To meet the specified requirements, configuring a backup copy job in periodic copy (pruning) mode with 62 days of retention and six monthly Grandfather-Father-Son (GFS) restore points is the best approach. This setup allows for daily backups for up to two months and weekly backups for up to six months, aligning with the desired restore capabilities.References:Veeam Backup & Replication User Guide, Veeam GFS Retention Policy Explanation

When backing up NAS file shares, it's essential to ensure that more than one copy of the data exists to adhere to best practices for data protection, often referred to as the 3-2-1 Rule (3 total copies of your data, 2 of which are local but on different devices, and at least 1 copy offsite). Implementing a secondary copy of the data provides this additional layer of redundancy. In Veeam, this can be achieved by creating backup copy jobs, which ensure that there is a secondary copy of the data, possibly stored in a different location. This is not to be confused with the simple use of backup repositories, which are storage locations for primary backups, or with cache/archive repositories which serve different purposes in the backup process.

For the given scenario where daily backups need to be sent from Site B to a repository at Site A and the connection between the two sites is limited, thus impacting the ability to complete direct backups within the backup window, a Backup Copy Job would be appropriate. A Backup Copy Job is a feature in Veeam Backup & Replication that allows you to create several instances of the same backup files across different locations (repositories). In the image provided, the components involved in the Backup Copy Job from Site B to Site A would be:
6 (Repository at Site B): This is the source repository where the hourly backup jobs are stored.
4 and 3 (WAN Accelerators at both sites): These components optimize data transfer over the WAN.
2 (Repository at Site A): This is the target repository where the backup copies will be stored. This setup would minimize impact on the source VMs at Site B since the Backup Copy Job works with backup data rather than directly with the production VMs, thus reducing the load on those VMs during the process.