VMware recently discovered a bug in CBT that can cause data coruption when trying to restore a backed up virtual machine. This can result in missing data and blue screen on restored VM.
This only seems to effect VMs that disk/disks have been expanded beyond 128GB.
http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2090639
Per the VMware KB article mentioned above:
"The Virtual Disk Development Kit (VDDK) method QueryChangedDiskAreas() is used by some backup applications to determine which areas of a virtual disk have changed since the last backup, this information is used to decide which blocks to include in a backup.
To work around this issue with any backups going forward, CBT needs to be reset. To manually reset CBT on VM you need to disable CBT >> stun/un-stun (taking and deleting of snapshot) >> enable CBT >> stun/un-stun VM.
Attached is a powershell script that can be run in the vRanger console. It automates the process by checking for VMs 128GB or larger and runs through the process outlined above.
The script requires VMware powecli to be installed on the vRanger machine, the script will try and load the snapin and warns if not installed.
Powercli can be downloaded from link below but does require vmware login credentials to download.
https://my.vmware.com/web/vmware/details?downloadGroup=PCLI550&productId=352
Steps:
1. Download script to vRanger box
2. Open the vRanger console
3. change directoy to where script was placed.
4. run the command ./CBT_reset_01.ps1
5. When prompted enter virtual center or ESX host information.
6. select y/n to use credentials you are currently logged into vRanger box with to authenticate to VC. If you select no you will be prompted for credentials with windows pop-up.
7. once list is enumerated you will be prompted to select y/n to reset CBT
**NOTE Any VMs that have disks over a 128GB will have message in red underneath stating "Disk larger than 128GB attached"
