[Openstack-operators] 500 Internal Server error and [Errno 113] EHOSTUNREACH when adding a new node

Andrew Glen-Young andrew.glen-young at canonical.com
Tue Dec 4 15:24:08 UTC 2012


Hello,

On Tue 04-12-2012 13:01 ›, Gui Maluf wrote:
> Hy, I've a ubuntu 12.04 cloud controller + node runing essex in multi_node
> mode.
> So I'm trying to install a new node but without success.
> 
> Node and controller have the same /etc/network/interface/;
> Node is running nova-{api-metadata,compute,network,volume};
> nova.conf: http://paste.openstack.org/show/27390/
> Node have this line above  uncommented and --my_ip changed
> --enabled_apis=ec2,osapi_compute,osapi_volume,metadata
> 
> Here is the nova-compute.log: http://paste.openstack.org/show/27387/ with
> complete error.
> 
> I'm trying to install two nodes, and from both I can reach all services on
> cloud-controller.
> telnet cloud-controller 9292 works from both nodes. They all have the same
> nova.conf. I dont know what else can be this error.
> I've check many things and I cant find a solution to this.
> Thanks in advance!
> 

A quick look at your log seems to indicate that your Glance API server
is returning the '500' error (line 71).

It looks like your Glance server is failing to contact your Swift proxy
server (line 149).

I would suggest checking your Glance server's logs and making sure that
the service is configured correctly and testing the service with the
glance client (package: glance-client).

HTH

-- 
Regards.
Andrew Glen-Young



More information about the OpenStack-operators mailing list