Open Stack

Hi All,

I am running a cluster of 4 nodes:  3 all-in-one control/compute nodes and
1 compute-only node.   I am running the quantum-dhcp-agent on the three
all-in-one nodes in active/active/active fashion.   I noticed that any
instance launched after stopping quanutm-dhcp-agent on control-node2 failes
to obtain an address.   It was my understanding that if one of the
dhcp-agents went down the others would take over for it.   Am I missing
something?   I have added all subnets that use dhcp to all of the quantum
dhcp agents.  Help would greatly appreciated.   I need to stop
control-node2 for maintenance but doing so brings down all the VMs.

PS I am using the latest grizzly release from Ubuntu cloud-archive.

Thanks,
Sam
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20131121/b558cd3f/attachment.html>