Replica failback size

[matteu]

Hello,

I test a DR scenario with a customer these last days.
Site1 : 1 proxy (with Wan accelerator)
DR Site : 1 proxy (with Wan accelerator)

1 replica job with 3 VM
On the replica job since 1 month : Arround 1GB data transfered / days

We launch failover from Site1 to DR site and it was ok.
1 day after, we launch failback, I select manually proxy (Veeam doesn't picked the good ones) and I check quick rollback because it was compatible with test scenario.
However, Data transfered are here are more than 10GB and time necessary is not anymore 1hour but more than 7 hours.

I know the time is longer because transfer is over WAN and bandwith is low. What I would like to understand is : Why it's so much bigger data to transfert on failback than on replica everyday.
It's not performance issue in this case...

Thanks for your help

[HannesK]

Hello,
Veeam transfers the data, that VMware change block tracking gives back. If you like to know which blocks changed and then try to find out why these blocks changed, I guess VMware is a good starting point.

I know from my environment, that a Windows reboot (or boot in case of failover) causes some GB to change. So maybe the question could also be asked to Microsoft

Best regards,
Hannes

[matteu]

And the VM will automatically reboot because when you poweron replica....
That means failback will always be some GB right ?

[HannesK]

at least for Windows, I see no other way

[matteu]

Thank you for the answer