OpenStack Compute (nova)

VM on Compute-node Cannot Connect the Internet

Asked by xmdr

I install 2 nodes on CentOS 6.2 with openstack-essex.

When I launch vms, they can be connected on both physical server (with ping and ssh).However, the vms on the main node with all openstack service can connect to the Internet while vms on the other node with only compute and network running, cannot connect the Internet. Is there something wrong with my nova configuration?

When I stop the network service on the second node, vms on the node cannot be launched turning into error state.

I want to know where the problem is? The nova.conf file, the physical server network configuration or something else?

Here's my nova.conf file on both nodes. And the logs on both nodes have no obvious errors.

FlatDHCP mode with one NIC on each node.

[DEFAULT]
# LOGS/STATE
verbose=True
logdir=/var/log/nova
state_path=/var/lib/nova
lock_path=/var/lock/nova
# AUTHENTICATION
auth_strategy=keystone
# SCHEDULER
compute_scheduler_driver=nova.scheduler.filter_scheduler.FilterScheduler
# VOLUMES
volume_group=nova-volumes
volume_name_template=volume-%08x
iscsi_helper=tgtadm
# DATABASE
sql_connection=mysql://nova:nova@192.168.128.10/nova
# COMPUTE
libvirt_type=kvm
connection_type=libvirt
instance_name_template=instance-%08x
api_paste_config=/etc/nova/api-paste.ini
allow_resize_to_same_host=True
# APIS
osapi_compute_extension=nova.api.openstack.compute.contrib.standard_extensions
ec2_dmz_host=192.168.128.10
s3_host=192.168.128.10
# RABBITMQ
rabbit_host=192.168.128.10
# GLANCE
image_service=nova.image.glance.GlanceImageService
glance_api_servers=192.168.128.10:9292
# NETWORK
network_manager=nova.network.manager.FlatDHCPManager
force_dhcp_release=False
dhcpbridge_flagfile=/etc/nova/nova.conf
dhcpbridge=/usr/bin/nova-dhcpbridge
firewall_driver=nova.virt.libvirt.firewall.IptablesFirewallDriver
my_ip=192.168.128.10
public_interface=eth0
#vlan_interface=eth0
flat_network_bridge=br100
flat_interface=eth0
fixed_range=172.16.1.0/24
flat_network_dhcp_start=172.16.1.2
# NOVNC CONSOLE
novncproxy_base_url=http://192.168.128.10:6080/vnc_auto.html
vncserver_proxyclient_address=192.168.128.11
vncserver_listen=192.168.128.11

root_helper=sudo nova-rootwrap

Please help……

Question information

Language:
English Edit question
Status:
Expired
For:
OpenStack Compute (nova) Edit question
Assignee:
No assignee Edit question
Last query:
Last reply:
Revision history for this message
A.T. Cheng (cheng-aiti-taiwan) said : 		#1

Is my_ip setting correctly?
It seems strange to me that my_ip and vncserver_proxyclient_address are different in this conf file.

Revision history for this message
xmdr (xiamianderen) said : 		#2

Hi, Cheng

192.168.128.10 is the full-node, the other compute node is 192.168.128.11

I'm not very clear about the conf you said, could you tell me the configuration with my_ip and vncserver on both nodes? Or could you show me your conf param?

Thanks for your relpy.

Revision history for this message
A.T. Cheng (cheng-aiti-taiwan) said : 		#3

Hi back xmdr,

Sorry for the confusion. I mean when I look at your vncserver_proxyclient_address, I think it's the compute node's conf file. But when I look at your my_ip, it looks like the main node's conf file. That's why I say it's strange for me.
So, is it the main node's conf or the compute node's conf file?
Other than that I cannot see what's wrong on your conf file.

I have a similar multi-node setup and your conf file looks alright. Maybe you can provide more information?

By the way, with only 1 nic platforms, takeing off my_ip setup should also work.

Revision history for this message
xmdr (xiamianderen) said : 		#4

Hi Cheng,

On the full-node with 192.168.128.10, the configuration is:

my_ip=192.168.128.10
vncserver=192.168.128.10

on the compute node with 192.168.128.11

my_ip=192.168.128.10
vncserver=192.168.128.11

The configuration files on both nodes have only one difference on vncserver's ip.

When I trying to use 192.168.128.11 for my_ip on compute node, I cannot connect the vm launched on this node at all.

Thanks for your relpy~

Revision history for this message
Launchpad Janitor (janitor) said : 		#5

This question was expired because it remained in the 'Open' state without activity for the last 15 days.