LUN file system deleted - how to reuse it now?

[Sateesh_vCloud]

We have a LUN mapped to 2 ESXi 5.1 hosts in LAB. This is coming via FC (IBM SAN) and other Administrator mentioned he deleted the file system (not clear). Now I got a task to re-use the same LUN for another project. Any suggestion about steps to perform?

[a_p_]

If only the file system (i.e. the VMFS datastore) was deleted, the LUN should be available in the "Add Storage" wizard again to create a VMFS Datasore on it.

André

[Sateesh_vCloud]

That is valid point but unfortunately it is NOT  visible in Add storage wizard.

Do we need to check in devices and search for the device ID?

[a_p_]

Assuming the LUN itself is presented to the host, i.e. visible under Storage Adapters, but isn't available in the Add Storage Wizard, you may:

- check whether the LUN shows up as a datastore (which doesn't seem to be the case)

- check whether the LUN is used as a RDM

- the LUN is detected as a Snapshot LUN (see e.g. http://kb.vmware.com/kb/1011387)

André

[Sateesh_vCloud]

- check whether the LUN shows up as a datastore (which doesn't seem to be the case)      -->   yes it is not

- check whether the LUN is used as a RDM --> It is not RDM

- the LUN is detected as a Snapshot LUN  --> No details given about Snapshot LUN

Do we have any chances like

File System deleted and it still hangs in devices?

Using Mounting option will allow us to see the LUN in Add Storage?

Thanks Again

[a_p_]

Strange. So can you confirm that the LUN is presented to the host/initiator from the storage?

Maybe the LUN/datastore has been detached!? In this case is should show up running esxcli storage core device detached list, and you need to remove it from this list running esxcli storage core device detached remove -d NAA_ID. (see http://kb.vmware.com/kb/2004605 for details)

André

[Whaler]

OhMyGosh, Andre, I've got the same problem.  Six LUNS used for Datastores on an old cluster (4.1 I believe) were unmounted & detached, then presented to a server in a new 5.5 cluster.

Looking at the Configuration->StorageAdapter screen for the server's FC-HBA, I see all the LUNS showing up.  But when I try to add Disk/LUN storage to the server, they don't show up on the "Select Disk/LUN" screen.

Using "esxcli storage core device detached list" just shows one device I don't recognize:  naa.60060e80104b276004f3217e00000056  off

Thanks....Lyle

[a_p_]

I assume you already rescanned the storage adapter for new LUNs and VMFS datastores?

What about esxcli storage vmfs snapshot list? If the command shows the "missing" datastores, take a look at VMware KB: vSphere handling of LUNs detected as snapshot LUNs for how to manually mount them.

André

[Whaler]

Rescanned multiple times, and as I say, they are showing up when viewing the HBA storage adapter.

"esxcli storage vmfs snapshot list" run on the server shows blank.

Thanks for the quick reply.  Asking "Andre the Giant" is quicker than logging a support call.  8-)

....Lyle

[a_p_]

Just to rule this out, can you confirm that the LUNs were presented in read/write mode?

Another step in troubleshooting is to use partedUtil getptbl ... to check whether the presented LUNs contain a VMFS partition. Since they were created with ESXi 4.x it may also work to run fdisk -lu to get the partition information!?

André

PS: I'm not Andrew Mauro

PS2: Forgot to post the KB article VMware KB: Using the partedUtil command line utility on ESXi and ESX

[Whaler]

Thanks for the help, Andre(notTheGiant).

SAN administrator assures me read/write is the default presentation, and that's all he uses.

partedUtil says:

# partedUtil getptbl /dev/disks/t10.HITACHI_D60044040015

unknown

104535 255 63 1679370240

fdisk says:

Disk /dev/disks/t10.HITACHI_D60044040015: 859.8 GB, 859837562880 bytes

255 heads, 63 sectors/track, 104535 cylinders, total 1679370240 sectors

Units = sectors of 1 * 512 = 512 bytes

Disk /dev/disks/t10.HITACHI_D60044040015 doesn't contain a valid partition table

.....Lyle

[a_p_]

If this is the case for all six LUNs you need to find out how exactly the datastores were detached. Maybe it's possible to re-create the partition table, but before trying to do this I'd strongly recommend you ask the storage admin to create SAN snapshots or clones of the LUNs to have a way back in case things become worse.

André

[Whaler]

Sorry, I should have made it clear.  I don't need what's on the LUNS, just want to reuse them for new Datastores but can't seem to add them.

Thanks....Lyle

[a_p_]

Out of curiosity, are these LUNs available if you try to add them as RDMs to a virtual machine?

André

[Whaler]

The RDM option is grey'd-out when trying to add a disk to a Windows or Centos VM.

I really appreciate the help, Andre, and as I say, you're very fast with the responses.  But should I place a support call instead?  I'm totally happy to look around & try more if you want to though.

thanks....Lyle

[a_p_]

The only option that comes in mind is to try and clean the partition information (although I think this may not help, as no partition seems to exit anyway). For details see VMware KB: Troubleshooting the Add Storage Wizard error: Unable to read partition information from t...

If this still doesn't help you may open a SR to see what VMware support can figure out. In this case please provide some feedback on what caused the issue, i.e. whether it was related to some stale data on the LUNs, the presentation, an issue with ESXi, ...

André

[Whaler]

Mystery solved....

Support did a webex session, and it turns out that we had to use the vSphere client connected to the one server in the cluster that had the HBA and LUNs.  Then I was able to add the LUNS as Datastores as normal.

Trying to do this from vCenter wouldn't work, although once the Datastores were created via the single server, they magically appeared in the vCenter view.

Thanks to Andre for all the advice.....Lyle