Frequently asked questions about VM data recovery
- VMware boot image corrupted?
- VMFS volume not mounted?
- VMDK file not recognized?
- Need to recover VMDK file?
- vSphere: Unable to read partition?
- ESXi could not read volume header?
- Bank6 (or Bank5): Not a VMWare boot bank No hypervisor found
- VMFS datastore inaccessible?
- ESX Server datastore not visible?
- How to backup and restore ESX installation configuration
VMware boot image corrupted?
During install ESXi Server creates several FAT partitions for its system purposes and forbid access to these partitions to any 3-rd party software. One of these partitions is used to store images of the ESXi Server. During boot of ESXi Server, it creates a RAM disk, loads stored image from forementioned FAT, and proceeds to the system boot from RAM disk. Also, there are multiple checks of integrity of the stored boot image, and effectively prevent it’s corruption.
Should you have any issues with such an error, it may be caused by either updates or hardware. Especially RAM modules as they are the actual disk media for the boot image. If it’s software to blame, try to reinstall ESXi Server.
VMFS volume not mounted?
If VMFS volume not mounted, try to use ESX Service console or Direct console (ESXi). Please note that console is to be enabled before use in Server’s options. To mount VMFS volume use ESXCLI or vicfg-volume commands:
ESX CLITo list all volumes that have been detected as snapshots:
esxcli <conn_options> storage filesystem list
Mount a volume:
esxcli <conn_options> storage filesystem volume mount --volume-label=<label>|--volume-uuid=<VMFS-UUID>
To unmount a volume:
esxcli <conn_options> storage filesystem volume unmount --volume-label=<label>|--volume-uuid=<VMFS-UUID>
To list all available volumes:
vicfg-volume <conn_options> --list
To mount a volume:
vicfg-volume <conn_options> --persistent-mount <VMFS-UUID|label>
Unmount a volume:
vicfg-volume <conn_options> --unmount <VMFS-UUID|label>
Also, you can use VMFS Recovery™ to access VMFS volume. As we use direct disk access, we do not require VMFS volume to be mounted or the ESX server running at all. Connect HDDs as local hard drives or use SSH\iSCSI to access data on the server.
VMDK file not recognized?
A common error when *-FLAT.VMDK misses its descriptor file, a small text *.vmdk. ESX requires both files to exist in the same directory and have corresponding names. Vsphere moves both files by default and also may add necessary changes to configuration files for proper work of VM in a new location. However, if you have used SSH and forgot to copy descriptor files, you may create a new VM disk with the same name and features that had one with lost descriptor and then change new *.flat.VMDK with an older one.
Need to recover VMDK file?
If the VMDK file can not be found on the disk, you need to run VMFS Recovery™ in Uneraser or Full Recovery-> VMFS modes and check if data is recoverable. Demo version can mount recovered VMDK files and preview files inside VM to make sure that integrity is preserved.
Please note: if the VMDK file was lost because of deletion or during moving data from one place to another, in this case, there is a vSphere disk driver involved. This driver often erases links to the data pieces on the disk, which makes further recovery very complicated. Yes, the data is still on the disk, and you did stop all the operations, and yes, VMFS is claimed it has got a journal. But in reality, the journal doesn't have records of previous files, it is used for multi-user transactions only and because of fragmentation, it's impossible to reconstruct file pieces in correct order not to mention they are to be found and distinguished between empty space and other files pieces first.
vSphere: Unable to read partition?
Such a problem is normally caused when the partition table doesn’t comply with MS-DOS partitioning standard or the sector 0 is corrupt.
Solution without recovering data on the disk would be format HDD in MS-DOS compatible layout. If the format command fails or there are valuable data present on the disk, you need to proceed to recovery. We strongly recommend creating a disk image from such drives to reduce the load on the damaged disk and stop further possible degradation. After creating a disk image, mount it to VMFS Recovery™, scan in Reader\Uneraser\Full Recovery modes and check the integrity of recovered data.
ESXi could not read volume header?
Most probable cause of this error is a formatted or heavily damaged VMFS disk. In this case, you need to run VMFS Recovery™ in Uneraser or Full Recovery-> VMFS modes and check if data is recoverable. Demo version can mount recovered VMDK files to its disk list and allow preview files inside VM to make sure that integrity is preserved.
Bank6 (or Bank5): Not a VMWare boot bank No hypervisor found
This error message means that ESX failed to locate its installation. This may be because a boot record or a partition table was altered, damaged or these clusters became bad blocks. Hypervisor partition was damaged corrupted or removed.
The solution to this issue would be reinstalling an ESX new. To speed up and ease the process, you can install ESX on a USB flash drive or SD card many vendors integrate into their hardware. The performance will not be an issue, as ESX creates a RAM disk before boot and transfer its installation to RAM entirely. A little delay in boot won't affect workflow performance at all.
The ESXi installer, will prompt you not to overwrite the existing VMFS disk, but we'd back up the VMFS partition anyway, just in case.
Also, you can gain access to the VMFS disk and extract information using VMFS Recovery™. It can be done by one of the following:
- Boot ESX's hardware to Windows using WindowsPE or installing it onto external drive
- Boot with Linux Live CD and configure it to act like a iSCSI target, then connect from Windows remote machine. You can also use SSH, but it has limited performance and is less recommended
- Boot a back up or fresh ESX from USB stick\SD card as described above and use SSH connection and VMFS Recovery™ to extract all the information.
- Extract RAID disks and reconstruct them on another machine
Next, use VMFS Recovery™ as described in the guides to recover your data.
VMFS datastore inaccessible?
ESX Server datastore not visible?
There are too many possible reasons, that may cause such issue. You need to narrow possibilities by checking hardware and software configuration. Check if hardware works fine using server logs and monitoring and proceed to software.
How to backup and restore ESX installation configuration
You can back up ESX\ESXi configuration to re-use it on fresh installation and quickly restore production state of the server. Use CLI and following commands to operate:
Backup the host config to the file C:\backup.txt:
Restore a saved config C:\backup.txt to the host:
Back up ESX configuration regularly to avoid problems.