After successfully creating my vSAN environment, I deleted the resources associated with it in vCenter since I needed to document, with screen shots the process of creating it.
What I found, the method I used to delete from vCenter had it short comings, namely the ESXi still believe the disks are still vSAN disks, and has such can not be used. For instance ...
vdq -q
{
"Name" : "naa.5000cca09500c2c4",
"VSANUUID" : "52df7db4-7f26-268f-3a66-5005d88d58df",
"State" : "In-use for VSAN",
"Reason" : "None",
"IsSSD" : "1",
"IsCapacityFlash": "0",
"IsPDL" : "0",
}
failed attempts to remedy the condition include commands such as this:
[root@mran01:/etc] partedUtil delete /dev/disks/naa.5000cca09500c2c4 1
Error: Read-only file system during write on /dev/disks/naa.5000cca09500c2c4
Unable to delete partition 1 from device /dev/disks/naa.5000cca09500c2c4
Even after: chmod 755 /dev/disks/naa.5000cca09500c2c4
the result is still the same.
Using the browser to get into the ESXi to either Edit the partition table, or delete the partition table failed as well.
Any suggestions on how to clean this up, so I can use it again?
You can try deleting the disk group (and thus making the disks available) from CLI:
esxcli vsan storage list
If the disks show up as part of a disk group from this output, then the command below should let you remove the disk group and free the underlying disks
esxcli vsan storage remove -u uuid
UUID would be the disk group UUID
Hello jediller
Have the disk-groups been removed?
Cluster > Manage > Disk Management > Select Disk-group > Delete (pick option e.g. No Action if you want to permanently delete this without moving anything off it)
(NOTE: This is permanent and any data on the disk-group cannot be retrieved after removal)
Disk Mappings can also be checked via the CLI using 'vdq -Hi' and/or:
# esxcli vsan storage list
If they have been removed then try removing the partitions via the Web Client:
Host > Manage/Configure > Storage > Storage Devices > Select device > 'All Actions' > Erase Partitions
(Again - this is permanent and irreversible)
If this doesn't work (can happen after bad disk issues) there are some other CLI methods.
Bob
The problem, I suspect is when I deleted the prior resources, I deleted the Cluster (and/or removed it from Inventory), before I took the step of deleting the disk group that was in the cluster. Thus, at least some remnants of that group appear to exist.
The only cluster I have now, does not have vSAN enabled, mostly from lack of disks to put in it
From the Web client, I get Failed to clear disk Local ... when I attempt to clear the partition table. Similar result when I try to edit the partition table.
Also tried (has mentioned in the original post):
partedUtil delete /dev/disks/naa.5000cca09500c2c4 1
Error: Read-only file system during write on /dev/disks/naa.5000cca09500c2c4
Unable to delete partition 1 from device /dev/disks/naa.5000cca09500c2c4
as well as partedUtil mklabel
What other utilities can I use from the (ssh) command prompt, or from anywhere else that would help?
You can try deleting the disk group (and thus making the disks available) from CLI:
esxcli vsan storage list
If the disks show up as part of a disk group from this output, then the command below should let you remove the disk group and free the underlying disks
esxcli vsan storage remove -u uuid
UUID would be the disk group UUID
Hello jediller,
It is possible you just removed the vSphere-level cluster but the actual vSAN cluster is still intact - thus try as elerium advised removing the disk-groups or the cache-tier devices e.g. esxcli vsan storage remove -s <SSD-naa> .
If esxcli doesn't work, try localcli.
If this doesn't work, I will PM you in a moment with a sure-fire method that takes slightly longer (PM as it is a process that is generally done under GSS guidance).
Bob