nova-compute should logout an iSCSI target on migration failure

After failing to migrate an instance with an iSCSI volume to HOST A, we can no longer migrate any instances with an iSCSI volume to HOST A.

The error message from nova-compute is as follows:
2011-03-25 11:16:26,640 ERROR nova.rpc [-] ['Traceback (most recent call last):\n', ' File "/home/nttdata/nova/nova/rpc.py", line 183, in _receive\n rval = node_func(context=ctxt, **node_args)\n', ' File "/home/nttdata/nova/nova/compute/manager.py", line 875, in pre_live_migration\n self.volume_manager.setup_compute_volume(context, v[\'id\'])\n', ' File "/home/nttdata/nova/nova/volume/manager.py", line 164, in setup_compute_volume\n path = self.driver.discover_volume(context, volume_ref)\n', ' File "/home/nttdata/nova/nova/volume/driver.py", line 456, in discover_volume\n self._run_iscsiadm(iscsi_properties, "--login")\n', ' File "/home/nttdata/nova/nova/volume/driver.py", line 428, in _run_iscsiadm\n iscsi_command)\n', ' File "/home/nttdata/nova/nova/utils.py", line 173, in execute\n cmd=\' \'.join(cmd))\n', "ProcessExecutionError: Unexpected error while running command.\nCommand: sudo iscsiadm -m node -T iqn.2010-10.org.openstack:volume-00000001 -p 10.192.12.34:3260 --login\nExit code: 255\nStdout: 'Logging in to [iface: default, target: iqn.2010-10.org.openstack:volume-00000001, portal: 10.192.12.34,3260]\\n'\nStderr: 'iscsiadm: Could not login to [iface: default, target: iqn.2010-10.org.openstack:volume-00000001, portal: 10.192.12.34,3260]: \\niscsiadm: initiator reported error (15 - already exists)\\n'\n"]