Hyper-V unable to find mounted disks when attaching volume to instance on IBM Storwize V7K
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Cinder |
Fix Released
|
Critical
|
Jay Bryant | ||
Icehouse |
Fix Released
|
Undecided
|
Unassigned |
Bug Description
Steps to create:
1.) Create and instance, hosted on a Hyper-V system
2.) Create a new volume that is carved out of IBM Storwize V7K
3.) Confirm both are available/active
3.) Attach volume to instance
$ nova volume-attach ldbragst-instance ldbragst-volume auto
The following stacktrace was pulled from the nova logs on the hyper-v node that hosts the instance.
[Hyper-V Volume Attach Issue]: Unable to find a mounted disk
http://
After this happens I can confirm that an iSCSI connection exists and is 'Connected' through the iSCSI Initiator. The iqn that it is connected to is the same iqn that is in the above stacktrace: iqn.1986-
If I don't attempt to clean up anything on the Hyper-V node and don't 'Disconnect' the Discovered Targets (iqn.1986-
[Hyper-V Volume Attach Issue]: WMI Error Code
http://
Which I believe is being addressed in https:/
Here, what I think is happening is that the device_number here https:/
Here is some manual testing I did in a Python terminal on the compute node, specifically testing the case here https:/
summary: |
Hyper-V unable to find mounted disks when attaching volume to instance + on IBM Storwize V7K |
Changed in cinder: | |
status: | Confirmed → In Progress |
Changed in cinder: | |
status: | Fix Committed → Fix Released |
tags: | removed: icehouse-backport-potential |
Changed in cinder: | |
milestone: | juno-1 → 2014.2 |
Hello Alex, outside of the linked traces and testing I did in Powershell, can you think of any additional things to try? We've been able to map a volume from the iSCSI target to the compute node running Hyper-V manually.