Replication and backup of VM at the same time

[veremin]

Though I didn’t have a chance to test the tool you’ve cited earlier, probably, it will meet your expectations. However, you should use it at your own risk.

Apart from that, the only available option is to let sdelete tool zeroing out blocks from deleted files and perform migration right after it.

Thanks.

[mephisto]

I did the vmotion after sdelete, the disks are inflated to the maximum and the VM still using lofs of space, the disks were not shrinked.

How can I shrink those disks? Are you guys sure it's not like on this guide:

http://kb.vmware.com/selfservice/micros ... Id=2004155

Storage vMotion to thin disk does not reclaim null blocks
Symptoms
When using Storage vMotion to thin disk between two VMFS volumes with the same block size, null blocks are not reclaimed.
Null blocks are copied from source to destination resulting in a VMDK that is the same size, despite thin being chosen as the destination disk type.
When attempting to thin a disk after either using VMware Tools Prepare to shrink option or using a third-party product to zero disk blocks within a guest operating system, null blocks are copied.

To reclaim the unused space of a virtual disk in ESXi/ESX 4.1 or later:

Ensure the disk has no Snapshots.

In a Windows virtual machine, use the SDelete command (or a tool with similar functionality) to zero out all unused space.If you use SDelete, ensure that you use version 1.6 or later.

Note: Zeroing all unused blocks inflates the disk to its full size and converts it into an eagerzeroed disk. If the original disk is a thin provisioned disk, ensure there is sufficient space on the datastore to allow the disk to grow to its full size. For more information, see Determining if a VMDK is zeroedthick or eagerzeroedthick (1011170).

Shut down the virtual machine or temporarily remove the virtual disk from the virtual machine to ensure it is not in use.

Erase all unused blocks by running the command:

vmkfstools -K /path/to/disk-name.vmdk
This option de-allocates all zeroed out blocks and leaves only those blocks that were allocated previously and contain valid data. The resulting virtual disk is in thin format. For more information on the vmkfstools command, see Removing Zeroed Blocks in the vSphere documentation.

[veremin]

My answer regarding vmkfstools –k was completely based on this VMware KB.

And yes, It seems to be a known issue that if are using Storage vMotion to thin disk between two VMFS volumes with the same block size, null blocks are not reclaimed.
More information about block size impact can be found at this article. Sorry for not letting you know before.

In addition, as the practice we’ve described can’t be implemented in your particular case, you should probably rely on VMware KB article cited in your latest post and follow necessary steps.

Thanks.

[Vitaliy S.]

Hi Mephisto,

The proposed approach was described by one of our community members, and it does seem to work for him:
Veeam v6 and sdelete/defrag optimisations for CBT.

Feel free to follow the KB article you've referenced to reclaim free space.

Thanks!

[mephisto]

My answer regarding vmkfstools –k was completely based on this VMware KB.

And yes, It seems to be a known issue that if are using Storage vMotion to thin disk between two VMFS volumes with the same block size, null blocks are not reclaimed.
More information about block size impact can be found at this article. Sorry for not letting you know before.

In addition, as the practice we’ve described can’t be implemented in your particular case, you should probably rely on VMware KB article cited in your latest post and follow necessary steps.

Thanks.

Well, the vmkfstools –k command didn't do the trick, I inflated the VMDK with sdelete, but the vmkfstools –k didn't shrink it.

I'm using VMware converter now as a last resort, but this is a very time consuming and lame option in my opinion :/

[mephisto]

Hi Mephisto,

The proposed approach was described by one of our community members, and it does seem to work for him:
Veeam v6 and sdelete/defrag optimisations for CBT.

Feel free to follow the KB article you've referenced to reclaim free space.

Thanks!

I understand the limitations of CBT in this case, not a problem with it. So even if I sdelete the VMDKs, as I have all 0's written on the "free space" should Veeam backup job be blazing fast went it read these blocks and do an absurd compression?

[mephisto]

Well, I used VMware converter on the VM, from 477GB went down to 376GB, all disks are thin provisioned.

Looking inside the guest, 142GB across all partitions. How the hell I keep ending up with the VM with this size? My backups are being veyr large, like 200GB+ for a VM that actually has 142GB.

How can I work around this? I simply can't afford this situation, several VM's are having the same problem and my backup storage is maxed out

[mephisto]

Question, if I disable CBT, am I going to achieve better compression as it will not process the whole 0's blocks?

[Vitaliy S.]

Question, if I disable CBT, am I going to achieve better compression as it will not process the whole 0's blocks?

CBT does not have any affect on compression ratio.

[mephisto]

Well I've tried a few other things and got this resolved:

sdelete -z instead of -c

vmotion to datastore using different VFS version like VFS5 to VFS3, then the machine shrinked

[veremin]

The parameter which is used in SDelete utility for zeroing free space depends on SDelete version.

In Secure Delete v1.51:

Code: Select all

-c Zero free space

In Secure Delete v1.6 (and newer):

Code: Select all

-z Zero free space
-c Clean free space

Anyway, thank you for your feedback. I’ve already updated previous answers in order to avoid further misunderstanding.

[mikegodwin]

Can two jobs backup the same VM at the same time?

[guillermo.lozano]

Hi Community,

Is it possible/recommended to replicate and backup a VMv at the same time

For instance, I want to replicate constantly a VM and back it up every 15 -20 minutes.

Thanks again.
Best Regards,
Guillermo.-

[guillermo.lozano]

In my particular case i have a source proxy with VB&R doing the backups and a destination VB&R doing the replications.

Thank you.
Guillermo.-

[foggy]

Then please review another existing thread discussing the same setup.

[guillermo.lozano]

Hi Foggy, thank you for the information.

Regards,
Guillermo.-