Ask Your Question
0

RDO openstack: multinode setup Fedora 21

asked 2015-08-14 00:48:32 -0500

Sanjay_1987 gravatar image

updated 2015-08-14 02:37:20 -0500

HI, Trying to create a multinode setup RDO kilo + Fedora21. single node setup works fine (created using packstack --allinone ). Now adding a compute node to the same and edited the answer file as instructed by RDO quickstart page. Setup completed successfully but VM cannot be spawned.

Following is the error I get on dashboard

"Failed to perform requested operation on instance "router1", the instance has an error status: Please try again later [Error: Build of instance 4ad9b3a9-9496-44d4-83d1-d13066fbdccb aborted: Failed to allocate the network(s), not rescheduling.]. "

Following is the error in /var/log/neutron/serverl.log

2015-08-14 11:10:00.893 6510 WARNING keystonemiddleware.auth_token [-] Authorization failed for token 2015-08-14 11:10:00.893 6510 WARNING keystonemiddleware.auth_token [-] Identity response: {"error": {"message": "Could not find token: bafb5a3af04643d6995f85120caee157", "code": 404, "title": "Not Found"}} 2015-08-14 11:10:00.894 6510 WARNING keystonemiddleware.auth_token [-] Authorization failed for token

Any idea what is causing this ?

edit retag flag offensive close merge delete

1 answer

Sort by ยป oldest newest most voted
0

answered 2015-08-14 02:16:59 -0500

dbaxps gravatar image

updated 2015-08-14 08:22:26 -0500

Try to pick up answer-file from here :- http://bderzhavets.blogspot.com/2015/...

Run:-  packstack --answer-file=./answer3node.txt  on virgin HW

To change to 2 node setup

CONFIG_CONTROLLER_HOST=192.169.142.127
CONFIG_COMPUTE_HOSTS=192.169.142.137
CONFIG_NETWORK_HOSTS=192.169.142.127

Assume that both F21 have 2 Ethernet interfaces ( names should the same on both enp2s0,enp5s1)

CONFIG_NOVA_COMPUTE_PRIVIF=enp5s1
CONFIG_NOVA_NETWORK_PUBIF=enp2s0
CONFIG_NOVA_NETWORK_PRIVIF=enp5s1 
CONFIG_NEUTRON_OVS_TUNNEL_IF=enp5s1

Otherwise, see :- http://bderzhavets.blogspot.com/2015/...

edit flag offensive delete link more

Comments

Thanks buddy! But your setup is of 3 node and mine is of 2 node (controller+network is on same machine and compute is on another machine). The answer file on given link uses eth0 and eth1 which is same across all the machines in the given setup. However my machines uses differnet NIC card names.

Sanjay_1987 gravatar imageSanjay_1987 ( 2015-08-14 04:45:44 -0500 )edit
1

which eth0 is pointing to which machines eth0? CONFIG_NOVA_COMPUTE_PRIVIF=enp3s1 ----> compute machine i/f connected to controller CONFIG_NOVA_NETWORK_PRIVIF=enp13s0 ----> controller machine i/f connected to compute CONFIG_NOVA_NETWORK_PUBIF=--> I left this field empty Can you please guide ?

Sanjay_1987 gravatar imageSanjay_1987 ( 2015-08-14 04:49:02 -0500 )edit

That'exactly what I have entered. I am doubtful regarding interface name,under fields CONFIG_NOVA_COMPUTE_PRIVIF, CONFIG_NOVA_NETWORK_PRIVIF and CONFIG_NOVA_NETWORK_PUBIF. Can you help me out here!

Sanjay_1987 gravatar imageSanjay_1987 ( 2015-08-14 08:20:43 -0500 )edit

See updated answer.

dbaxps gravatar imagedbaxps ( 2015-08-14 08:23:46 -0500 )edit

Tried using the same but again no luck :(. Still the same error. I changed follwiing in nova.conf in both compute and controller: vif_plugging_is_fatal: false vif_plugging_timeout: 0

Now the VM are spawned successfuly. But VM cannot ping each other. Any idea what's happening ?

Sanjay_1987 gravatar imageSanjay_1987 ( 2015-08-14 11:06:34 -0500 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2015-08-14 00:48:32 -0500

Seen: 445 times

Last updated: Aug 14 '15