Open Stack

Actually, I found that the tcpdump -i 'tapf9884dcb-e5' got dhcp reply. However, tcpdump -i br-int only has dhcp request.

Also, when I tried to get ip address on the dhcp agent, I was not able to see 'tapf9884dcb-e5'

ip netns exec qdhcp-7e73481a-d100-46d4-8dd1-9b3bb7949333
No cmd specified
[root at blade04 quantum]# ip netns exec qdhcp-7e73481a-d100-46d4-8dd1-9b3bb7949333 ip address
15: 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
    inet 127.0.0.1/8 scope host lo
    inet6 ::1/128 scope host
       valid_lft forever preferred_lft forever
18: ns-f9884dcb-e5: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether fa:16:3e:a1:d5:e2 brd ff:ff:ff:ff:ff:ff
    inet 30.0.0.3/24 brd 30.0.0.255 scope global ns-f9884dcb-e5
    inet6 fe80::f816:3eff:fea1:d5e2/64 scope link
       valid_lft forever preferred_lft forever

Would this be an issue?

Thanks,

YuLing
From: C, Yuling
Sent: Tuesday, December 17, 2013 9:37 PM
To: 'openstack at lists.openstack.org'
Subject: RE: vm instances could not get ip address from dhcp agent

B.T.W, I just noticed that 'brctl show' command does not show any attached interface on this test bed. Could this be the root cause?
# brctl show
bridge name     bridge id               STP enabled     interfaces
virbr0          8000.525400046ff3       yes             virbr0-nic

Also, I saw error message as '2013-12-17 21:33:47.586 32628 ERROR nova.virt.libvirt.driver [-] Getting disk size of foreman: [Errno 2] No such file or directory: '/var/lib/libvirt/images/foreman.img''.

Any clue?

Thanks,

YuLing
From: C, Yuling
Sent: Tuesday, December 17, 2013 6:47 PM
To: openstack at lists.openstack.org<mailto:openstack at lists.openstack.org>
Subject: vm instances could not get ip address from dhcp agent

Hi All,

I have an openstack installation environment of two nodes. One node contains Controller, Network, Nova, Glance, Keystone, Client, and Cinder. The other node is just a nova_compute node.   Everything worked very well before. Suddenly we hit the issue that vm instances could not get ip addresses from dhcp agent.  I looked into dhcp-agent.log, but didn't find much useful information. Just wondering if someone could shed some light on this issue.

The behavior is that even the VMs created on node1, which contains Controller, Network, Nova, Glance, keystone, Client and Cinder could not get the IP from dhcp agent. When I use tcpdump -i br-ehter1, or tcpdump -i < physical Ethernet port id>, I was able to see the DHCP requests, but no reply from dhcp agent.

Any idea?

Thanks,

YuLing



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20131218/bea7d953/attachment.html>