Questions on Changed Block Tracking (CBT)

[coolgod]

what mechanism is used to determine changed block by veeam when CBT is not used ?

what's the performance and processing impact when CBT is disabled ?

[Shestakov]

Hello,

what mechanism is used to determine changed block by veeam when CBT is not used ?

When CBT is not in use it fails over to Veeam’s proprietary filtering mechanism.
Instead of tracking changed blocks of data, VBR filters out unchanged data blocks. During backup, VBR consolidates virtual disk content, scans through the VM image and calculates a checksum for every data block. Checksums are stored as metadata to backup files next to VM data. When incremental backup is run, VBR opens all backup files in the chain of previous full and incremental backups, reads metadata from these files and compares it with checksums calculated for a VM in its current state. If a match is found (which means the block already exists in the backup), the corresponding block is filtered out.

what's the performance and processing impact when CBT is disabled?

If you disable CBT same blocks will be filtered, but with less efficient mechanism, means it will take more time.
Thank you.

[coolgod]

thanks shestakov for clarifying this.

[NightBird]

What about this VMware kb ?
http://kb.vmware.com/selfservice/micros ... Id=2090639

What do Veeam think about it ?

[VladV]

Interesting. Why would invalidating the CBT by expanding the disk jeopardize a running VM? Curious about the explanation on why this would happen.

[Vitaliy S.]

What do Veeam think about it ?

When virtual disk allocation is changed, then Veeam backup server uses our own proprietary change tracking engine to transfer changed data. See this topic for more info > Procedure for expanding virtual disks

[NightBird]

So what about Gostev forum digest post ? if Veeam is so magical

[TommyB]

So what about Gostev forum digest post ? if Veeam is so magical

Regarding Gostev's email and the CBT issue - I would like to know if there is any way to find out if my VM's are actually affected by this bug?

[foggy]

Only actual restore attempt (in a virtual lab, using SureBackup, for example) will tell you whether the restore point is recoverable. Data integrity on data disks should be verified manually.

Btw, there's a separate discussion regarding that CBT issue, covering similar questions. Worth reviewing.

[TommyB]

Thanks foggy