Open Stack

Thanks Aaron and John for your fast answers.

Unfortunatelly I forgot to include the subject on this message.

To be honest I'm not totally sure what is going on, but what I notice is
that when I do start a VM on top of OpenStack, after the bridge br100 is
created, I cannot login the rest of the machines on the lab (those which
are not related to the OpenStack tests, but on the same physical network)

*Here is a cat of my /var/log/messages on the controller node*

[root@*controller* ~]# *tail -20 /var/log*/messages

Sep 25 16:51:40 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
e0:b9:ba:ae:78:dd no address
available

Sep 25 16:51:41 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
4c:bc:a5:92:e1:c7 no address
available

Sep 25 16:51:42 controller dnsmasq-dhcp[5860]: DHCPINFORM(br100)
192.168.30.52
00:1b:b1:28:64:ae

Sep 25 16:51:42 controller dnsmasq-dhcp[5860]: DHCPACK(br100) 192.168.30.52
00:1b:b1:28:64:ae
Leliane-NB

Sep 25 16:51:42 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
30:39:26:83:7c:6a no address
available

Sep 25 16:51:42 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
84:8f:69:c7:45:26 no address
available

Sep 25 16:51:44 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
4c:b1:99:83:77:82 no address available
Sep 25 16:51:45 controller dnsmasq-dhcp[5860]: BOOTP(br100)
14:5a:05:1e:43:1a no address available
Sep 25 16:51:48 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
18:3f:47:ba:0b:a8 no address available
Sep 25 16:51:48 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
e0:b9:ba:ae:78:dd no address available
Sep 25 16:51:48 controller dnsmasq-dhcp[5860]: DHCPREQUEST(br100) 10.0.0.2
fa:16:3e:64:cd:82
Sep 25 16:51:48 controller dnsmasq-dhcp[5860]: DHCPACK(br100) 10.0.0.2
fa:16:3e:64:cd:82 ubuntu01
Sep 25 16:51:50 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
30:39:26:83:7c:6a no address available
Sep 25 16:51:51 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
84:8f:69:c7:45:26 no address available
Sep 25 16:51:52 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
7c:c3:a1:ac:a1:ed no address available
Sep 25 16:51:53 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
4c:b1:99:83:77:82 no address available
Sep 25 16:51:53 controller dnsmasq-dhcp[5860]: DHCPINFORM(br100)
192.168.60.2 00:25:11:cf:f0:54
Sep 25 16:51:53 controller dnsmasq-dhcp[5860]: DHCPACK(br100) 192.168.60.2
00:25:11:cf:f0:54 sed06
Sep 25 16:51:53 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
7c:c3:a1:ac:a1:ed no address available
Sep 25 16:51:55 controller dnsmasq-dhcp[5860]: DHCPDISCOVER(br100)
7c:c3:a1:ac:a1:ed no address available

I don't think I'm using neutron, as far as I know I'm using nova-network.
On the example I booted an ubuntu VM and it got the IP 10.0.0.2, and then I
cound't logon on another computer from the lab.

*Here are some lines from dmesg from a regular workstation that was already
logged in while DevStack has a running VM:*

albert ~ $ dmesg
[23898.208033] lockd: server lua.eclipse.ime.usp.br not responding, still
trying
[23898.212053] lockd: server lua.eclipse.ime.usp.br not responding, still
trying
[23907.504018] lockd: server lua.eclipse.ime.usp.br not responding, still
trying
[23907.504038] lockd: server lua.eclipse.ime.usp.br not responding, still
trying
[23907.504041] lockd: server lua.eclipse.ime.usp.br not responding, still
trying
[23921.824036] lockd: server lua.eclipse.ime.usp.br not responding, still
trying
[24041.754304] lockd: server lua.eclipse.ime.usp.br OK
[24041.754316] lockd: server lua.eclipse.ime.usp.br OK
[24101.753756] lockd: server lua.eclipse.ime.usp.br OK
[24161.753178] lockd: server lua.eclipse.ime.usp.br OK
[24161.808018] lockd: server lua.eclipse.ime.usp.br not responding, still
trying

*Here is my localrc (controller)*

[stack@*controller* ~]$ *cat devstack/localrc*
#VIRT_DRIVER=docker

#SERVICE_HOST=10.11.0.40 # REMOVE THIS LINE FOR THE CONTROLLER
# Stop DevStack polluting /opt/stack
DESTDIR=/opt/stack/src/openstack

# Switch to use QPid instead of RabbitMQ
disable_service rabbit
disable_service n-cpu
enable_service qpid
#enable_service qpid, n-cpu,n-net,n-api,n-vol

# Replace with your primary interface name
HOST_IP_IFACE=em1
PUBLIC_INTERFACE=em1
VLAN_INTERFACE=em1
FLAT_INTERFACE=em1

# Replace with whatever password you wish to use
MYSQL_PASSWORD=badpassword
SERVICE_TOKEN=badpassword
SERVICE_PASSWORD=badpassword
ADMIN_PASSWORD=badpassword

# Pre-populate glance with a minimal image and a Fedora 17 image
IMAGE_URLS="
http://launchpad.net/cirros/trunk/0.3.0/+download/cirros-0.3.0-x86_64-uec.tar.gz#,
http://berrange.fedorapeople.org/images/2012-11-15/f17-x86_64-openstack-sda.qcow2
"

#ENABLED_SERVICES=n-cpu,n-net,n-api,n-vol # REMOVE THIS LINE FOR THE
CONTROLLER

*Here is my localrc on any compute node:*

[stack@*compute02* ~]$ cat *devstack/localrc*
#SERVICE_HOST=10.7.22.7 # REMOVE THIS LINE FOR THE CONTROLLER
SERVICE_HOST=marte.eclipse.ime.usp.br # REMOVE THIS LINE FOR THE CONTROLLER
# Stop DevStack polluting /opt/stack
DESTDIR=/opt/stack/src/openstack



# Switch to use QPid instead of
RabbitMQ

##disable_service rabbit c-sch c-api
c-vol

disable_service
c-vol

#disable_service
g-api

#disable_service
g-reg

disable_service key
disable_service n-crt
disable_service n-obj
disable_service n-cond
disable_service cinder
disable_service c-sch
disable_service c-api
disable_service c-vol
disable_service c-sch
disable_service n-novnc
disable_service n-xvnc
disable_service n-cauth
disable_service horizon
disable_service rabbit
disable_service tempest
##disable_service mysql
##enable_service qpid
#enable_service n-cpu,n-net,n-api,n-vol
#
g-api,g-reg,key,n-api,n-crt,n-obj,n-cpu,n-net,n-cond,cinder,c-sch,c-api,c-vol,n-sch,n-novnc,n-xvnc,n-cauth,horizon,rabbit,tempest,mysql,n-cpu,n-net,n-api,n-vol

disable_all_services
enable_service qpid n-cpu n-net n-api n-vol


# Replace with your primary interface name
HOST_IP_IFACE=em1
PUBLIC_INTERFACE=em1
VLAN_INTERFACE=em1
FLAT_INTERFACE=em1

# Replace with whatever password you wish to use
MYSQL_PASSWORD=badpassword
SERVICE_TOKEN=badpassword
SERVICE_PASSWORD=badpassword
ADMIN_PASSWORD=badpassword

# Pre-populate glance with a minimal image and a Fedora 17 image
IMAGE_URLS="
http://launchpad.net/cirros/trunk/0.3.0/+download/cirros-0.3.0-x86_64-uec.tar.gz,http://berrange.fedorapeople.org/images/2012-11-15/f17-x86_64-openstack-sda.qcow2
"

#ENABLED_SERVICES=n-cpu,n-net,n-api,n-vol # REMOVE THIS LINE FOR THE
CONTROLLER
#RABBIT_PASSWORD=d83acc200f2fa34b127a

MYSQL_HOST=$SERVICE_HOST
RABBIT_HOST=$SERVICE_HOST
GLANCE_HOSTPORT=$SERVICE_HOST:9292
Q_HOST=$SERVICE_HOST

*Here is my hosts file (all nodes):*

[stack@*controller* ~]$ cat */etc/hosts*
127.0.0.1   localhost localhost.localdomain localhost4
localhost4.localdomain4
::1         localhost localhost.localdomain localhost6
localhost6.localdomain6
143.107.45.200  marte.eclipse.ime.usp.br marte controller
143.107.45.201  jupiter.eclipse.ime.usp.br jupiter compute01 c01
143.107.45.202  saturno.eclipse.ime.usp.br saturno compute02 c02
143.107.45.193  venus.eclipse.ime.usp.br venus compute03 c03

All of the IPs are public. I didn't understand about the FlatDHCP to be
honest, I will research about it.

Any ideas?

Many thanks!

Regards,
Albert.


On Wed, Sep 25, 2013 at 4:55 PM, John Griffith
<john.griffith at solidfire.com>wrote:

>
>
>
> On Wed, Sep 25, 2013 at 1:26 PM, Aaron Rosen <arosen at nicira.com> wrote:
>
>> Hi Albert,
>>
>> Are you sure this is happening. I'm positive that neutron's dhcp agent
>> will only hand out ip addresses for ports that it knows about and I'm sure
>> nova-network does the same as well.
>>
>> Aaron
>>
>> On Wed, Sep 25, 2013 at 12:17 PM, Albert Vonpupp <vonpupp at gmail.com>wrote:
>>
>>> Hello,
>>>
>>> I'm trying DevStack at the university lab. When I tried to deploy a VM I
>>> noticed that all the machines from the lab started renewing their leases
>>> with the DevStack DHCP server. That is inconvenient for me since I'm not
>>> the only user of this lab and it could cause troubles. I thought that
>>> perhaps changing the default port on the controller as on the compute nodes
>>> would work, but I don't know how to do that.
>>>
>>> How can I change the dnsmasq DHCP port on DevStack? (controller and
>>> compute nodes)
>>>
>>> Thanks a lot!
>>>
>>> Albert.
>>>
>>> _______________________________________________
>>> Mailing list:
>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>> Post to     : openstack at lists.openstack.org
>>> Unsubscribe :
>>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>>
>>>
>>
>> _______________________________________________
>> Mailing list:
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>> Post to     : openstack at lists.openstack.org
>> Unsubscribe :
>> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
>>
>> Hi Albert,
>
> I inadvertently did this once in our lab.  The issue I believe (if my
> memory is correct) you're probably using nova-networking and you've
> configured FlatDHCP.  The problem is that you're that your public network
> is accessing your internal/private network (check your bridge setting) so
> the result is that external DHCP requests can be received from your
> OpenStack private network.
>
> It might be helpful if you include your localrc file and some info
> regarding your systems nics and how they're configured.
>
> John
>
>
>
>


-- 

Albert.

http://www.albertdelafuente.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20130925/8435e2fc/attachment.html>