Restore changes disk Format from Thick Eager to Thick Lazy

[rurouni]

Hi,

I've been experiencing a very annoying problem after performing lots of vm disks restore.

While choosing "VM Hard Disk" Restore or VM Files Restore, no matter how i set the setting "Restore disks" (As on Original VM or Force Thick) at the "Disk Mapping" Step, the disks is never restored as THICK EAGER ZEROED.
At the end it's always "Thick Lazy Zeroed".

I'm not satisfied with this behaviour since the only workaround i have is to perform a storage vmotion with vsphere client to change the disk format settings to Thick Eager Zeroed.

I already have opened a case which proves there is a BUG in the restoration process.

CASE ID was: # 00519033

We're using veeam backup v7.0.0.771.

I really need it fixed since Lazy Zero is not the recommendation from my storage vendor. (HP 3PAR Storeserv)

Thank you for the help.

[foggy]

Thanks for bringing this to our attention. As I can see from the case, this was already passed to R&D, so most likely will be addressed in one of the following updates.

[dellock6]

Hi,
aside the request, that is totally legit, I'm interested in this reccomendation. 3Par technology removes all the zeros from any disk, so what is the point to forcefully write them if then the 3Par removes them on the storage? Sounds to me like a waste of I/O operations...

Luca.

[rurouni]

From what i ve been told from HP consultant, the fact that volume is Zero initialized at creation or upgrade, means there won't be any need to do it on the fly upon Guest File System Growth. So it would save some IO at the end as they are initialized at creation or extension.
Do you agree with that?

On the other hand i also observed VMware themselves didn't support Eager Zeroing in every tool they provide. Converter for example do not allow to initialize a volume with Eager Zeroes. It only offers two options Thin, Or Thick (which is Lazy Zero here) as Veeam does upon VM file restore option.
The Last thing we were surprised to see is that Extending a Thick Zeroed Volume converts it to Thin Zeroed Volume. Vmware said that by extending the guest os disk, the extend would be in all cases Thin Zeroed initialized. That s the way it works. If you need to do have the Disk extend to be Zeroed, you have to use vmkfstool command to do it.

[rurouni]

By the way,
I found the HP/Vmware Best Practice documents which states that for IO intensive environments, you'd better use Eager Zero Provisionning on Vmware Side as HP 3Par Handles the Zero Block Writing flawlessly:

When creating VMs, there are a number of options that are available for the VMDK files. VMware vSphere creates VMs using
the “Lazy Zeroed Thick” option by default. With this option, when a new VM is created, the full size of the VMDK is not
immediately zeroed. Instead, zeros are returned upon reads from unwritten areas, but not actually backed by physical
storage until actual write operations. For performance-intensive environments and security concerns, VMware recommends
using “Eager Zeroed Thick” (EZT) virtual disks. EZT disks have the smallest overhead but require zeros to be written across
all of the capacity of the VMDK at the time of creation. Unlike many other storage vendors, HP 3PAR Thin Persistence
Software and HP 3PAR Zero Detect enabled virtual volumes allow clients to retain the thin provisioning benefits when using
Eager Zeroed Thick VMDKs without sacrificing any of the performance benefits offered by this VMDK option. Please see the
Virtual SCSI adapters and virtual disk types section of this document for a comparison of the available disk types with
VMware vSphere 5.

http://www.vmware.com/files/pdf/partner ... erv-so.pdf