Running VMware’s ESXi and diligently backing up your data in the belief it’s safe as houses? Think again.
VMware has quietly ‘fessed up to the existence of a bug affecting all versions of its bare-metal hypervisor. It copped the problem in its knowledge base as users began cottoning on to the fact something was amiss in their data backups.
The bug affects virtual machines with Changed Block Tracking (CBT) turned on and that have been increased in size by more than 128Gb.
The problem only presents itself when it comes to execution of the command QuaryChangedDisckAreas (“*”). It's only then that portions of the virtual machine disk vmdk files are returned and you get an inaccurate list of allocated virtual machine disk sectors.
All this means you’ll have been happily pouring gigabytes of data into your newly expanded virtual machine blissfully unaware you likely won’t see it again.
The problem affects VMware ESXi 4.x and ESXi 5 and the virty giant admits it is stumped.
In its VMware knowledge base article here, VMware said it has not got a fix.
128GB won't affect everybody running ESXi, but it will hit those running very large clouds and virtual instances.
For really big VMware customers running thousands of virtual machines that keep historical backups, this hidden bug poses a major headache.
Virtualisation management specialist Veeam Software told its customers it is working on a hot fix versions 7.0 and 8.0 of its backup and replication software.
In the meantime, Veeam recommended a manual CBT reset for all expanded VMs while VMware recommended turning it off and on again "disabling and then enabling Changed Block Tracking (CBT) on the virtual machine".
The Reg contacted VMware to find out whether it’s working on a permanent fix and when it would be delivered.
Also, we asked when the giant became aware of the problem, but a company spokesperson said it didn’t have a comment.
Thanks to Reg reader Mark Lansley for the tip. ®