Upgrading to Havana "Failed to invoke ems"

[SAM_MORRISON]

Just upgraded our test environment to Havana and am getting the following error:

2013-11-15 14:50:41.698 30720 INFO cinder.volume.manager [-] Updating volume status
2013-11-15 14:50:41.699 30720 DEBUG cinder.volume.drivers.netapp.iscsi [-] Updating volume stats _update_volume_stats /usr/lib/python2.7/dist-packages/cinder/volume/drivers/netapp/iscsi.py:1040
2013-11-15 14:50:41.699 30720 WARNING cinder.volume.drivers.netapp.iscsi [-] Cluster ssc is not updated. No volume stats found.
2013-11-15 14:50:41.701 30720 DEBUG cinder.openstack.common.lockutils [-] Got semaphore "safe_set_attr" for method "set_safe_attr"... inner /usr/lib/python2.7/dist-packages/cinder/openstack/common/lockutils.py:191
2013-11-15 14:50:41.702 30720 DEBUG cinder.openstack.common.lockutils [-] Got semaphore "refresh_ssc_44796880" for method "get_latest_ssc"... inner /usr/lib/python2.7/dist-packages/cinder/openstack/common/lockutils.py:191
2013-11-15 14:50:41.703 30720 INFO cinder.volume.drivers.netapp.ssc_utils [-] Running cluster latest ssc job for server: netapp-iscsi-qh2-test.mgmt.melbourne.rc.nectar.org.au and vserver rc_test_volumes
2013-11-15 14:50:41.787 30720 DEBUG cinder.volume.drivers.netapp.utils [-] Failed to invoke ems. Message : NetApp api failed. Reason - Not found:No records found provide_ems /usr/lib/python2.7/dist-packages/cinder/volume/drivers/netapp/utils.py:114
2013-11-15 14:50:42.107 30720 DEBUG cinder.openstack.common.lockutils [-] Got semaphore "safe_set_attr" for method "set_safe_attr"... inner /usr/lib/python2.7/dist-packages/cinder/openstack/common/lockutils.py:191

So we can't schedule any volume builds because the scheduler thinks the netapp doesn't have any capacity.

We are using stable/havana (HEAD = ad05556a2f3acf11e7ab633fa59bf23a6cad11da)

Our ONTAP version is: 8.2P2

[SAM_MORRISON]

Should mention our cinder.conf options

netapp_server_port = 443
netapp_storage_family = ontap_cluster
netapp_storage_protocol = iscsi
netapp_transport_type = https
volume_driver = cinder.volume.drivers.netapp.common.NetAppDriver

[esker]

Hi Sam,

What occurs when you run "event status show" from the Clustered ONTAP console?

Thanks!

[singn · Accepted solution]

Hi Sam. It arises because of the vsadmin role problem. The issue has been fixed in the change https://review.openstack.org/#/c/60183/ The file to fix ems in particular will be cinder/volume/drivers/netapp/utils.py. Let me know if it works for you.

Thanks

Navneet