HPE StoreOnce Active Full vs Synthetic full

[LickABrick]

Hi, I have a question about HPE StoreOnce.

The Veeam BP site states:
It’s been tested on multiple, different environments that Active Fulls perform faster than Synthetic Fulls, regardless of the Catalyst bandwidth modes. The difference has been registered for up to double the performance.
(https://bp.veeam.com/vbr/2_Design_Struc ... eonce.html)

While the HPE StoreOnce Veeam Best practices states the following:
A synthetic full backup generally completes in 50% or less of the time required for a traditional active full backup.
(https://www.hpe.com/psnow/doc/a00023056 ... -psnow-red page 31)

Which one is correct? I would assume the synthetic fulls would be the best option since everything would be done on the StoreOnce without the need to transfer any data between the VMware datastores and the repository. As also explained in the HPE best practices:
HPE StoreOnce executes this task at high speed because it does not need to move the actual data, but only the pointers to the stored deduplicated data segments.

[foggy]

Hi, I will check the origin of the statement in our best practices. Meanwhile, you can test what is faster in your particular environment since there are a lot of variables that can affect the performance. Storage I/O wise, virtual synthetics should be faster, indeed.

[david.domask]

One thing to add is that in v10, we did a pretty nice upgrade to the engine that handles the Virtual Full process (for tape), but in fact the Virtual full process for the Storeonce is the same, and thus receives the same benefit. I had quite a few cases where people were pleasantly surprised by the performance increase (sometimes as much as a 3x reduction in Synthetic Full creation time)

Nothing beats a real world test with your Storeonce and environment, but if you can find happy results with the Virtual Full, just to take pressure off the production environment, I'd say go with the Virtual Full, as it's better to stress the Storeonce than to stress production. (In most cases )

Edit: To clarify the time reduction, keep in mind this is very environmentally specific, so your results may vary.

[foggy]

We will update the statement in our best practices since it looks like it was based on some particular cases while in general it might not be the case.