VMware Cloud Community
SparkyRih
Contributor
Contributor

Can't see VMFS3 datastore in ESXI 5.5

For some reason I can't paste anything in the text box, so please bear with me...

I moved 3 physical HDD's from an ESXi 4.1 to an ESXi 5.5 server. The first 2 160GB disks went fine, but than I switched one of the 160GB disks with a 1TB disk (also containing an existing datastor) but I aacidently made a mistake in the RAID controller, so it removed the partition table of this 1TB disk. If read KB2046610 of the VMware Knowledge Base, and I managed to get a partition on the disk, I can also see the original datastore label while running the script in the CLI, it's called "datastore3", and it is present. But I can't get it to show in vSphere, refreshing or rescanning won't work... Adding new storage doesn't work because it doesn't ask me for the 3 options "keep signature, resignature or format" it goes straight to the "VMFS5 or VMFS3" choice. If I choose one of these it goes toe the next step and shows that there is a VMFS partition present using the entire disk, but if I go on from this step it will format the disk (I didn't do that of course).

Once again I'm sorry that I can't paste anything, it makes it very hard for me to explain the issue and the steps that I've already taken in attempt to solve this issue :S But I hope it's clear so that someone might be able point me in the right direction...

Making IT Possible While Keeping IT Simple
Reply
0 Kudos
7 Replies
a_p_
Leadership
Leadership

Welcome to the Community,

Please take a look at VMware KB: vSphere handling of LUNs detected as snapshot LUNs to persistently mounting the datastore from the command line helps.

André

Reply
0 Kudos
SparkyRih
Contributor
Contributor

It keeps coming back with the error: "No unresolved VMFS snapshots with volume label "VMFS".

Not sure where to get the correct label or UUID from.....

Making IT Possible While Keeping IT Simple
Reply
0 Kudos
a_p_
Leadership
Leadership

>>> "No unresolved VMFS snapshots with volume label "VMFS".

What datastore name (label) do you use in the command. Shouldn't it be "datastore3" rather than "VMFS"?

esxcli storage vmfs snapshot mount -l "datastore3"

can you confirm that the "datastore3" shows up when you run

esxcli storage vmfs snapshot list

André

Reply
0 Kudos
SparkyRih
Contributor
Contributor

Yeah sorry, I tried datastore3, but same result.

The command you advice doesn't return any output...

Making IT Possible While Keeping IT Simple
Reply
0 Kudos
a_p_
Leadership
Leadership

So you say that after recreating the partition you are able to see the partition information and the label correctly?

Just a quick question, is already a datastore attached with the same label "datastore3"? If this is the case, please rename this datastore to see whether this solves the issue (you may need to hit rescan/refresh after renaming the datatastore).

André

Reply
0 Kudos
SparkyRih
Contributor
Contributor

I can see the datastore label here: http://misc.schattorie.nl/esx/esxcli.jpg

There is currently only one other datastore connected called OperatingSystems... So there are no duplicates...

Edit: Is this output of any use for you? It does return an error, the correct label and the UUID: http://misc.schattorie.nl/esx/esxcli2.jpg

I'm not sure about the numbers given In the error message, I can't give them any representation...

Making IT Possible While Keeping IT Simple
Reply
0 Kudos
SparkyRih
Contributor
Contributor

I knew that the disk was still fine... After connecting it to a Ubuntu machine using vmfs-tools I managed to get into the disk and backup the smallest vmdk, and the file appears to be fine (connected it to another VM)... Too bad that I have to wait for a new HDD to be able to backup the larger vmdk which holds the data that I'm worried about, but I think it will be fine as well...

Making IT Possible While Keeping IT Simple
Reply
0 Kudos