Open Stack

I've been following the install/setup directions here:

http://docs.openstack.org/juno/install-guide/install/yum/content/ch_preface.html

and have completed steps 1-9.I have 3 nodes: controller[runs all the
controller components, mariadb, rabbitmq], compute[runs compute, cinder,
and swift services], and network[just neutron, and gateway to outside
network].

I can launch an instance, get a DHCP IP address, login to the instance via
console, and ping the outside world. So, it seems almost everything is
working (had a lot of difficulty with neutron networking, but eventually
figured it out and now works).

However, when the instance tries to contact the metadata service API at
169.254.169.254, I get no response. Within the instance, the ARP cache
shows incomplete, so it's almost like nothing is servicing the metadata API.

Here's some stuff I did on the network node based on info I got from:

https://ask.openstack.org/en/question/66516/instance-can-not-access-openstack-metadata-service/

root at network neutron]# ip netns list
qrouter-61e6eb07-a052-4248-925b-268f4f71765c
qdhcp-93fc034f-d010-4e89-aa01-ac8e563b5927

[root at network neutron]# ip netns exec
qrouter-61e6eb07-a052-4248-925b-268f4f71765c iptables-save | grep -i
redirect
-A neutron-l3-agent-PREROUTING -d 169.254.169.254/32 -p tcp -m tcp --dport
80 -j REDIRECT --to-ports 9697

[root at network neutron]# ip netns exec
qrouter-61e6eb07-a052-4248-925b-268f4f71765c netstat -antp
Active Internet connections (servers and established)
Proto Recv-Q Send-Q Local Address           Foreign Address
State       PID/Program name
tcp        0      0 0.0.0.0:9697            0.0.0.0:*
LISTEN      1335/python

[root at network neutron]# ps -f --pid 1335 | fold -s -w 82
UID        PID  PPID  C STIME TTY          TIME CMD
root      1335     1  0 14:56 ?        00:00:00 /usr/bin/python
/bin/neutron-ns-metadata-proxy
--pid_file=/var/lib/neutron/external/pids/61e6eb07-a052-4248-925b-268f4f71765c.pid
 --metadata_proxy_socket=/var/lib/neutron/metadata_proxy
--router_id=61e6eb07-a052-4248-925b-268f4f71765c
--state_path=/var/lib/neutron
--metadata_port=9697 --verbose
--log-file=neutron-ns-metadata-proxy-61e6eb07-a052-4248-925b-268f4f71765c.log

--log-dir=/var/log/neutron

[root at network neutron]# ip netns exec
qrouter-61e6eb07-a052-4248-925b-268f4f71765c curl http://169.254.169.254

At this point, I get no response from that curl command. It just hangs
until I break out of it.

However, similar to the comment in the post above, I get a response in the
qdhcp namespace:

[root at network neutron]# ip netns exec
qdhcp-93fc034f-d010-4e89-aa01-ac8e563b5927 curl http://169.254.169.254
1.0
2007-01-19
2007-03-01
2007-08-29
2007-10-10
2007-12-15
2008-02-01
2008-09-01
2009-04-04

How do I further troubleshoot this? What does this mean? Do I have
something severely misconfigured?

Also, going over the instructions, this is my understanding:

network node runs metadata_agent service, but points to
nova_metadata_ip=controller. on the controller, I have
service_metadata_proxy=true under [neutron] sub-section of nova.conf. I'm
guessing this means that network/metadata_agent talks to the metadata_proxy
on controller, but if this is a proxy, where is the actual metadata service
running? And is there an additional step to configure that metadata
service? I did not see anything like that in the install/setup
instructions. Is my understanding even correct?

Can anyone please help?

Thanks,
TOG
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20150923/2970db3d/attachment.html>