Open Stack

Hi Antonio
I verified your suggestion is fine and now iscsi volumes runs over storage
network.
On yesterday it did not work fike because iscsi_ip_address was in DEFAULT
section.
I created a lvm section like the followinf and now it works:

[lvm]

volume_driver = cinder.volume.drivers.lvm.LVMVolumeDriver
iscsi_helper = tgtadm
# volume group name just created

# IP address of Storage Node

iscsi_ip_address = 192.168.1.62
volumes_dir = $state_path/volumes
iscsi_protocol = iscsi


Many Many thanks for your help

2015-07-17 20:59 GMT+02:00 Antonio Messina <antonio.s.messina at gmail.com>:

> It might only work for new volumes, created after you restart the cinder
> service. We don't have a separate network for srorage, but i had to use it
> while preparing this tutorial:
> https://github.com/gc3-uzh-ch/gridka-school/blob/master/tutorial/overview.rst
> and as far as i remember that option was the only thing needed.
>
> In our case we only had a controller node, but i guess the option must be
> defined where cinder-volume is running
>
> .a.
> Il 17/lug/2015 08:30 PM, "Ignazio Cassano" <ignaziocassano at gmail.com> ha
> scritto:
>
>> I have already tried to insert what you suggested but looking at
>> /var/lib/iscsi directory on computing note I saw its iscsi initiator is
>> still using the target on management network.
>> Must I modify it either on storage nod or controller node ?
>> Does controller node pass the configuratin to computing node ?
>> Many thanks
>> Il giorno 17/lug/2015 19:48, "Antonio Messina" <
>> antonio.s.messina at gmail.com> ha scritto:
>>
>>> I think you have to set iscsi_ip_address option in cinder.conf.
>>>
>>> .a.
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20150718/70ad0af6/attachment.html>