Ask Your Question
0

dhcp reply lost between int-br1 and br-int

asked 2013-07-09 07:42:04 -0600

paolo-veronesi gravatar image

Hi, first of all, here my configuration: O.S. CentOS 6.4 x86_64 (SELinux disabled) Grizzly from RDO repos; Quantum with ovs plugin and vlan (3 vlan id configured at switch level, from 3501 to 3503); cloudctrl01 acts as keystone, glance, quantum-, rabbitmq, mysql, nova- (except network and compute); three interfaces: eth0 - management; eth1 - data; eth2 external; nova01 acts as nova-compute; two interfaces: eth0 - management, eth1 - data.

As far i understand, the path in the compute node should be: vm:demo01:eth0 <---> tapafa41705-77 <---> qbrafa41705-77 <---> qvbafa41705-77 <---> qvoafa41705-77 <---> br-int <---> int-br1 <---> phy-br1 <---> br1 <---> eth1

I suppose there something wrong on the compute node (nova01): with tcpdump i saw that dhcp reply arrive at int-br1 but not at br-int.

Additional info:

uname -r

2.6.32-358.111.1.openstack.el6.x86_64

rpm -qa |grep openvs

openvswitch-1.10.0-1.el6.x86_64 openstack-quantum-openvswitch-2013.1.2-1.el6.noarch kmod-openvswitch-1.10.0-1.el6.x86_64

ip li

1: lo: <loopback,up,lower_up> mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: eth0: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether e0:cb:4e:55:37:c9 brd ff:ff:ff:ff:ff:ff 3: eth1: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether e0:cb:4e:55:36:79 brd ff:ff:ff:ff:ff:ff 4: ovs-system: <broadcast,multicast> mtu 1500 qdisc noop state DOWN link/ether a2:ae:a7:4e:81:6c brd ff:ff:ff:ff:ff:ff 43: br-int: <broadcast,up,lower_up> mtu 1500 qdisc noqueue state UNKNOWN link/ether 0a:4c:e0:02:1d:4c brd ff:ff:ff:ff:ff:ff 45: br1: <broadcast,up,lower_up> mtu 1500 qdisc noqueue state UNKNOWN link/ether e0:cb:4e:55:36:79 brd ff:ff:ff:ff:ff:ff 46: phy-br1: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether e6:98:6d:71:2b:24 brd ff:ff:ff:ff:ff:ff 47: int-br1: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 56:e1:03:b6:07:e5 brd ff:ff:ff:ff:ff:ff 60: qbrafa41705-77: <broadcast,multicast,up,lower_up> mtu 1500 qdisc noqueue state UNKNOWN link/ether 4e:47:72:51:fe:7e brd ff:ff:ff:ff:ff:ff 61: qvoafa41705-77: <broadcast,multicast,promisc,up,lower_up> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether a6:c9:e2:a5:5e:52 brd ff:ff:ff:ff:ff:ff 62: qvbafa41705-77: <broadcast,multicast,promisc,up,lower_up> mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 4e:47:72:51:fe:7e brd ff:ff:ff:ff:ff:ff 63: tapafa41705-77: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 500 link/ether fe:16:3e:0f:0b:f3 brd ff:ff:ff:ff:ff:ff

ovs-vsctl show

cef7bfff-248f-438c-82f9-499505f51197 Bridge br-int Port "int-br1" Interface "int-br1" Port "qvoafa41705-77" tag: 2 Interface "qvoafa41705-77" Port br-int ... (more)

edit retag flag offensive close merge delete
add a comment

3 answers

Sort by ยป oldest newest most voted
0

answered 2013-07-09 07:46:28 -0600

paolo-veronesi gravatar image

I forgot to put the pluing.ini:

cat /etc/quantum/plugin.ini |grep -v '#'

[DATABASE] sql_connection = hidden sql_max_retries = 10 reconnect_interval = 2

[OVS] tenant_network_type=vlan network_vlan_ranges=physnet1:3501:3503 integration_bridge=br-int bridge_mappings=physnet1:br1 enable_tunneling=False

[AGENT] polling_interval = 2

[SECURITYGROUP] firewall_driver = quantum.agent.linux.iptables_firewall.OVSHybridIptablesFirewallDriver

edit flag offensive delete link more
add a comment
0

answered 2013-07-09 09:04:43 -0600

paolo-veronesi gravatar image

I manually assigned the ip to the vm (its mac address is fa:16:3e:0f:0b:f3): ifconfig eth0 10.50.1.3 netmask 255.255.255.0 up

And i tried to ping the dnsmasq ip ping -c 1 10.50.1.3 (dnsmasq ip)

I have the same behaviour: replies get lost between int-br1 -> br-int on the compute node:

tcpdump -nne -i br-int ether host fa:16:3e:0f:0b:f3 10:55:31.618919 fa:16:3e:0f:0b:f3 > ff:ff:ff:ff:ff:ff, ethertype 802.1Q (0x8100), length 46: vlan 1, p 0, ethertype ARP, Request who-has 10.50.1.3 tell 10.50.1.2, length 28 10:55:32.618834 fa:16:3e:0f:0b:f3 > ff:ff:ff:ff:ff:ff, ethertype 802.1Q (0x8100), length 46: vlan 1, p 0, ethertype ARP, Request who-has 10.50.1.3 tell 10.50.1.2, length 28 10:55:33.618751 fa:16:3e:0f:0b:f3 > ff:ff:ff:ff:ff:ff, ethertype 802.1Q (0x8100), length 46: vlan 1, p 0, ethertype ARP, Request who-has 10.50.1.3 tell 10.50.1.2, length 28

tcpdump -nne -i int-br1 ether host fa:16:3e:0f:0b:f3 10:55:31.618922 fa:16:3e:0f:0b:f3 > ff:ff:ff:ff:ff:ff, ethertype 802.1Q (0x8100), length 46: vlan 1, p 0, ethertype ARP, Request who-has 10.50.1.3 tell 10.50.1.2, length 28 10:55:31.619411 fa:16:3e:64:b6:fb > fa:16:3e:0f:0b:f3, ethertype ARP (0x0806), length 60: Reply 10.50.1.3 is-at fa:16:3e:64:b6:fb, length 46 10:55:32.618868 fa:16:3e:0f:0b:f3 > ff:ff:ff:ff:ff:ff, ethertype 802.1Q (0x8100), length 46: vlan 1, p 0, ethertype ARP, Request who-has 10.50.1.3 tell 10.50.1.2, length 28 10:55:32.619128 fa:16:3e:64:b6:fb > fa:16:3e:0f:0b:f3, ethertype ARP (0x0806), length 60: Reply 10.50.1.3 is-at fa:16:3e:64:b6:fb, length 46 10:55:33.618780 fa:16:3e:0f:0b:f3 > ff:ff:ff:ff:ff:ff, ethertype 802.1Q (0x8100), length 46: vlan 1, p 0, ethertype ARP, Request who-has 10.50.1.3 tell 10.50.1.2, length 28 10:55:33.619054 fa:16:3e:64:b6:fb > fa:16:3e:0f:0b:f3, ethertype ARP (0x0806), length 60: Reply 10.50.1.3 is-at fa:16:3e:64:b6:fb, length 46

edit flag offensive delete link more
add a comment
0

answered 2013-07-15 13:29:42 -0600

paolo-veronesi gravatar image

Hi, problem solved. There was an issue in the configuration of two switches between cloudctrl01 and nova-compute.

We have two DELL switches, but different model and o.s. Their configuration was the same: switchport mode general switchport general allowed vlan add 3501-3510

This configuration works with "PowerConnect 7048R-RA, 4.2.1.3, VxWorks 6.6" (!System Software Version 4.2.1.3), but not on "PowerConnect 6224, 3.3.5.5, VxWorks 6.5" (!System Software Version 3.3.5.5). The right configuration for the PowerConnect 6224 is the following: switchport mode trunk switchport trunk allowed vlan 3501-3510

kind regards

edit flag offensive delete link more
add a comment

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

subscribe to rss feed

Stats

Asked: 2013-07-09 07:42:04 -0600

Seen: 71 times

Last updated: Jul 15 '13

Related questions

Neutron Port Create

Floating IP and 'system' ports

CLI Tool and Rest API for Quantum

the nova endpoint URL is throwing error

Dhcp and metadata not working

How to enable SNAT after disable it

kolla aio external network with single NIC

can't ping instance from any ip

vRouter does not forward multicast packet

Backup of instances?