After enabling space-allocation raw device mapping has all paths down
Applies to
- ONTAP
- VMware
- Windows Server Failover Clustering (WSFC)
- Physical Raw Device Mapping (RDM) LUNs
- Cluster Across Boxes
Issue
- After setting a LUN offline and online(to enable space_allocation), all related paths stay down in one of two ESXi hosts
- Windows disk was offlined within the Microsoft cluster node which is a virtual machine (VM) before the procedure but not ESXi
- The other host with the VM cluster node still attaching the RDM in question has no access issues.
- On the affected host:
- Initiator rescans do not rediscover the LUN.
esxcli storage core device list
still contains the LUNesxcli storage nmp path list
displays all paths for that LUN as permanently lost.esxcli storage core device world list
displays the VM name with the NAA of the LUN in question
Example:
[root@esx:~] esxcli storage core device world list | grep my-vm
naa.600a0980774f6a352d5d4a684exxxxxx 2121675 1 vmm0:my-vm