This is a general question on if there are any constraints or recommendations on how long you can leave a VPG in a failover state before clicking commit with reverse protection ?
Viola from Zerto here.
Thanks for your reply.. what are the scratch volumes and if large enough can we sustain 2-3 hours before committing the VPG ?
Guy from Zerto support here.
A scratch volume is a temporary disk that is attached to the Recovery site VRA during a failover operation.
The idea behind the scratch volume is that all IOs generated during the failover will be written to it.
As Viola stated in her response, it is not recommended to leave a VPG in a pre-commit state for too long because IOs will be constantly written to the scratch volume, which could potentially negatively affect replication.
Please note that 2-3 hours is a reasonable timeframe to wait before committing a failover operation, as it is lower than the maximal delay of 24 hours.