Comment 7 for bug 1398078

jgriffith pointed this out:

http://logs.openstack.org/17/117417/11/gate/gate-tempest-dsvm-neutron-pg-full/c3fe760/logs/screen-c-vol.txt.gz#_2014-11-26_14_16_02_494

Sounds like that shouldn't happen, bug since it's logged at debug level logstash won't pick it up so we don't know the correlation to gate failures. He's going to push a change to make that WARNING level so we can see if it's related.

The weird thing is that when looking at the command failure here:

http://logs.openstack.org/17/117417/11/gate/gate-tempest-dsvm-neutron-pg-full/c3fe760/logs/screen-c-vol.txt.gz#_2014-11-26_14_16_03_152

That's doing an op new but the code at:

File "/opt/stack/new/cinder/cinder/brick/iscsi/iscsi.py", line 252, in create_iscsi_target

is doing an op show. It seems like there is some processutils/oslo.concurrency weirdness in here.