The host was decommissioned and the initiators were not logged in for weeks prior to running the order.
The user is unable to successfully run an unexport and remove block volume with the following errors:
The controllersvc logs show the following errors:
vipr5 al1dvipre.cn5 controllersvc <date><time> [10752|zoneExportMasksDelete|eb23041b-86eb-4baf-b14d-3d8ff7954d2f740e8ea4-0705-4919-9c02-172fa3ef3e33] INFO NetworkScheduler.java (line 1087) Generating remove zoning targets for ExportMask <export mask> (urn:storageos:ExportMask:f0a54fd5-d1b5-470f-a0e8-81a0b0e12345:vdc1)
vipr5 al1dvipre.cn5 controllersvc <date><time> [10752|zoneExportMasksDelete|eb23041b-86eb-4baf-b14d-3d8ff7954d2f740e8ea4-0705-4919-9c02-172fa3ef3e33] INFO NetworkUtil.java (line 114) networklite could not be found for endpoint <host initiator WWN>
vipr5 al1dvipre.cn5 controllersvc <date><time> [10752|zoneExportMasksDelete|eb23041b-86eb-4baf-b14d-3d8ff7954d2f740e8ea4-0705-4919-9c02-172fa3ef3e33] WARN NetworkScheduler.java (line 632) Could not find any network systems that can be used to zone.
vipr5 al1dvipre.cn5 controllersvc <date><time> [10752|zoneExportMasksDelete|eb23041b-86eb-4baf-b14d-3d8ff7954d2f740e8ea4-0705-4919-9c02-172fa3ef3e33] ERROR NetworkDeviceController.java (line 1374) Exception zoning delete Export Masks
java.lang.NullPointerException
at com.emc.storageos.networkcontroller.impl.NetworkScheduler.getZoningNetworkSystems(NetworkScheduler.java:582)
at com.emc.storageos.networkcontroller.impl.NetworkScheduler.unexportVolumes(NetworkScheduler.java:1254)
at com.emc.storageos.networkcontroller.impl.NetworkScheduler.getZoningRemoveTargets(NetworkScheduler.java:1119)
...
vipr5 al1dvipre.cn5 controllersvc <date><time> [10752|zoneExportMasksDelete|eb23041b-86eb-4baf-b14d-3d8ff7954d2f740e8ea4-0705-4919-9c02-172fa3ef3e33] INFO WorkflowService.java (line 652) Updating workflow step: eb23041b-86eb-4baf-b14d-3d8ff7954d2f740e8ea4-0705-4919-9c02-172fa3ef3e33 state ERROR : Failed to remove zones for removed export mask because of: null
The following is observed in the environment:
- There is an empty export group in the ViPR Controller database related to the order. The export group contains no hosts, initiators, or volumes.
- The storage was unmapped on the XtremIO array.
- The volume was not deleted.
- The zones were not removed from the fabric.
The host's initiators were no longer logged into the fabric because the host was decommissioned. ViPR Controller returned a null value when the initiator could not be found causing the order to fail.
Workaround:
If the host is up and the initiators are still logged into the fabric this error will not occur.
Interim Resolution:
Manual clean up of the array, switch, and ViPR Controller database may be required. A Service Request should be opened to investigate the issue further.
Resolution:
ViPR Engineering is currently addressing this problem, but has not provided a fix in a released patch. This solution will be updated with the patch when it has been released.