Design solutions for backup and disaster recovery

Pellham Media runs a video-rendering farm of 30 Azure virtual machines in the East US 2 region that use Premium SSD v2 data disks. The company requires an RPO of 4 hours for the VMs and requires that the backup data remain available if a single availability zone in East US 2 fails. Cross-region recovery is not required and cost must be minimized.

What should you recommend?

1. AA Standard backup policy with a daily backup in a Recovery Services vault that uses locally redundant storage.A Standard policy backs up once per day (RPO 24 hours), does not support Premium SSD v2 disks, and locally redundant vault storage does not survive a zone failure.
2. BAn Enhanced backup policy with a backup every 4 hours in a Recovery Services vault that uses zone-redundant storage.The Enhanced policy supports Premium SSD v2 disks and backups as frequent as every 4 hours, and zone-redundant vault storage keeps the backup data available during a zone outage at lower cost than geo-redundant storage.
3. CAn Enhanced backup policy with a backup every 12 hours in a Recovery Services vault that uses zone-redundant storage.A 12-hour backup interval gives an RPO of 12 hours, which exceeds the 4-hour requirement.
4. DAn Enhanced backup policy with a backup every 4 hours in a Recovery Services vault that uses locally redundant storage.Locally redundant storage keeps three copies in one zone, so the backup data would be unavailable during a zone failure.

Corvina Freight hosts 12 databases on an Azure SQL Managed Instance named mi-ops. The company requires that any database can be restored to any point in time within the last 35 days and that one backup per year of each database is retained for 10 years. The solution must not depend on SQL Server Agent jobs or scripts that the DBA team has to maintain.

What should you recommend?

1. ASet the point-in-time restore retention of each database to 35 days and configure a long-term retention policy with yearly backups kept for 10 years.SQL Managed Instance supports PITR retention of 1 to 35 days and a built-in LTR policy that retains weekly, monthly or yearly backups for up to 10 years, all managed by the platform.
2. BSet the point-in-time restore retention of each database to 35 days.PITR alone covers the 35-day window but provides nothing beyond it, so the 10-year yearly backup requirement is not met.
3. CCreate a SQL Server Agent job on mi-ops that takes a COPY_ONLY backup of each database to Azure Blob Storage once a year.Native backups to URL do satisfy retention, but the scenario explicitly rules out a solution that depends on Agent jobs maintained by the DBAs.
4. DRegister mi-ops with a Recovery Services vault and use the SQL Server in Azure VM backup workload with yearly retention.The SQL Server in Azure VM backup workload is for SQL Server running on IaaS virtual machines; it cannot be used with a managed instance.

Marlowe Studios protects an Azure virtual machine named VM-EDIT01 with a daily Azure VM backup in a Recovery Services vault. An administrator accidentally deleted VM-EDIT01, including its disks. The studio requires that the VM is running again within 1 hour, in the same region and virtual network, from the most recent recovery point, with the least administrative effort.

What should you recommend?

1. AUse the Create new virtual machine restore option from the latest recovery point in the vault.The Create new VM restore type rebuilds the VM with its disks, network settings and configuration directly from the recovery point, which is the fastest path and needs no manual assembly.
2. BUse the Restore disks option, then manually create a new VM from the restored disks.Restoring disks only works, but it adds manual steps to create and configure the VM, so it is more effort than the Create new VM option.
3. CUse Cross Region Restore to recover the VM in the paired region.Cross Region Restore targets the paired region, which violates the requirement to recover in the same region and network.
4. DUse File Recovery to mount the recovery point and copy the files to a new VM.File Recovery mounts a recovery point as a drive for item-level recovery; it does not rebuild an entire VM.

Quarry Lane Games stores player profiles in an Azure Cosmos DB for NoSQL account named cdb-players with a single write region. A recent incident deleted a container by mistake. The company requires that the operations team can restore the account to any second within the last 30 days without opening a support ticket, and that the restore can target a new account in the same region.

What should you recommend?

1. AEnable continuous backup with the 30-day retention tier on cdb-players.Continuous backup lets the customer self-service a point-in-time restore to any timestamp within the retention window (7 or 30 days) into a new account.
2. BKeep periodic backup with a backup interval of 1 hour and a retention of 720 hours.Periodic backups are taken at fixed intervals, so the granularity is one hour rather than any second, and a restore requires a support request to Microsoft.
3. CEnable continuous backup with the 7-day retention tier on cdb-players.The 7-day tier provides self-service point-in-time restore but cannot reach back 30 days.
4. DUse the change feed to copy every write into a second Cosmos DB account.A change-feed mirror replays deletes as well as writes and does not provide restore to an arbitrary earlier point in time.