Troubleshooting Host to VM networking

asked 2017-05-01 19:51:30 -0500

tiger gravatar image

I have seven Ubuntu 14.04LTS VM's running with IP addresses 10.10.1.10-10.10.1.16 They can ping each other from each other I have a network with subnet 192.168.140.0/24. I have a router that encompasses both networks. From the host, I run the "ip netns" command to learn the namespace of the qrouter.

[root@kvm8 ~]# ip netns
qdhcp-0284a8c4-8a46-455e-875e-72dae2cd0124
qrouter-00e726f8-31c5-491b-9eb3-958f3cdbde13

Then I try to ping the cluster member:

[root@kvm8 ~]# ip netns exec qrouter-00e726f8-31c5-491b-9eb3-958f3cdbde13 ping 10.10.1.10
PING 10.10.1.10 (10.10.1.10) 56(84) bytes of data.
^C
--- 10.10.1.10 ping statistics ---
3 packets transmitted, 0 received, 100% packet loss, time 1999ms

and.....nothing. How is it I can follow someone's recipe (purportedly a simple Openstack networking tutorial) and get different results? So if anyone can give me a hint where to look, I'd appreciate it. I am not trying to split the atom or perform brain surgery. Just want to have a cluster of seven NFV's that are accessible from the outside world.

edit retag flag offensive close merge delete

Comments

Troubleshooting depends very much on the core plugin you use. If it's ML2, it then depends on the type and mechanism drivers you configured. Another huge factor is the router type: Distributed, centralized, HA etc.

Bernd Bausch gravatar imageBernd Bausch ( 2017-05-01 22:36:16 -0500 )edit
Bernd Bausch gravatar imageBernd Bausch ( 2017-05-01 22:39:11 -0500 )edit

I also recommend Assaf Muller's blog: Distributed router and Neutron Troubleshooting.

There is much more out there.

Bernd Bausch gravatar imageBernd Bausch ( 2017-05-01 22:42:47 -0500 )edit