Ask Your Question
0

dhcp reply lost between int-br1 and br-int

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

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

3 answers

Sort by ยป oldest newest most voted
0

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

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
0

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

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
0

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

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

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

Stats

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

Seen: 66 times

Last updated: Jul 15 '13