[Openstack-operators] [Openstack] OpenStack on Hetzner

George Shuklin george.shuklin at gmail.com
Mon May 12 14:19:19 UTC 2014


(note: please, reply to maillist too. When I reply to maillist - this is 
community activity, when I reply to someone in private, this is 
consulting service).

All operations below can cause loose access to server if something goes 
wrong (typo, wrong iface/bridge name, ip address, etc). And they should 
be ran under screen/nohup/local console, not from plain SSH.

You need:
1. to remove IP from eth0 (ip addr del)
2. to add interface eth0 to br-ex (ovs-vsctl add-port)
3. to add IP (from step 1) to br-ex (ip addr add - do not forget to add 
address with netmask /xx )
4. Fix back default routing (ip route add)

I'll not give you exact commands to type in, because plain copypaste 
with 99.98% probability will leave you without working server. Think, 
test and add agruments you need

P.S. And you need to change your networking configuration scripts to 
make those changes permanent.

One important notice: operations with ovs are permanent by default - if 
you add eth0 to br-ex, there gonna be no 'magic reboot restore all'.

On 05/12/2014 05:07 PM, Marius Stanca wrote:
> Hello George,
>
> Thank you a lot for your help.
>
> Hmm, Hetzner doesn't allow to have two or more NIC.
> I will try to move management IP to bridge and move interface to br-ex,
> so can you advise me to do not lose net access?
> Also, can you tell me if you know a Host that offer two or more NIC?
>
> Thank you!!!
> Best,
>
> Marius
>
>
> Pe 5/11/2014 9:51 PM, George Shuklin a scris:
>> Heh, systems with single interface are suck. Technically, you can use it
>> for management and client traffic, but my advice - add more network cards
>> to system or use other server with two or more interfaces.
>>
>> If you really can not, than move management IP to bridge and move interface
>> to br-ex. Note: if you make mistake here, you will lose net access to you
>> hetzner desktop they call 'server'.
>> On May 11, 2014 9:24 PM, "Marius Stanca" <salut at mariuss.net> wrote:
>>
>>> Hello,
>>>
>>> Thank you for reply.
>>>
>>> 1. It seems eth1 is not found:
>>> tcpdump: eth: No such device exists
>>> (SIOCGIFHWADDR: No such device)
>>>
>>> P.S. I have only one interface (eth0).
>>>
>>> 2. I did.
>>>
>>> 3.
>>> 10: 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
>>> 11: qr-1de7a5a2-10: <BROADCAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state
>>> UNKNOWN
>>>      link/ether fa:16:3e:88:7a:01 brd ff:ff:ff:ff:ff:ff
>>>      inet 10.0.0.1/24 brd 10.0.0.255 scope global qr-1de7a5a2-10
>>>      inet6 fe80::f816:3eff:fe88:7a01/64 scope link
>>>         valid_lft forever preferred_lft forever
>>>
>>> I print the openvswitch logs:
>>>
>>> 2014-05-11T01:22:01Z|00055|vlog|INFO|opened log file
>>> /var/log/openvswitch/ovs-vswitchd.log
>>> 2014-05-11T16:16:18Z|00056|bridge|INFO|bridge br-ex: added interface
>>> br-ex on port 65534
>>> 2014-05-11T16:16:18Z|00057|bridge|INFO|bridge br-ex: using datapath ID
>>> 00002a899f2cd442
>>> 2014-05-11T16:16:18Z|00058|connmgr|INFO|br-ex: added service controller
>>> "punix:/var/run/openvswitch/br-ex.mgmt"
>>> 2014-05-11T16:16:22Z|00059|netdev_linux|WARN|ethtool command
>>> ETHTOOL_GFLAGS on network device eth1 failed: No such device
>>> 2014-05-11T16:16:22Z|00060|dpif|WARN|system at ovs-system: failed to add
>>> eth1 as port: No such device
>>>
>>>
>>>
>>>
>>> Pe 5/9/2014 12:13 PM, George Shuklin a scris:
>>>> tcpdump -ni eth eth1
>>> --
>>> Stanca Marius
>>> email: salut at mariuss.net | web: www.mariuss.net
>>> tel: +4.0767.955.578 | skype: wmariuss
>>>




More information about the OpenStack-operators mailing list