Ask Your Question

DHCP not working: Quantum + RHEL6.3 + Linuxbridge

asked 2013-03-29 22:00:45 -0500

a-s-o-anas gravatar image

Hello, I'm deploying OpenStack Folsom on CentOS6.3, and I'm using Quantum with LinuxBridge as a L2 plugin. For the installation I followed these two posts: ( (

I did all what is needed, and I finished the install without any error in the log files (/var/log/quantum/.... ), but I'm seeing a strange behaviour. 1 - The created bridge brqcxxxxxxx does not have an IP address. Is it normal ?! I think it should have one, because it'll be the gateway for the VMs attached to it. 2 - An interface named ns-xxxxxxx is created an got the first available address in the created subnet. What is for?! 3 - dnsmasq log file shows this message : <dhcp, static="" leases="" only="" on="",="" lease...="" &gt;="" .="" ""="" is="" the="" subnet="" i="" created.="" shouldn't="" be="" listening="" on="" a="" host="" ip="" address="" ("" for="" example="" ),="" which="" should="" be="" the="" same="" address="" given="" to="" the="" bridge="" ?!!="" 4="" -="" and="" of="" course="" when="" i="" create="" a="" vm,="" it="" sends="" its="" dhcpdiscover="" messages,="" no="" one="" of="" them="" is="" received="" by="" dnsmasq,="" and="" the="" vm="" can't="" get="" an="" ip="" address.<="" p="">

here are my configs :

//quantum.conf [DEFAULT] verbose = True bind_host = bind_port = 9696 core_plugin = quantum.plugins.linuxbridge.lb_quantum_plugin.LinuxBridgePluginV2 api_paste_config = api-paste.ini auth_strategy = keystone control_exchange = quantum notification_driver = quantum.openstack.common.notifier.list_notifier list_notifier_drivers = quantum.openstack.common.notifier.rabbit_notifier rpc_backend = quantum.openstack.common.rpc.impl_qpid qpid_hostname =

// dhcp_agent.ini [DEFAULT] state_path = /var/lib/quantum interface_driver = quantum.agent.linux.interface.BridgeInterfaceDriver dhcp_driver = quantum.agent.linux.dhcp.Dnsmasq use_namespaces = False root_helper = sudo quantum-rootwrap /etc/quantum/rootwrap.conf auth_url = admin_username = admin admin_password = verybadpassword admin_tenant_name = admin

//l3_agent.ini [DEFAULT] interface_driver = quantum.agent.linux.interface.BridgeInterfaceDriver auth_url = auth_region = RegionOne admin_tenant_name = admin admin_user = admin admin_password = verybadpassword root_helper = sudo quantum-rootwrap /etc/quantum/rootwrap.conf use_namespaces = False external_network_bridge =

I omitted all the commented lines, and I left the default config for all the other files.

thanks in advance

edit retag flag offensive close merge delete

6 answers

Sort by ยป oldest newest most voted

answered 2013-04-01 10:31:55 -0500

a-s-o-anas gravatar image

Thanks yong sheng gong, that solved my question.

edit flag offensive delete link more

answered 2013-04-01 02:38:43 -0500

gongysh gravatar image

Since you are using redhat 6.3, which does not support namespace, you had better separate DHCP agent and L3 agent. Typically, these two agents cannot coexist well.

edit flag offensive delete link more

answered 2013-04-01 10:33:09 -0500

a-s-o-anas gravatar image

I want to add that iptables was blocking the communication with the dhcp daemon, so I had to open the needed port (67 udp). thank you for your help.

edit flag offensive delete link more

answered 2013-04-01 10:03:17 -0500

a-s-o-anas gravatar image

should I run them both on separate machine, or disabling L3 agent will be sufficient? I don't know if running quantum-server, quantum-linuxbridge-agent and quantum-dhcp-agent is enough for quantum to give a normal behaviour ?! I tried disabling the L3 agent but I didn't work. As I said before this is only a proof of concept scenario. thanks,

edit flag offensive delete link more

answered 2013-03-30 09:49:02 -0500

a-s-o-anas gravatar image

thanks yong for your answer, and here is the needed informations: 1 - "quantum net-show net1" output: (net1 is the name of network I've created for tests )

+------------------------------+------------------------------------------------+ | Field | Value | +------------------------------+------------------------------------------------+ | admin_state_up | True | | id | e7de2673-cd46-4510-8df9-298bba8d1d36 | | name | net1 | | provider:network_type | vlan | | provider:physical_network | physnet1 | | provider:segmentation_id | 1000 | | router:external | False | | shared | True | | status | ACTIVE | | subnets | 084c9068-d28f-415c-8ac1-f65db72d6086 | | tenant_id | e84fdf0dce7e4068bc7fb5730c4413ef | +------------------------------+-------------------------------------------------+

2 - linuxbridge_config.ini file :

[VLANS] tenant_network_type = vlan network_vlan_ranges = physnet1:1000:1100

[DATABASE] sql_connection = mysql://quantum:quantum@Controller.OpenStack/quantum_linux_bridge reconnect_interval = 2

[LINUX_BRIDGE] physical_interface = eth0 physical_interface_mappings = physnet1:eth0

[AGENT] polling_interval = 2

root_helper = sudo quantum-rootwrap /etc/quantum/rootwrap.conf

I am using one quantum "router" as an internal gateway, which has an interface connected to subnet. No external gateway is created.

I forgot to say that I'm using a single node (all the OpenStack daemons are running on the same host), this a proof of concept scenario.

If you need any other information do not hesitate to ask for it. I really need to get our of this problem.

edit flag offensive delete link more

answered 2013-03-30 00:08:45 -0500

gongysh gravatar image

run quantum net-show and give us your linuxbridge_conf.ini, we will see what is your network's provider type.

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower


Asked: 2013-03-29 22:00:45 -0500

Seen: 96 times

Last updated: Apr 01 '13