[Openstack] Cloud-init not pulling metadata when booting instances

Brian Haley brian.haley at hp.com
Fri Jun 13 16:47:47 UTC 2014


On 06/13/2014 04:13 AM, Ricardo Carrillo Cruz wrote:
> Hi there
> 
> I installed a one node Icehouse cloud at home, with Neutron and GRE.
> Everything is ok, except I can't ssh to ubuntu instances.
> The problem is that cloud-init is unable to reach the metadata service, thus the
> keypair is never injected into the instance. 
> This is the error:
> 
> <snip>
> 
> cloud-init start-local running: Fri, 13 Jun 2014 07:51:10 +0000. up 3.33 seconds
> no instance data found in start-local
> ci-info: lo    : 1 127.0.0.1       255.0.0.0       .
> ci-info: eth0  : 1 192.168.0.21    255.255.255.0   fa:16:3e:41:82:7b
> ci-info: route-0: 0.0.0.0         192.168.0.1     0.0.0.0         eth0   UG
> ci-info: route-1: 192.168.0.0     0.0.0.0         255.255.255.0   eth0   U
> cloud-init start running: Fri, 13 Jun 2014 07:51:10 +0000. up 3.85 seconds
> 2014-06-13 07:51:13,829 - util.py[WARNING]: 'http://169.254.169.254/2009-04-04/meta-data/instance-id' failed [2/120s]: url error [[Errno 113] No route to host]
> 2014-06-13 07:51:16,829 - util.py[WARNING]: 'http://169.254.169.254/2009-04-04/meta-data/instance-id' failed [5/120s]: url error [[Errno 113] No route to host]
> 2014-06-13 07:51:19,829 - util.py[WARNING]: 'http://169.254.169.254/2009-04-04/meta-data/instance-id' failed [8/120s]: url error [[Errno 113] No route to host]
> 2014-06-13 07:51:22,829 - util.py[WARNING]: 'http://169.254.169.254/2009-04-04/meta-data/instance-id' failed [11/120s]: url error [[Errno 113] No route to host]
> 2014-06-13 07:51:25,829 - util.py[WARNING]: 'http://169.254.169.254/2009-04-04/meta-data/instance-id' failed [14/120s]: url error [[Errno 113] No route to host]

The metadata proxy should be started by the l3-agent automatically as it
defaults to True, have you verified it's running?  There will be a
metadata-proxy log file with your router id in it in /var/log/neutron.  Also,
make sure the metadata-agent is actually up and running by looking in the log file.

The other thing that could be happening is that your instance is trying to send
to 169.254.169.254 directly and not via the default route, I've seen Windows do
this before.  I'm blanking on the fix for that.

I'd run tcpdump on the tap and/or in the router namespace to see where the
packets are getting dropped, that should help you narrow it down at least.

> # IP address used by Nova metadata server
> nova_metadata_ip = controller

'controller' is resolvable via DNS, correct?

-Brian





More information about the Openstack mailing list