[Openstack-operators] nova milestone - issue with iscsi volumes

Linux Datacenter linuxdatacenter at gmail.com
Fri Feb 3 11:59:11 UTC 2012


I fixed it by launching discovery in function connect_volume in
virt/libvirt/volume.py:

(out, err) =
utils.execute('iscsiadm','-m','discovery','-t','st','-p',iscsi_properties['target_portal'],run_as_root=True)

 Cheers,

On 2 February 2012 20:38, Linux Datacenter <linuxdatacenter at gmail.com>wrote:

> Hi,
>
> I've run into an issue with nova-volume.
>
> Each time I want to attach a volume to an instance, I get the following
> error on the compute:
>
> 2012-02-02 20:33:04,546 DEBUG nova.utils [-] Running cmd (subprocess):
> sudo nova-rootwrap iscsiadm -m node -T
> iqn.2010-10.org.openstack:volume-00000011 -p <NOVA_VOLUME_IP>:3260 --rescan
> from (pid=19855) debug /usr/lib/python2.7/dist-packages/nova/log.py:175
> 2012-02-02 20:33:04,596 DEBUG nova.virt.libvirt.volume [-] iscsiadm
> ('--rescan',): stdout= stderr=iscsiadm: No portal found.
>  from (pid=19855) debug /usr/lib/python2.7/dist-packages/nova/log.py:175
>
> The error disappears when I perform a target discovery on the compute
> node, and once more try issue euca-attach-volume:
>
> iscsiadm -m discovery -t st -p <NOVA_VOLUME_IP>
> 10.189.0.31:3260,1 iqn.2010-10.org.openstack:volume-cc1-00000011
>
> euca-attach-volume (works well this time)
>
> Anyone knows if this is a bug and how to work it around? (I tried to
> inject discovery into the volume driver code and libvirt code. It's not
> executed - apparently I put it into some wrong place in the code)
>
> Cheers,
>
>
> --
> checkout my blog on linux clusters:
> -- linuxdatacenter.blogspot.com --
>



-- 
checkout my blog on linux clusters:
-- linuxdatacenter.blogspot.com --
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-operators/attachments/20120203/00922d2d/attachment-0002.html>


More information about the Openstack-operators mailing list