack | 16:36 | |
| *** wilmoore has joined #openstack | 16:36 | |
| Tristani3D | yeah | 16:37 |
| Tristani3D | sqlite3 is working | 16:37 |
| betodalas | Hello, anyone know about Openstack with Xen?[ | 16:37 |
| mjfork | otaku2: ou are right, there isn't much use for swift with hard drive images | 16:37 |
| mjfork | otaku2: swift's use with nova is very limited, outside storing base VM images | 16:37 |
| Tristani3D | mjfork: never used sqlite, dont know the queries | 16:37 |
| otaku2 | so is there a reason to even install it? | 16:38 |
| mjfork | otaku2: what are ou trying to do? | 16:38 |
| Spirilis | gnu111: yeah, same here | 16:38 |
| mjfork | Tristani3D: are you connected to the DB in MySql or sqllite? | 16:38 |
| gnu111 | Spirilis: did you get this error: Error: That port is already in use. | 16:38 |
| Tristani3D | mjfork: well i am inside sqlite, want to see the db's | 16:38 |
| Tristani3D | to verify if keystone uses. sqlite or mysql | 16:39 |
| otaku2 | run vm | 16:39 |
| Spirilis | gnu111: yeah, I had "nova-ajax-console-proxy" running, I stopped that and it cleared up | 16:39 |
| mjfork | Tristani3D: do you have a file on your system named keystone.db | 16:39 |
| Spirilis | I didn't see that nova-ajax-console-proxy stuff running in the devstack environment, so, I figure it's not necessary? | 16:39 |
| otaku2 | I want a portal where people can provision vm's. thats pretty much it | 16:39 |
| mjfork | otaku2: then no, you probably don't need swift just to run VMs | 16:39 |
| Spirilis | Ok now new error: ImportError: Could not import settings 'dashboard.settings' (Is it on sys.path?): No module named dashboard.settings | 16:39 |
| Tristani3D | mjfork: MySQL it seems | 16:40 |
| mjfork | Spirilis: is this when starting the dfashboard? | 16:40 |
| otaku2 | k. so how do I run nova without it? It wants an e2 or s3 server I believe | 16:40 |
| Tristani3D | mysql> show databases; | 16:40 |
| Tristani3D | +--------------------+ | 16:40 |
| Tristani3D | | Database | | 16:40 |
| Tristani3D | +--------------------+ | 16:40 |
| Tristani3D | | information_schema | | 16:40 |
| Tristani3D | | keystone | | 16:40 |
| Tristani3D | +--------------------+ | 16:40 |
| mjfork | Tristani3D: use keystone; SELECT * FROM endpointTemplates; | 16:40 |
| Spirilis | mjfork: This is when hitting it, that error was in apache's log. I was able to do the dashboard/manage.py runserver piece without error | 16:40 |
| mjfork | never used apache...can you stay with runserver? | 16:41 |
| Tristani3D | mjfork: | endpoint_templates | you mean? | 16:41 |
| mjfork | yes. sorry. | 16:41 |
| Spirilis | I can hit it at port 8000 as it's bound directly I guess... | 16:41 |
| betodalas | jaypipes Hello | 16:41 |
| uvirtbot | New bug: #904830 in swift "Swift should search for "type" field in serviceCatalog" [Undecided,New] https://launchpad.net/bugs/904830 | 16:41 |
| Spirilis | Hmm, though on port 8000 I get: ImproperlyConfigured: Error importing middleware django_openstack.middleware.keystone: "No module named openstackx" | 16:41 |
| betodalas | *jaypipes* hello | 16:41 |
| Spirilis | I thought I'd installed that one already | 16:42 |
| mjfork | Spirilis: check the dashboard logs, guessing you need openstackx on nova-compute | 16:42 |
| Tristani3D | mjfork: http://pastebin.com/g4SJBA22 | 16:42 |
| *** mchenetz has quit IRC | 16:44 | |
| mjfork | Tristani3D: what exactly does the dashboard show? | 16:44 |
| jaypipes | betodalas: heya, ok, back now... whatup? | 16:44 |
| *** perestrelka has quit IRC | 16:45 | |
| Spirilis | hm | 16:46 |
| Tristani3D | mjfork: http://imageshack.us/photo/my-images/194/naamlooslod.png/ | 16:46 |
| Spirilis | I do have --osapi_extensions_path set in nova.conf | 16:46 |
| Spirilis | restarting the nova services for good measure, let me see now | 16:46 |
| mjfork | Spirilis: what is the path passed to --osapi_extensions_path and what is at that folder? | 16:47 |
| Spirilis | no difference. humm | 16:47 |
| Spirilis | --osapi_extensions_path=/usr/local/lib/python2.7/dist-packages/openstackx-0.2-py2.7.egg/extensions | 16:47 |
| *** llang629 has joined #openstack | 16:47 | |
| Spirilis | looks like admin.py/pyc, __init__.py/pyc | 16:47 |
| mjfork | Tristani3D: where does Swift live for yuo? | 16:48 |
| *** ppradhan has joined #openstack | 16:49 | |
| mjfork | Spirilis: on Ubuntu? | 16:49 |
| Spirilis | yeah | 16:49 |
| Tristani3D | mjfork: the swift-proxy-acct and swift-ring-compute on the same node, swift-storage on 5 other nodes | 16:49 |
| mjfork | ok, what IP though? | 16:49 |
| Tristani3D | for what service? | 16:50 |
| *** katkee has joined #openstack | 16:50 | |
| mjfork | Spirilis: those look right to me - can you pasetbin full stack? | 16:50 |
| Spirilis | though I did modify horizon's pip-requires, I think it asked for openstackx and I removed that since I'd already d/l'ed it and installed before | 16:50 |
| mjfork | did it error out? or did you remove it proactively?\ | 16:51 |
| *** perestrelka has joined #openstack | 16:51 | |
| Spirilis | Proactively | 16:51 |
| *** llang629 has left #openstack | 16:51 | |
| Spirilis | I was having enough trouble d/l'ing the packages (as gnu111 had above too) I lightened the load a bit... | 16:51 |
| Spirilis | Let me add that back and see what it does | 16:51 |
| mjfork | Spirilis: add it back and try again | 16:51 |
| Tristani3D | mjfork: IP's of the swift-proxy-acct and swift-ring-compute node: http://pastebin.com/UsEhnuk6 | 16:52 |
| *** deshantm has quit IRC | 16:52 | |
| *** betodalas has quit IRC | 16:52 | |
| mjfork | Tristani3D: pretty sure this si the problem public_url = http://192.168.122.5:8080/v1.1/AUTH_%tenant_id% and admin_url = http://192.168.124.81:8080/v1.0/AUTH_%tenant_id% | | 16:53 |
| mjfork | look @ IPs | 16:53 |
| *** Leander has joined #openstack | 16:54 | |
| Tristani3D | 122.5 is from the swift-proxy-acct and swift-ring-compute node / public: eth0: 192.168.122.5 | 16:54 |
| mjfork | i think the admin_url needs udpated | 16:54 |
| mjfork | and itnernal_url | 16:54 |
| *** dev_sa has left #openstack | 16:55 | |
| Leander | hello! I'm in need of som help regarding the configuration of openstack with keystone | 16:55 |
| Tristani3D | dont follow, to what? | 16:55 |
| Leander | some* | 16:55 |
| Spirilis | ah still rocking timeout hell, let me try it a few more times | 16:56 |
| Spirilis | or loop it with a script until it returns exit code 0 lol | 16:56 |
| Leander | i have configured all the services according to the page: http://keystone.openstack.org/configuringservices.html (except for swift which is not installed at the moment), however i'm not sure how test if the integration is complete | 16:56 |
| mjfork | Tristani3D: nevermind, i see now. can dashbaord telnet to http://192.168.124.81:8080/v1.0/AUTH_%tenant_id% | 16:57 |
| Leander | i've tried to run the following command "sudo nova-manage project zipfile demo admin", i have both the tenant demo and the user admin configured there | 16:57 |
| Spirilis | hmm, connection refused when doing the git's for git://github.com/rackspace/python-novaclient.git | 16:58 |
| Spirilis | that might be our internal firewalling at fault, not sure | 16:58 |
| Tristani3D | mjfork: root@d00-0c-29-40-bc-63:~# telnet http://192.168.124.81:8080/v1.0/AUTH_%tenant_id% | 16:58 |
| Tristani3D | telnet: could not resolve http://192.168.124.81:8080/v1.0/AUTH_%tenant_id%/telnet: Temporary failure in name resolution | 16:58 |
| Leander | however, it keeps failing and according to the logs it keeps performing a database look up | 16:58 |
| Leander | am i missing something | 16:58 |
| Leander | ? | 16:58 |
| mjfork | Tristani3D: telnet 192.168.124.81 8080 | 16:58 |
| Spirilis | well no good reason I can't switch that to using https I guess | 16:58 |
| mjfork | Leander: i don't beleive zipfile works with keystone | 16:58 |
| *** Ramonster has quit IRC | 16:58 | |
| Tristani3D | root@d00-0c-29-40-bc-63:~# telnet 192.168.124.81 8080 | 16:58 |
| Tristani3D | Trying 192.168.124.81... | 16:58 |
| Tristani3D | Connected to 192.168.124.81. | 16:58 |
| Tristani3D | Escape character is '^]' | 16:58 |
| *** cloudgeek has quit IRC | 16:59 | |
| Leander | then how can use the euca commands? | 16:59 |
| Leander | or do they not work as well with keystone? | 17:00 |
| mjfork | Leander: they do not work well with keystone, you can search the mailing list archives for more deatils | 17:00 |
| mjfork | Tristani3D: do you see anything in dashbaord logs? | 17:02 |
| *** Oneiroi has quit IRC | 17:03 | |
| *** lorin1 has quit IRC | 17:03 | |
| *** rods has joined #openstack | 17:03 | |
| Spirilis | mjfork: hot dang, that worked :) | 17:04 |
| *** Ryan_Lane has quit IRC | 17:05 | |
| Kiall | Leander / mjfork: all the euca commands bar those needing a pem cert (ie image upload) work fine with keystone | 17:05 |
| Tristani3D | mjfork: where are the dashboard logs? | 17:05 |
| *** nati2 has joined #openstack | 17:05 | |
| mjfork | Tristani3D: how do you start dashbaord? | 17:06 |
| Tristani3D | its started by crowbar | 17:06 |
| mjfork | Tristani3D: in that case...i don't know :-) search for something like dashboard.log? | 17:06 |
| Spirilis | GUI looks nice... just gotta see why nova-volume and objectstore show up red now | 17:06 |
| mjfork | Kiall: is it the standand zipfile create to get ecua working? | 17:06 |
| Spirilis | ah nova-volume of course, the LVM volume group isn't active | 17:07 |
| Kiall | mjfork: no, none of the "nova-manage project" or "nova-manage user" commands should be used when you use keystone | 17:07 |
| mjfork | Kiall: what is required then to use euca commands with keystone? | 17:07 |
| Kiall | (Really, those commands should disappear if keystone is used as they are useless and cause confusion) | 17:07 |
| Leander | Kial: then how exactly do i proceed? | 17:08 |
| Kiall | Create a novarc file, something like this: https://github.com/managedit/openstack-setup/blob/master/openrc.tmpl | 17:08 |
| Kiall | And then add EC2 credentials to keystone like this: https://github.com/managedit/openstack-setup/blob/master/keystone_data.sh.tmpl#L36 | 17:09 |
| *** katkee has quit IRC | 17:11 | |
| *** ninkotech has joined #openstack | 17:13 | |
| *** RobertLaptop has joined #openstack | 17:13 | |
| Leander | according to the example configuration in "http://keystone.openstack.org/configuringservices.html", the NOVA_REGION should be RegionOne right? | 17:14 |
| Leander | (from https://github.com/managedit/openstack-setup/blob/master/openrc.tmpl) | 17:14 |
| *** RobertLaptop has quit IRC | 17:14 | |
| *** clauden has joined #openstack | 17:15 | |
| Tristani3D | mjfork: gotta run, thx sofar | 17:16 |
| *** Tristani3D has quit IRC | 17:16 | |
| *** maplebed has joined #openstack | 17:17 | |
| *** arBmind has quit IRC | 17:17 | |
| Kiall | Leander: it should be set to whatever you have in nova.conf and keystone.. | 17:17 |
| Spirilis | hot damn, Houston we have liftoff! | 17:17 |
| Spirilis | Got a KVM process started from the "tty" default image that devstack uses... | 17:17 |
| Kiall | That doc suggests its RegionOne, but I'm betting nova.conf has something different since that doc doesnt explain that they need to match | 17:17 |
| Spirilis | All done through the GUI too... | 17:17 |
| mjfork | Leander: run keystone-manage endpointTemplates list | 17:18 |
| Leander | i'm checing my nova.conf and i can't find any references to regions | 17:19 |
| Leander | novaRegionOnehttp://172.16.225.138:8774/v1.1/ glanceRegionOnehttp://172.16.225.138:9292/v1.1/ keystoneRegionOnehttp://http://172.16.225.138:6000/v2.0 | 17:19 |
| *** jeblair has quit IRC | 17:19 | |
| Leander | (ouput from keystone-manage) | 17:19 |
| mjfork | use RegionOne | 17:19 |
| *** woleium has joined #openstack | 17:19 | |
| Spirilis | and I am able to login to that instance. Wow! | 17:19 |
| Spirilis | mjfork: Thank you so much for your assistance. Now I'm gonna document everything I did, blow away the server and redo it from scratch... crossing fingers... lol | 17:20 |
| mjfork | good luck - what is your goal? developemnt? POC? | 17:20 |
| *** mnour has joined #openstack | 17:21 | |
| *** deshantm has joined #openstack | 17:21 | |
| *** huslage has joined #openstack | 17:22 | |
| *** huslage has quit IRC | 17:22 | |
| Leander | mjfork: thks. Do you know where the euca-* tools store their logs, if any? i think i have one port that's not well configured since i got a "[Errno 111] Connection refused" as a result | 17:24 |
| *** jkyle_ has joined #openstack | 17:24 | |
| *** uksysadmin has quit IRC | 17:24 | |
| *** jkyle has quit IRC | 17:24 | |
| mjfork | Leander: what are you trying todo? | 17:24 |
| Leander | euca-describe-instances | 17:25 |
| Leander | i'm trying to check if everything is well configured | 17:25 |
| Spirilis | POC for now... | 17:26 |
| *** jeblair has joined #openstack | 17:26 | |
| Spirilis | The end-goal is to have a dev environment for our web applications folks | 17:26 |
| mjfork | Spirilis: i would suggest not using devstack the | 17:26 |
| Spirilis | Yeah | 17:26 |
| Spirilis | That's why I'm rolling it by hand to get down into the muck of things | 17:27 |
| Spirilis | Devstack did at least provide a good learning reference | 17:27 |
| Spirilis | I had a lot of fun with Swift (cactus), we have a POC of that running at the moment, nobody's ponying up cash to build a real cluster though | 17:28 |
| Spirilis | so it's mostly a toy for us IT/storage folks | 17:28 |
| Spirilis | I think the nova cluster will be the first foot in the door | 17:28 |
| *** andrewsben has joined #openstack | 17:31 | |
| *** vdo has quit IRC | 17:32 | |
| Leander | mjfork: i'm trying to check if everything is well configured and ran euca-describe-instances | 17:32 |
| *** bepernoot has joined #openstack | 17:33 | |
| *** jkyle_ has quit IRC | 17:33 | |
| *** Razique has quit IRC | 17:35 | |
| uvirtbot | New bug: #904850 in glance "Glance service selection from ServiceCatalog when using keystone" [Undecided,New] https://launchpad.net/bugs/904850 | 17:36 |
| mjfork | Spirilis: very good - good luck! | 17:36 |
| *** pixelbeat has quit IRC | 17:37 | |
| *** MarcMorata has joined #openstack | 17:38 | |
| *** pixelbeat has joined #openstack | 17:39 | |
| *** dotdevops has joined #openstack | 17:41 | |
| *** anpl has quit IRC | 17:42 | |
| *** lorin1 has joined #openstack | 17:43 | |
| *** vladimir3p has quit IRC | 17:44 | |
| *** sdake has quit IRC | 17:45 | |
| *** sdake_ has joined #openstack | 17:45 | |
| *** koolhead11 has quit IRC | 17:45 | |
| Leander | are the ports 6000/6001 reserved for something on linux? | 17:45 |
| *** andrewbogott has joined #openstack | 17:45 | |
| *** deshantm has quit IRC | 17:45 | |
| Leander | keystone wouldn't launch with 5000/5001 so i had to change them to 6000/60001 | 17:45 |
| *** sdake_ has quit IRC | 17:46 | |
| *** jdurgin has joined #openstack | 17:46 | |
| Spirilis | 6000+ is often used for the X11 display | 17:47 |
| Spirilis | if your server doesn't have X running on its console then it shouldn't be a problem | 17:47 |
| *** sdake has joined #openstack | 17:47 | |
| Leander | it's running on a server | 17:47 |
| *** adjohn has quit IRC | 17:47 | |
| Leander | if that's the case i'm not sure what the euca commands can't dial to | 17:48 |
| Leander | connect* | 17:48 |
| Leander | i always get a connection refused error | 17:48 |
| mjfork | Leander: what command are you running? euca-describe-instances? | 17:48 |
| Leander | yes | 17:48 |
| mjfork | pastebin the output and novarc file | 17:49 |
| *** sdake has quit IRC | 17:50 | |
| Leander | output: [Errno 111] Connection refused | 17:50 |
| *** sdake has joined #openstack | 17:50 | |
| mjfork | ok, and pastebin novsrc | 17:51 |
| Leander | novarc: NOVA_API_HOST=172.16.225.138 GLANCE_API_HOST=172.16.225.138 KEYSTONE_API_HOST=172.16.225.138 KEYSTONE_TENANT="admin" KEYSTONE_USERNAME="admin" KEYSTONE_PASSWORD="nova" NOVA_REGION=RegionOne # Common Stuff Dont change export NOVA_USERNAME=$KEYSTONE_USERNAME export NOVA_PROJECT_ID=$KEYSTONE_TENANT export NOVA_PASSWORD=$KEYSTONE_PASSWORD export NOVA_API_KEY=$KEYSTONE_PASSWORD export NOVA_REGION_NAME=$NOVA_REGION export NOV | 17:51 |
| Leander | export OS_AUTH_USER=$KEYSTONE_USERNAME export OS_AUTH_KEY=$KEYSTONE_PASSWORD export OS_AUTH_TENANT=$KEYSTONE_TENANT export OS_AUTH_URL="http://${KEYSTONE_API_HOST}:6000/v2.0/" export OS_AUTH_STRATEGY="keystone" | 17:51 |
| Leander | export EC2_URL="http://${NOVA_API_HOST}:8773/services/Cloud" export EC2_ACCESS_KEY="$KEYSTONE_USERNAME:$KEYSTONE_TENANT" export EC2_SECRET_KEY=$KEYSTONE_PASSWORD export S3_URL="http://${GLANCE_API_HOST}:3333" | 17:51 |
| *** yshh has quit IRC | 17:54 | |
| mjfork | Leander: please paste it at paste.openstack.org | 17:54 |
| Leander | mjfork: sorry about that, http://paste.openstack.org/show/3843/ | 17:55 |
| mjfork | can you telnet 172.16.225.138 8773 on machine running ecua | 17:56 |
| *** huslage has joined #openstack | 17:57 | |
| *** RobertLaptop has joined #openstack | 17:57 | |
| Leander | nope, connection refused | 17:58 |
| *** RobertLaptop has left #openstack | 17:58 | |
| mjfork | there is your issue | 17:58 |
| mjfork | netstat -tnap | grep LISTEN | grep 8773 | 17:59 |
| mjfork | anything there? | 17:59 |
| Leander | empty result | 17:59 |
| *** MarkAtwood has quit IRC | 17:59 | |
| mjfork | so, nothing is liistening on 8773 | 18:00 |
| mjfork | on yor controlleris nova-api listening? | 18:00 |
| Leander | this is my nova.conf btw, http://paste.openstack.org/show/3844/ | 18:00 |
| uvirtbot | New bug: #904861 in horizon "Tenant id error on syspanel overview page." [Undecided,New] https://launchpad.net/bugs/904861 | 18:01 |
| *** pixelbeat has quit IRC | 18:01 | |
| uvirtbot | New bug: #904863 in glance "glance-manage db_sync fails with "no such option: add_group"" [Undecided,New] https://launchpad.net/bugs/904863 | 18:01 |
| mjfork | why did you set ec2_dmz_host? | 18:01 |
| mjfork | do you have multiple nodes? | 18:01 |
| Leander | at the moment i do not | 18:02 |
| mjfork | also - can you verify nova-api is running? | 18:02 |
| Leander | i do that with nova-manage service list right? | 18:02 |
| *** rnorwood has quit IRC | 18:03 | |
| mjfork | ps auxww | grep nova-api | 18:04 |
| *** MarkAtwood has joined #openstack | 18:04 | |
| Leander | not running | 18:05 |
| mjfork | start that service | 18:05 |
| *** bepernoot has quit IRC | 18:05 | |
| mjfork | are you a single node? | 18:05 |
| Leander | yes | 18:05 |
| Spirilis | Hmm, so it appears nova-api is returning an http 400 when I run euca-describe-images | 18:05 |
| Spirilis | or euca-describe-instances | 18:05 |
| *** nacx has quit IRC | 18:06 | |
| Leander | nova-api is not launching | 18:06 |
| Leander | the log say something like "CRITICAL nova [-] No module named ec2_token" | 18:06 |
| Leander | says* | 18:06 |
| mjfork | ok, sounds like your paste-api is not right | 18:06 |
| Leander | api-paste.ini > http://paste.openstack.org/show/3845/ | 18:09 |
| *** jdg has joined #openstack | 18:09 | |
| *** spiffxp has joined #openstack | 18:09 | |
| *** Hakon|mbp has quit IRC | 18:10 | |
| *** bepernoot has joined #openstack | 18:10 | |
| *** mchenetz has joined #openstack | 18:10 | |
| *** ejat has joined #openstack | 18:10 | |
| *** ejat has joined #openstack | 18:10 | |
| mjfork | [filter:totoken] | 18:11 |
| mjfork | paste.filter_factory = keystone.middleware.ec2_token:EC2Token.factory | 18:11 |
| mjfork | guessing that line, try googling it. | 18:11 |
| *** TheOsprey has joined #openstack | 18:12 | |
| *** whit has joined #openstack | 18:12 | |
| *** koolhead17 has joined #openstack | 18:13 | |
| *** opensanjib has joined #openstack | 18:18 | |
| opensanjib | hi all i have just installed openstack in a virtual system ubuntu 11.10 using devstack script...but i could not login to dashboard | 18:20 |
| *** bepernoot has quit IRC | 18:20 | |
| *** rustam has quit IRC | 18:21 | |
| *** Ryan_Lane has joined #openstack | 18:22 | |
| *** nati2_ has joined #openstack | 18:22 | |
| *** nati2 has quit IRC | 18:25 | |
| Leander | mjfork: i've tried to see if this was related to the issue http://osdir.com/ml/openstack-cloud-computing/2011-10/msg00704.html | 18:25 |
| Leander | but when i checked that file i did not find any "ec2_token.py" | 18:26 |
| Leander | could it be that the version of keystone bundled with ubuntu 11.10 is outdated by now? | 18:26 |
| mjfork | Leander: what do you have in keystone/middleware dir? | 18:29 |
| Leander | auth_basic.py auth_token.pyc nova_auth_token.py swift_auth.pyc auth_basic.pyc glance_auth_token.py nova_auth_token.pyc url.py auth_openid.py glance_auth_token.pyc remoteauth.py url.pyc auth_openid.pyc __init__.py remoteauth.pyc auth_token.py __init__.pyc swift_auth.py | 18:29 |
| opensanjib | does anyone know default user name and password for openstack dashboard that has been installed using devstack.org script | 18:30 |
| mjfork | where did you install from? | 18:30 |
| Leander | apt-get | 18:30 |
| mjfork | diablO? | 18:30 |
| Leander | i think it's the latest release | 18:30 |
| *** localhost has quit IRC | 18:30 | |
| mjfork | Leander: i see that file in both stable/diable and maseter - no idea why you don't ahve it | 18:31 |
| mjfork | but you need it | 18:31 |
| Leander | i'm clueless as well | 18:31 |
| *** localhost has joined #openstack | 18:32 | |
| mjfork | Leander: grab it from github and drop it in, see what happens. | 18:32 |
| *** huslage has quit IRC | 18:35 | |
| *** jog0 has joined #openstack | 18:36 | |
| *** cfant has joined #openstack | 18:36 | |
| *** mchenetz has quit IRC | 18:38 | |
| *** rnorwood has joined #openstack | 18:39 | |
| *** troytoman-away is now known as troytoman | 18:39 | |
| *** mattray has joined #openstack | 18:40 | |
| *** jakedahn has quit IRC | 18:42 | |
| *** lxu has joined #openstack | 18:43 | |
| lxu | hi | 18:43 |
| Leander | mjfork: nova-api is now running | 18:43 |
| Leander | i was also missing nova_keystone_context for some reason | 18:43 |
| *** jog0 has left #openstack | 18:44 | |
| *** mchenetz has joined #openstack | 18:44 | |
| mjfork | ok | 18:45 |
| Leander | now it get the following: Warning: failed to parse error message from AWS: <unknown>:1:0: syntax error \n None: None | 18:47 |
| Leander | when i run euca-describe-instances | 18:48 |
| *** adjohn has joined #openstack | 18:49 | |
| *** GheRivero has quit IRC | 18:49 | |
| *** opensanjib has quit IRC | 18:50 | |
| *** mchenetz has quit IRC | 18:50 | |
| *** mchenetz has joined #openstack | 18:52 | |
| mjfork | check nova-api.log | 18:54 |
| *** Tony_ has joined #openstack | 18:55 | |
| Tony_ | Hello all! | 18:55 |
| *** lloydde has joined #openstack | 18:56 | |
| *** sdake has quit IRC | 18:57 | |
| Leander | this is all i get INFO nova.api [-] 0.72431s 172.16.225.138 POST /services/Cloud/ None:None 400 [Boto/2.0 (linux2)] application/x-www-form-urlencoded text/plain | 18:58 |
| *** Tony_ has quit IRC | 18:59 | |
| *** javiF has quit IRC | 18:59 | |
| jdg | Leander: I was just getting ready to ask the same question! | 18:59 |
| mjfork | sorry, i don't think i can be of much more help here. i don't use euca2ools with Keystone. | 19:00 |
| mjfork | best bet is general mailing list or launchpad answers | 19:00 |
| Leander | thanks for all your help so far!! | 19:00 |
| mjfork | Leander: no porblem, too bad i couldn't finish it off. | 19:01 |
| uvirtbot | New bug: #904888 in nova "nova db sync failure on precise" [Undecided,New] https://launchpad.net/bugs/904888 | 19:02 |
| *** rnorwood has quit IRC | 19:05 | |
| *** dpippenger has joined #openstack | 19:07 | |
| Leander | mjfork: You brought me further than i possibly could | 19:07 |
| Leander | well, i got to run | 19:07 |
| Leander | thanks for all the help | 19:07 |
| *** Leander has quit IRC | 19:08 | |
| *** rnorwood has joined #openstack | 19:08 | |
| *** jakedahn has joined #openstack | 19:11 | |
| *** fghaas has joined #openstack | 19:11 | |
| *** zykes- has quit IRC | 19:11 | |
| *** jakedahn has quit IRC | 19:12 | |
| Spirilis | yeah I'm getting that same error | 19:13 |
| Spirilis | can't use euca-describe-images or anything else with it | 19:13 |
| Spirilis | of course the other thing is I'm not entirely sure what EC2_* variables to set | 19:13 |
| Spirilis | with the keystone backended version | 19:13 |
| Spirilis | keystone's got some token feature but I'm not sure how to craft the env variables from that | 19:13 |
| *** darraghb has quit IRC | 19:14 | |
| mjfork | can you use the equivalent nova commands in place of euca | 19:14 |
| Spirilis | hmm | 19:15 |
| Spirilis | are those based on nova-manage? | 19:15 |
| mjfork | what are you trying to do? | 19:15 |
| Spirilis | view what images I have available, and maybe launch an image from CLI | 19:15 |
| Spirilis | both of those functions work OK in the Horizon gui so far | 19:16 |
| Spirilis | also, add a public IP to a node, and add a nova-volume disk volume | 19:16 |
| mjfork | ok, i think nova & nova-manage can do some of that | 19:17 |
| mjfork | but not all | 19:17 |
| Spirilis | man it's just so confusing b/c the openstack docs all use euca-* commands in their examples, but I haven't seen anything about configuring the EC2_* variables that these cmds need | 19:18 |
| mjfork | yeah, did you verify all the entries you have in novarc are valid hosts/ports? | 19:18 |
| *** katkee has joined #openstack | 19:20 | |
| Spirilis | novarc? | 19:20 |
| Spirilis | probably need to generate that first :) | 19:20 |
| annegentle | Spirilis: yeah we've got a doc bug logged about taking out all euca commands and replacing with nova commands | 19:20 |
| Spirilis | are the euca-* commands being deprecated (i.e. new installations, should we ignore them for now?) | 19:21 |
| mjfork | annegentle: great news! that woudl be great | 19:21 |
| annegentle | Spirilis: http://docs.openstack.org/diablo/openstack-compute/admin/content/creating-certifications.html is where you learn how to create EC2 creds | 19:22 |
| mjfork | comes up alot on the forums and here. | 19:22 |
| *** crayon_z has quit IRC | 19:22 | |
| Spirilis | annegentle: ah k thx :) Does the fact that I'm using keystone affect any of that? | 19:22 |
| annegentle | Spirilis: well it's more nuanced than that, as I understand it. If you're using "deprecated auth" then you can continue to use EC2 and euca commands. If you're using Keystone, you want nova client commands. | 19:22 |
| Spirilis | cause I noticed nova-manage user list ... didn't show any users | 19:22 |
| Spirilis | ahh k | 19:23 |
| annegentle | Spirilis: yup, with Keystone nova-manage is also deprecated | 19:23 |
| annegentle | Spirilis: and yeah, not all docs describe this nuance :) | 19:23 |
| Spirilis | ahh k it's all making more sense now | 19:23 |
| *** jakedahn has joined #openstack | 19:24 | |
| Spirilis | is there a good command reference out there for the nova CLI? | 19:24 |
| *** lloydde has quit IRC | 19:24 | |
| *** dirkx_ has joined #openstack | 19:25 | |
| Spirilis | er nm, "nova help" seems to work :-) | 19:26 |
| annegentle | Spirilis: yeah that's your best one for now :) | 19:27 |
| *** llang629 has joined #openstack | 19:28 | |
| *** zykes- has joined #openstack | 19:29 | |
| *** mchenetz has quit IRC | 19:29 | |
| *** mchenetz has joined #openstack | 19:29 | |
| Spirilis | so to fill out the nova auth variables/options, username and password are obvious, what do you typically use for --apikey ? | 19:30 |
| *** stuntmachine has joined #openstack | 19:31 | |
| *** map_nw has quit IRC | 19:32 | |
| Spirilis | nm, figured it out, had to point my --url to Keystone rather than nova itself | 19:35 |
| *** lxu has quit IRC | 19:36 | |
| *** llang629 has left #openstack | 19:39 | |
| ppradhan | mjfork: hi. what might be the reason the brigde interface on compute node doesnot get the IP. One node gets the IP but the 2nd node doesnot | 19:39 |
| mjfork | i what IP do you expect to be assigned to it? | 19:40 |
| ppradhan | mjfork: 10.0.0.x | 19:40 |
| ppradhan | the 1st node gets 10.0.0.1 | 19:40 |
| *** woleium has quit IRC | 19:40 | |
| ppradhan | 2nd node has no ip | 19:40 |
| mjfork | 1st node = compute node? guest running on a compute node? | 19:40 |
| *** woleium has joined #openstack | 19:40 | |
| mjfork | 1st node= controller? | 19:40 |
| ppradhan | 1st node controller, 2nd node compute | 19:41 |
| ppradhan | guest runs on 2nd node | 19:41 |
| mjfork | ok, and the 1st node has a 10.0.0.1 IP? | 19:41 |
| mjfork | the controller that is | 19:42 |
| ppradhan | yes correct | 19:42 |
| *** vladimir3p has joined #openstack | 19:42 | |
| Spirilis | sweet, successfully created a nova-volume disk & attached it | 19:42 |
| mjfork | using euca commands? | 19:43 |
| mjfork | ppradhan: what networkign? | 19:43 |
| ppradhan | vlan | 19:43 |
| ppradhan | interafces are up | 19:43 |
| mjfork | i am guessing 10.0.0.1 IP is your VLAN gateway and controller node got it because it is running nova-network | 19:44 |
| ppradhan | if you have lets say 5 nodes.. every nodes' br100 interface should get the IP? | 19:45 |
| mjfork | no | 19:45 |
| mjfork | 10.0.0.1 is a client network | 19:45 |
| mjfork | you would not want your infrastrecuture running on it and accessible to clients | 19:45 |
| ppradhan | only that runs the nova-network should get it? | 19:45 |
| mjfork | yes, because it is the gateway node | 19:45 |
| Spirilis | mjfork: nah using the 'nova' cmd | 19:45 |
| *** jog0 has joined #openstack | 19:46 | |
| ppradhan | mjfork: ok | 19:46 |
| mjfork | ok, good. nova create-volumes? | 19:46 |
| *** Draft has joined #openstack | 19:46 | |
| Spirilis | yeah nova volume-create, volume-attach | 19:46 |
| *** jog0 has left #openstack | 19:46 | |
| Spirilis | now I'm messing with deleting a volume, detaching one already attached, seems to work fine | 19:46 |
| mjfork | i need a newer nova command | 19:46 |
| mjfork | mine doesn't have the volume-* | 19:46 |
| Spirilis | though deleting a volume took a while | 19:46 |
| Spirilis | yeah I think I got python-novaclient from git | 19:46 |
| *** vladimir3p has quit IRC | 19:46 | |
| *** fghaas has left #openstack | 19:46 | |
| *** alekibango has joined #openstack | 19:46 | |
| Spirilis | the nova cmd looks so much simpler to understand than the multitude of euca-* cmds | 19:47 |
| Spirilis | only oddball is you have to specify numeric IDs for the servers, volumes, etc. and can't specify them by name. I'm hoping that'll be fixed in the future ;) | 19:47 |
| gnu111 | mjfork: I got the dashboard web interface running. Now, tacking the user management part. I already have a admin user in nova. But i need to create a new one right? | 19:49 |
| mjfork | hmm, i pastebin'ed command ealirer for you right? | 19:51 |
| mjfork | to setup admin user? | 19:51 |
| Draft | Hello. I want to use the nova API to start an instance but in a specific zone. With ec2 API I can use "euca-run-instances -z zoneA xxxxx" but how to do it with nova client? Is it supported? | 19:53 |
| *** bcwaldon has quit IRC | 19:53 | |
| *** whit has quit IRC | 19:53 | |
| gnu111 | mjfork: yes. this one: http://paste.openstack.org/show/3839/ | 19:54 |
| gnu111 | mjfork: how do you setup the password? | 19:55 |
| *** yshh has joined #openstack | 19:55 | |
| uvirtbot | New bug: #904916 in horizon "Allocate floating ips does not work" [High,Fix committed] https://launchpad.net/bugs/904916 | 19:56 |
| mjfork | log into the dashboard and create the users | 19:56 |
| mjfork | thats what i do | 19:57 |
| gnu111 | mjfork: not sure which user/pass to use to login? | 19:57 |
| gnu111 | keystone-manage user add admin admin is this the command for admin user/pass? | 19:58 |
| mjfork | admin/admin for dshaboard | 19:59 |
| *** yshh has quit IRC | 19:59 | |
| *** bcwaldon has joined #openstack | 19:59 | |
| *** siwos has joined #openstack | 19:59 | |
| gnu111 | mjfork: I get Error authenticating with keystone: [Errno 111] ECONNREFUSED | 19:59 |
| mjfork | in the dhasbord? | 20:00 |
| gnu111 | yes. | 20:00 |
| mjfork | pastebin stacktrace | 20:00 |
| gnu111 | mjfork: from the dashboard log? not sure how to get the stacktrace. | 20:02 |
| mjfork | yes, from dashbaord log | 20:03 |
| mjfork | alternaitive is from the screen | 20:04 |
| *** praefect has quit IRC | 20:04 | |
| gnu111 | mjfork: my keystone wasn't running. But now I am getting 401 http://paste.openstack.org/show/3848/ | 20:05 |
| *** nati2_ has quit IRC | 20:05 | |
| *** jdg has quit IRC | 20:06 | |
| *** jdg has joined #openstack | 20:06 | |
| *** vlan-os has quit IRC | 20:06 | |
| *** Draft has quit IRC | 20:07 | |
| Spirilis | hmmm | 20:07 |
| Spirilis | floating ips assign correctly it seems, but, I can't actually get to it from outside the compute/control/network server (they're all the same atm) | 20:08 |
| *** zigo has quit IRC | 20:08 | |
| Spirilis | got it working so it assigns floating ips to interface br100, which links to eth0 (the only interface on the server), but can't ping that or hit port 22 | 20:08 |
| Spirilis | iptables shows some DNAT rules set up for it that reroute to 192.168.10.2 | 20:09 |
| Spirilis | looks like the ARP is working, hosts outside that server correctly arp the MAC addr of the server for that floating IP | 20:09 |
| *** pothos_ has joined #openstack | 20:09 | |
| *** pothos has quit IRC | 20:11 | |
| gnu111 | Spirilis: did you do euca-authorize? | 20:13 |
| *** pothos_ has quit IRC | 20:13 | |
| *** pothos has joined #openstack | 20:13 | |
| Spirilis | nah, probably haven't done the 'nova' cli equiv anyway | 20:14 |
| Spirilis | let me look at that | 20:14 |
| *** Draft has joined #openstack | 20:14 | |
| mjfork | gnu111: what is your token specified in the dasboard's local settings | 20:14 |
| gnu111 | mjfork: ah! I forgot that part! | 20:15 |
| *** dprince has quit IRC | 20:16 | |
| *** mchenetz has quit IRC | 20:16 | |
| *** pothos has quit IRC | 20:16 | |
| *** tmichael has quit IRC | 20:16 | |
| *** pothos has joined #openstack | 20:16 | |
| *** siwos has quit IRC | 20:17 | |
| mjfork | Spirilis: what is your problem ? | 20:18 |
| *** Draft has quit IRC | 20:18 | |
| *** pothos has quit IRC | 20:19 | |
| *** woleium_ has joined #openstack | 20:19 | |
| *** pothos has joined #openstack | 20:19 | |
| Spirilis | mjfork: I assigned a floating IP to a running VM, but can't actually communicate with that floating IP from the compute host itself or from a host outside that openstack node (but on the same physical LAN segment, and shares the same subnet as the openstack node's main IP) | 20:20 |
| Spirilis | just started playing with the secgroup subcmds to authorize traffic | 20:20 |
| Spirilis | still doesn't seem to be helping though | 20:20 |
| uvirtbot | New bug: #904932 in nova "test_linux_net creates directories and files" [Undecided,New] https://launchpad.net/bugs/904932 | 20:21 |
| mjfork | Spirilis: VLANs? | 20:21 |
| *** pothos has quit IRC | 20:22 | |
| *** pothos has joined #openstack | 20:22 | |
| Spirilis | mjfork: not using VLANs in my openstack env, just FlatDHCPManager ... we use vlans in our larger enterprise environment, but this server has no awareness of it (its port isn't a trunk link) | 20:23 |
| *** woleium has quit IRC | 20:23 | |
| *** woleium_ is now known as woleium | 20:23 | |
| mjfork | Spirilis: assuming routing, etc is setup right for the floating IPs? did you run a tcpdump to see if they hit the nova-network controller | 20:24 |
| *** map_nw has joined #openstack | 20:25 | |
| Spirilis | let me try that | 20:25 |
| *** pothos has quit IRC | 20:25 | |
| *** pothos has joined #openstack | 20:25 | |
| Spirilis | just to recap real quick though, tell me if this doesn't make sense-- my server's main IP is 10.104.51.142, the "floating IP" subnet I added (with nova-manage floating create) is 10.104.51.192/28, the IP I've assigned is 10.104.51.192 | 20:26 |
| Spirilis | nothing else using that .192+ range on the network | 20:27 |
| Spirilis | my fixed IPs for the VMs is 192.168.10.0/24 | 20:27 |
| *** bsza has joined #openstack | 20:28 | |
| *** pothos has quit IRC | 20:28 | |
| *** pothos has joined #openstack | 20:28 | |
| *** map_nw has quit IRC | 20:29 | |
| Spirilis | yep traffic is making it through the openstack node's eth0 for sure | 20:29 |
| *** dirkx_ has quit IRC | 20:29 | |
| *** siwos has joined #openstack | 20:30 | |
| *** AhmedSoliman has joined #openstack | 20:30 | |
| *** dirkx_ has joined #openstack | 20:31 | |
| *** pothos has quit IRC | 20:31 | |
| *** pothos has joined #openstack | 20:31 | |
| mjfork | is it hitting the compute nodes? | 20:32 |
| Spirilis | ah I see one problem, the traffic isn't showing up on the bridge | 20:33 |
| Spirilis | eth0 sees the incoming traffic but br100 doesn't | 20:33 |
| mjfork | is iP_forwarding on and rp_filter in ssysctl | 20:33 |
| *** siwos has quit IRC | 20:34 | |
| *** pothos has quit IRC | 20:34 | |
| *** pothos has joined #openstack | 20:34 | |
| Spirilis | looks like it is, and rp_filter is enabled as well | 20:34 |
| mjfork | try rp-filter = 2? | 20:37 |
| *** pothos has quit IRC | 20:37 | |
| *** pothos has joined #openstack | 20:37 | |
| Spirilis | didn't seem to make any difference | 20:38 |
| Spirilis | one thing I have noticed though is that I don't see any iptables rules for the secgroup-add-rule cmd for tcp port 22 | 20:39 |
| *** rnorwood has quit IRC | 20:40 | |
| mjfork | Spirilis: i need to run, good lujck. back later if you need me. | 20:40 |
| Spirilis | errr actually I take that back | 20:40 |
| Spirilis | mjfork: thanks for all your help man! | 20:41 |
| Spirilis | this almost sounds like a bug, the iptables rule was updated for instance #3 but not instance #2 when I modified the secgroup | 20:41 |
| *** tmichael has joined #openstack | 20:45 | |
| *** bcwaldon has quit IRC | 20:45 | |
| Spirilis | boom, manually added the iptables rules to instance #2 (which holds the floating IP) identical to what instance #3 saw after I ran nova secgroup-add-rule | 20:46 |
| Spirilis | now I can ping and ssh from outside (pings return dupes though, probably since the server also registers the IP on its own interfaces) | 20:46 |
| Spirilis | sweet! | 20:47 |
| Spirilis | annegentle: also much thanks to you for clarifying things :) | 20:47 |
| Spirilis | this is exciting, now I get to build it all out again from scratch to see how much I learned ;) | 20:47 |
| *** byeager has quit IRC | 20:47 | |
| *** jakedahn_ has joined #openstack | 20:53 | |
| nhm | Forgive me if this has been answered on the web somewhere, but can keystone provide auth services for existing users in ldap? | 20:54 |
| *** spiffxp has quit IRC | 20:55 | |
| *** byeager has joined #openstack | 20:56 | |
| *** andrewsben has quit IRC | 20:56 | |
| *** MarcMorata has quit IRC | 20:56 | |
| *** jakedahn has quit IRC | 20:56 | |
| *** jakedahn_ is now known as jakedahn | 20:56 | |
| *** byeager has left #openstack | 20:57 | |
| *** andrewsben has joined #openstack | 20:57 | |
| *** byeager has joined #openstack | 20:58 | |
| gnu111 | nhm: This is in my todo list. If I understand correctly it can do authentication, not sure about authz. You probably seen this: http://mirantis.blogspot.com/2011/08/ldap-identity-store-for-openstack.html | 21:00 |
| *** zul has quit IRC | 21:00 | |
| gnu111 | anyone seen the Error authenticating: Unauthorized error while login to dashboard? | 21:02 |
| nhm | gnu111: yep, though it looks like there you are expected to use keystone-manage. I want to allow existing users from a centrally managed ldap installation to have acess to openstack. | 21:02 |
| gnu111 | nhm: yes correct. before keystone there was a ldapdriver. but now keystone is the overall auth service and one of the option there is LDAP. | 21:02 |
| gnu111 | nhm: here;s the older LDAP stuff: http://git.cloud.com/cgit/nova/tree/nova/auth/slap.sh | 21:03 |
| *** tmichael has quit IRC | 21:03 | |
| nhm | gnu111: yeah. I haven't been able to find any new docs on keystone LDAP tying into an existing LDAP setup. | 21:04 |
| *** sdake has joined #openstack | 21:05 | |
| *** zul has joined #openstack | 21:05 | |
| gnu111 | nhm: here's the new ObjectClasses: https://github.com/openstack/keystone/blob/master/keystone/backends/ldap/keystone.ldif | 21:06 |
| *** whit has joined #openstack | 21:07 | |
| *** jakedahn has quit IRC | 21:07 | |
| *** andrewsben has quit IRC | 21:08 | |
| *** whit_ has joined #openstack | 21:08 | |
| *** lxu has joined #openstack | 21:08 | |
| *** dirkx_ has quit IRC | 21:08 | |
| *** jpwalters has joined #openstack | 21:08 | |
| *** heydrick has quit IRC | 21:09 | |
| *** stuntmachine has quit IRC | 21:11 | |
| Spirilis | ah yunno, I don't think that secgroup problem was a bug, the two instances I was running were in separate tenants | 21:11 |
| Spirilis | and my NOVA_PROJECT_ID was set to instance #3's tenant string | 21:11 |
| Spirilis | woops :) | 21:11 |
| *** whit has quit IRC | 21:12 | |
| jpwalters | I'm trying to get devstack installed on a fresh Ubuntu 11.10 VM, and the script fails with: mkdir: cannot create directory '/opt/stack/horizon/openstack-dashboard/quantum': permission denied | 21:12 |
| *** whit_ is now known as whit | 21:12 | |
| jpwalters | anyone know what's going on? I'm running the script as root, so I wouldn't think that there should be any permissions issues | 21:12 |
| *** Ryan_Lane has quit IRC | 21:13 | |
| lxu | is possible to have data replication with nova-volume? | 21:15 |
| Spirilis | lxu: since nova-volume is backed by Linux LVM, I would think you could do data replication on the disk devices underlying that LVM VG if you wanted to | 21:16 |
| jdg | jpwalters: Not sure, but aren't you supposed to run stack.sh as a regular user? I believe it sudos and prompts for password to add to sudoers file. | 21:17 |
| Spirilis | though I doubt you could guarantee that, say, the system magically fails over in the event that your primary host goes south | 21:17 |
| *** Ryan_Lane has joined #openstack | 21:17 | |
| jdg | gnu111: I run into this from time to time, for me restarting all of the glance, keystone and nova services resolves the issue. | 21:18 |
| gnu111 | jdg: how do you restart keystone? I didn't create a service for it. | 21:19 |
| jdg | How did you install? Are you using screens? | 21:19 |
| *** heckj has joined #openstack | 21:19 | |
| lxu | Spirilis: ty | 21:20 |
| *** lxu has quit IRC | 21:20 | |
| *** leifmadsen has joined #openstack | 21:20 | |
| jpwalters | jdg: I actually tried both ways, and got a different error as a non-root user, so I went ahead and tried as root | 21:21 |
| jdg | jpwalters: I've seen some finiky behavior with stack.sh but never like what you're describing. Does the directory structure up to '.../openstack-dashboar' exist? | 21:22 |
| jdg | gnu111: Depending on how you installed, you can just try 'service keystone restart' and see if it works. | 21:22 |
| *** Ryan_Lane has quit IRC | 21:23 | |
| jdg | jpwalters: you're not giving it 'sudo' when you run as a normal user are you? | 21:23 |
| jpwalters | jdg: yes, /opt/stack/horizon exists | 21:23 |
| jpwalters | and no, I wasn't giving it sudo when running from a normal user | 21:24 |
| jdg | jpwalters: that was for gnu111 ;) | 21:24 |
| jdg | jpwalters: Does openstack-dashboard exist as well? | 21:24 |
| jdg | jpwalters: I'm running it on a fresh system right now... see if I run into the same problem. | 21:25 |
| jpwalters | jdg: no openstack-dashboard doesn't exist...everything up to horizon does though | 21:26 |
| *** vlan-os has joined #openstack | 21:28 | |
| *** andrewsben has joined #openstack | 21:28 | |
| jpwalters | jdg: I don't know what the difference was, but I restarted stack.sh as a normal user, and this time everything built correctly | 21:29 |
| jdg | jpwalters: Sadly I have the same experience on a regular basis :( | 21:29 |
| jpwalters | jdg: heh, well, for now I guess I'm good then. Thanks for your help | 21:30 |
| *** tmichael has joined #openstack | 21:31 | |
| *** AhmedSoliman has quit IRC | 21:32 | |
| gnu111 | jdg: restarting the services didn't help: http://paste.openstack.org/show/3849/ | 21:32 |
| *** _adjohn has joined #openstack | 21:32 | |
| *** Trebortech has quit IRC | 21:34 | |
| *** adjohn has quit IRC | 21:34 | |
| *** _adjohn is now known as adjohn | 21:34 | |
| jdg | gnu111: Unfortunatley you're delving into areas that I'm not that familiar with. | 21:34 |
| *** lorin1 has quit IRC | 21:34 | |
| jdg | gnu111: How did you install your setup? | 21:34 |
| gnu111 | jdg: thanks for your help. I think it is something basic I am missing. I just did the install following the steps here: http://forums.openstack.org/viewtopic.php?f=23&t=482&p=1745&hilit=github+horizon+keystone#p1745 | 21:35 |
| *** dirkx_ has joined #openstack | 21:35 | |
| *** andreask has left #openstack | 21:36 | |
| jdg | Ahhh.... I tried that once :( I ended up resorting to one of the openstack-installer script | 21:36 |
| jdg | It's toooo easy to miss a step | 21:36 |
| jdg | one suggestion is check your db for user entries. | 21:37 |
| gnu111 | jdg: ok. | 21:37 |
| jdg | solidfire | 21:37 |
| jdg | oops... wrong window | 21:37 |
| *** Nac_ has quit IRC | 21:38 | |
| uvirtbot | New bug: #904971 in nova "Error handler in Scheduler uses ID, not UUID" [Medium,In progress] https://launchpad.net/bugs/904971 | 21:38 |
| jdg | Does anybody know how to get nova commands or euca-xxxx commands to work with a devstack install? | 21:40 |
| *** rnorwood has joined #openstack | 21:41 | |
| gnu111 | how do select a db in sqlite3? | 21:41 |
| Spirilis | jdg: if annegentle's info earlier is correct, the euca-* commands don't work since devstack uses Keystone, so only the 'nova' cmd works and 'nova-manage's user/project commands won't have the intended effect | 21:41 |
| Spirilis | anyway here's what I had to do to get the 'nova' cmd working: | 21:42 |
| Spirilis | set environment variables NOVA_USERNAME, NOVA_PASSWORD, NOVA_URL=<url to keystone, e.g. http://127.0.0.1:5000/v2.0/>, NOVA_VERSION=1.1, NOVA_PROJECT_ID=<tenant> | 21:42 |
| Spirilis | I think the tenants created include admin and demo | 21:43 |
| uvirtbot | New bug: #904976 in nova "Floating ip association fails for non-admin users" [Undecided,New] https://launchpad.net/bugs/904976 | 21:43 |
| Spirilis | be mindful of which one you are in when you create instances, since the nova cmd will only change polices/perform work on the instances in that tenant | 21:43 |
| jdg | Spirilis: Thanks a ton!! I tried setting the env var's and even tried just using --username xxx --password xxx but no luck. Maybe I missed something in the URL/Version etc | 21:44 |
| *** stuntmachine has joined #openstack | 21:44 | |
| *** whit has quit IRC | 21:45 | |
| *** mchenetz has joined #openstack | 21:45 | |
| Spirilis | yeah, the URL has to point to the Keystone as I mentioned | 21:45 |
| Spirilis | and it will bomb if you don't specify a NOVA_PROJECT_ID or --projectid which is just a tenant | 21:46 |
| *** v0id has joined #openstack | 21:46 | |
| *** andrewsben has quit IRC | 21:46 | |
| Spirilis | and you can ignore the apikey part, I think the apikey portion is used when you create custom tokens in keystone | 21:46 |
| *** mszilagyi has joined #openstack | 21:47 | |
| *** mszilagyi has quit IRC | 21:48 | |
| *** mszilagyi has joined #openstack | 21:48 | |
| jdg | Spirlis: Thanks, tried it but still getting the dang 'Invalid OpenStack Nova credentials' | 21:48 |
| *** andreask has joined #openstack | 21:49 | |
| otaku2 | can anyone recommend a good linux-based nas server distro? anything to consider other than openfiler vs freenas | 21:49 |
| Spirilis | jdg: hmm, my devstack install works ok with this: nova --username admin --password <adminpw> --projectid admin --url http://127.0.0.1:5000/v2.0/ --version 1.1 list | 21:50 |
| uvirtbot | New bug: #904532 in nova "Provide a script to automate cleanup of _base" [Medium,Confirmed] https://launchpad.net/bugs/904532 | 21:51 |
| jdg | AHHH!!! that worked. I must have fat fingered it the first time! | 21:53 |
| jdg | Thank you thank you thank you | 21:53 |
| Spirilis | np, remember it's NOVA_USERNAME, NOVA_PASSWORD, NOVA_URL, NOVA_VERSION, NOVA_PROJECT_ID that are the variables | 21:53 |
| Spirilis | I kept doing NOVA_PROJECT instead of NOVA_PROJECT_ID and that screwed me up | 21:53 |
| *** whit has joined #openstack | 21:54 | |
| *** miclorb_ has joined #openstack | 21:54 | |
| *** katkee has quit IRC | 21:56 | |
| *** jj0hns0n has joined #openstack | 21:59 | |
| *** bepernoot has joined #openstack | 21:59 | |
| *** andrewsben has joined #openstack | 22:01 | |
| *** jdg has quit IRC | 22:02 | |
| *** Ryan_Lane has joined #openstack | 22:04 | |
| *** dolphm has joined #openstack | 22:10 | |
| *** whit has quit IRC | 22:12 | |
| *** dolphm has quit IRC | 22:14 | |
| *** donaldngo_hp has joined #openstack | 22:14 | |
| *** foexle has quit IRC | 22:21 | |
| *** donaldngo_hp has quit IRC | 22:26 | |
| *** mgoldmann has quit IRC | 22:27 | |
| *** whit has joined #openstack | 22:28 | |
| *** bepernoot has quit IRC | 22:31 | |
| *** dubsquared has quit IRC | 22:34 | |
| jasona | anyone done an openstack presentation lately ? :) | 22:39 |
| *** bepernoot has joined #openstack | 22:40 | |
| jasona | e.g to explain to a technical audience 'what is openstack' etc. | 22:40 |
| *** redconnection has joined #openstack | 22:40 | |
| *** rsampaio has quit IRC | 22:41 | |
| *** bepernoot has quit IRC | 22:42 | |
| *** markvoelker has quit IRC | 22:43 | |
| *** lborda has quit IRC | 22:45 | |
| *** jj0hns0n has quit IRC | 22:46 | |
| *** dragondm has quit IRC | 22:47 | |
| *** glenc has joined #openstack | 22:50 | |
| *** ldlework has quit IRC | 22:51 | |
| *** jj0hns0n has joined #openstack | 22:51 | |
| *** glenc_ has quit IRC | 22:52 | |
| *** TheOsprey has quit IRC | 22:54 | |
| *** perestrelka has quit IRC | 23:00 | |
| *** tmichael has quit IRC | 23:04 | |
| *** perestrelka has joined #openstack | 23:04 | |
| *** ldlework has joined #openstack | 23:05 | |
| *** heckj has quit IRC | 23:06 | |
| *** jakedahn has joined #openstack | 23:07 | |
| uvirtbot | New bug: #905020 in horizon "Fix formatting of django tags in templates" [Undecided,New] https://launchpad.net/bugs/905020 | 23:08 |
| *** rnirmal has quit IRC | 23:08 | |
| *** vlan-os has quit IRC | 23:09 | |
| *** dirkx_ has quit IRC | 23:09 | |
| *** dirkx_ has joined #openstack | 23:09 | |
| *** PotHix has quit IRC | 23:11 | |
| *** troytoman is now known as troytoman-away | 23:13 | |
| *** imsplitbit has quit IRC | 23:14 | |
| *** jj0hns0n has quit IRC | 23:16 | |
| *** tmichael has joined #openstack | 23:19 | |
| *** krow has joined #openstack | 23:20 | |
| *** rustam has joined #openstack | 23:22 | |
| *** whit has quit IRC | 23:22 | |
| *** rnorwood has quit IRC | 23:22 | |
| *** jj0hns0n has joined #openstack | 23:23 | |
| *** hugokuo has joined #openstack | 23:24 | |
| *** adjohn has quit IRC | 23:25 | |
| *** adjohn has joined #openstack | 23:25 | |
| *** HowardRoark has quit IRC | 23:25 | |
| *** cfant has quit IRC | 23:25 | |
| *** po has quit IRC | 23:27 | |
| *** cfant has joined #openstack | 23:32 | |
| *** PiotrSikora has quit IRC | 23:34 | |
| *** Ramonster has joined #openstack | 23:34 | |
| *** PiotrSikora_ has joined #openstack | 23:35 | |
| *** whit has joined #openstack | 23:35 | |
| *** ppradhan has left #openstack | 23:36 | |
| *** RobertLaptop has joined #openstack | 23:37 | |
| *** RobertLaptop has left #openstack | 23:38 | |
| *** clauden has quit IRC | 23:40 | |
| *** clauden has joined #openstack | 23:41 | |
| *** cfant has quit IRC | 23:43 | |
| *** AlanClark has quit IRC | 23:45 | |
| *** mattray has quit IRC | 23:45 | |
| *** Ramonster has quit IRC | 23:49 | |
| *** bsza has quit IRC | 23:54 | |
| *** cfant has joined #openstack | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
