I have lost a Datastore from my ESXi Host. The drive is still accessable, but it does not show up as Datastore (but as device). I can connect to it via SSH and copy data from it (I am getting all the vmdx at the moment, in case the drive dies).

Following issue came up in the log:

Host Connectivity Degraded in ESX/ESXi (1009557)

But it recovered

Host Connectivity Restored (1009556)

I can add it as a new Datastore but I would have to reformat it.

Is there a way to remount the drive so it is detected as a datastore?

It is also shown in the esxcli storage vmfs extent list

esxcli storage vmfs extent list
Volume Name  VMFS UUID                            Extent Number  Device Name                                                                Partition
-----------  -----------------------------------  -------------  -------------------------------------------------------------------------  ---------
datastore1   4cf78219-37a48610-9f53-001b21811bd8              0  t10.ATA_____SAMSUNG_HE502HJ_______________________________S2B6J90ZA08075           3
datastore3   4ea13fa7-6e0b3afc-8741-b8ac6f95a15a              0  t10.ATA_____ST2000DL0012D9VT156__________________________________5YD4J7JA          1
datastore2   4cf7821a-60777308-3711-001b21811bd8              0  t10.ATA_____SAMSUNG_HE502HJ_______________________________S2B6J90ZA08046           3


It reapeared after rescanning the drives, but it had an invalid Size, short after the drive disapeared again. Doesn't look good... but the data transfer via scp still works.

The blocksize of the device changed from 1 MB to 175 MB thuse the change in the device size

  • What the actual question here? – GregL Apr 13 '15 at 11:32
  • How can I remount this datastore so it can be used again by the esxi host - what might be the issue causing this behaviour and how to fix it – Maximilian Kindshofer Apr 13 '15 at 11:35
  • Do you have RAID? – ewwhite Apr 13 '15 at 12:27
  • No RAID installed, but there is a RAID-Controller inside the Host. The issue seems to be, that somehow the blocksize of the device became invalid – Maximilian Kindshofer Apr 13 '15 at 12:29

According to the vSphere documentation esxcli does indeed have a mount subcommand.

  1. List all volumes that have been detected as snapshots.
    esxcli <conn_options> storage filesystem list
  2. Run esxcli storage filesystem mount with the volume label or volume UUID. By default, the volume is mounted persistently, use --no-persist to mount persistently.
    esxcli <conn_options> storage filesystem volume mount --volume-label=<label>|--volume-uuid=<VMFS-UUID>

So for your case, you can run the following commands to get the VMFS remounted.

esxcli <conn_options> storage filesystem volume mount --volume-uuid=4cf78219-37a48610-9f53-001b21811bd8
esxcli <conn_options> storage filesystem volume mount --volume-uuid=4ea13fa7-6e0b3afc-8741-b8ac6f95a15a
esxcli <conn_options> storage filesystem volume mount --volume-uuid=4cf7821a-60777308-3711-001b21811bd8
  • This worked kind of - it reapeared with a size of 7,497,075.00 Tb. What could cause this? The other details and the partitions seem to be ok. - OK it looks like the block size changed from 1 MB to 175 MB – Maximilian Kindshofer Apr 13 '15 at 12:06
  • And what size is the volume actually? What does esxcli <conn_options> storage filesystem list report as the size? – GregL Apr 13 '15 at 12:09
  • It fails to get the information on the Error getting data for filesystem on '/vmfs/volumes/4cf7821a-60777308-3711-001b21811bd8': Unable to get FS Attrs for /vmfs/volumes/4cf7821a-60777308-3711-001b21811bd8, skipping. But it should be 460,75 GB. The shown size is about 175 times as much – Maximilian Kindshofer Apr 13 '15 at 12:15
  • It can't get the drive information? Is this local storage or remote? – GregL Apr 13 '15 at 12:30
  • 2
    Yeah, sounds like something has gone sideways with that disk. As you say, I'd get everything off it ASAP and replace it. – GregL Apr 13 '15 at 12:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.