[neutron] [nova] [ussuri] IP of the VM at the provider network is different than the one registered at the Openstack controller
Hello, Setup: i follow the main instruction for deploying openstack for learning purposes. I can successfully deploy vms in an internal network, and also connect them to a router that gives internet access using floating ips etc. My setup is based on ubuntu 18. I have virtualbox for deploying controller (10.0.0.11) and compute (10.0.0.31) etc. I have host only adapters at virtual box (dhcp enabled) for the two networks: mgmt: 10.0.0.0/24 and provider 203.0.111.0/24 as per the default instructions, and i use iptables to NAT the vms. Problem: When following the instructions from https://docs.openstack.org/install-guide/launch-instance-provider.html the vm is deployed successfully but the ip that it is assigned is different from the one that the controller has. It seems that there are more than one dhcp instances running? How can i debug this? Any advice? all the best, Pavlos.
Hi, On Wed, Nov 25, 2020 at 02:52:56PM +0200, Pavlos Basaras wrote:
Hello,
Setup: i follow the main instruction for deploying openstack for learning purposes. I can successfully deploy vms in an internal network, and also connect them to a router that gives internet access using floating ips etc. My setup is based on ubuntu 18. I have virtualbox for deploying controller (10.0.0.11) and compute (10.0.0.31) etc. I have host only adapters at virtual box (dhcp enabled) for the two networks: mgmt: 10.0.0.0/24 and provider 203.0.111.0/24 as per the default instructions, and i use iptables to NAT the vms.
Problem: When following the instructions from https://docs.openstack.org/install-guide/launch-instance-provider.html the vm is deployed successfully but the ip that it is assigned is different from the one that the controller has. It seems that there are more than one dhcp instances running? How can i debug this?
I'm not sure if I understand correctly but do You want to have Your virtualbox to run as DHCP server for vms run by OpenStack? Neutron has own dhcp agent and that is providing IPs to Your VMs according to the IP Allocation done by neutron. IPs are form the subnet created in the network to which Your vm is plugged.
Any advice?
all the best, Pavlos.
-- Slawek Kaplonski Principal Software Engineer Red Hat
Hello, many thanks for your prompt response. i dissbaled the dhcp from virtualbox and now the vms get the ip from the neutron dhcp as registered by the controller. Works fine! all the best, Pavlos. On Wed, Nov 25, 2020 at 4:58 PM Slawek Kaplonski <skaplons@redhat.com> wrote:
Hi,
Hello,
Setup: i follow the main instruction for deploying openstack for learning
I can successfully deploy vms in an internal network, and also connect
to a router that gives internet access using floating ips etc. My setup is based on ubuntu 18. I have virtualbox for deploying controller (10.0.0.11) and compute (10.0.0.31) etc. I have host only adapters at virtual box (dhcp enabled) for the two networks: mgmt: 10.0.0.0/24 and provider 203.0.111.0/24 as per the default instructions, and i use iptables to NAT the vms.
Problem: When following the instructions from https://docs.openstack.org/install-guide/launch-instance-provider.html
On Wed, Nov 25, 2020 at 02:52:56PM +0200, Pavlos Basaras wrote: purposes. them the
vm is deployed successfully but the ip that it is assigned is different from the one that the controller has. It seems that there are more than one dhcp instances running? How can i debug this?
I'm not sure if I understand correctly but do You want to have Your virtualbox to run as DHCP server for vms run by OpenStack? Neutron has own dhcp agent and that is providing IPs to Your VMs according to the IP Allocation done by neutron. IPs are form the subnet created in the network to which Your vm is plugged.
Any advice?
all the best, Pavlos.
-- Slawek Kaplonski Principal Software Engineer Red Hat
participants (2)
-
Pavlos Basaras
-
Slawek Kaplonski