Back up SQL Server on Azure Stack using Azure Backup

2025-04-12

This article describes how to configure Microsoft Azure Backup Server (MABS) to protect SQL Server databases on Azure Stack.

SQL Server databases protection workflow

Management of the SQL Server database backup to Azure and recovery from Azure involves:

Create a backup policy to protect SQL Server databases
Create on-demand backup copies
Recover the database from Disks, and from Azure

Cenários e limitações suportados

Before you back up SQL Server on Azure Stack, review the following supported scenarios and limitations:

Se você tiver um banco de dados com arquivos em um compartilhamento de arquivos remoto, a proteção falhará com a ID de erro 104. O MABS não oferece suporte à proteção de dados do SQL Server em um compartilhamento de arquivos remoto.
O MABS não pode proteger bancos de dados armazenados em compartilhamentos SMB remotos.
Certifique-se de que as réplicas do grupo de disponibilidade estejam configuradas como somente leitura.
Você deve adicionar explicitamente a conta do sistema NTAuthority\System ao grupo Sysadmin no SQL Server.
Ao executar uma recuperação de local alternativo para um banco de dados parcialmente contido, você deve garantir que a instância SQL de destino tenha o recurso Bancos de Dados Contidos habilitado.
Ao executar uma recuperação de local alternativo para um banco de dados de fluxo de arquivos, você deve garantir que a instância SQL de destino tenha o recurso de banco de dados de fluxo de arquivos habilitado.
Proteção para SQL Server Always On:
- O MABS deteta Grupos de Disponibilidade ao efetuar uma consulta durante a criação de um grupo de proteção.
- O MABS deteta um failover e continua a proteção do banco de dados.
- O MABS oferece suporte a configurações de cluster multissite para uma instância do SQL Server.
Quando você protege bancos de dados que usam o recurso Always On, o MABS tem as seguintes limitações:
- O MABS honrará a política de backup para grupos de disponibilidade definida no SQL Server com base nas preferências de backup, da seguinte maneira:
  - Preferir secundário - Os backups devem ocorrer em uma réplica secundária, exceto quando a réplica primária for a única réplica online. If there are multiple secondary replicas available, then the node with the highest backup priority will be selected for backup. If only the primary replica is available, then the backup should occur on the primary replica.
  - Apenas secundário - O backup não deve ser executado na réplica primária. Se a réplica principal for a única que está online, o backup não deve ocorrer.
  - Principal - Os backups devem sempre ocorrer na réplica primária.
  - Any Replica - Backups can happen on any of the availability replicas in the availability group. O nó a partir do qual será feito o backup será determinado com base nas prioridades de backup de cada um dos nós.
- Observação
  - Os backups podem ocorrer a partir de qualquer réplica legível, ou seja, primária, secundária síncrona, secundária assíncrona.
  - Se qualquer réplica for excluída do backup, por exemplo , Excluir réplica estiver habilitada ou marcada como não legível, essa réplica não será selecionada para backup em nenhuma das opções.
  - If multiple replicas are available and readable, then the node with the highest backup priority will be selected for backup.
  - Se o backup falhar no nó selecionado, a operação de backup falhará.
  - A recuperação para o local original não é suportada.
Problemas de backup do SQL Server 2014 ou superior:
- SQL server 2014 added a new feature to create a database for on-premises SQL Server on Microsoft Azure Blob storage. O MABS não pode ser usado para proteger essa configuração.
- There are some known issues with "Prefer secondary" backup preference for the SQL Always On option. MABS always takes a backup from secondary. Se nenhum secundário puder ser encontrado, o backup falhará.

Pré-requisitos

Before you back up SQL Server on Azure Stack, install and prepare Azure Backup Server.

Criar uma política de cópias de segurança

To create a backup policy to protect SQL Server databases to Azure, follow these steps:

On the Azure Backup Server, select the Protection workspace.
On the tool menu, select New to create a new protection group.

Azure Backup Server starts the Protection Group wizard, which leads you through creating a Protection Group. Selecione Seguinte.
On the Select Protection Group Type blade, select Servers.
On the Select Group Members blade, the Available members list displays the various data sources. Select + to expand a folder and reveal the subfolders. Select the checkbox to select an item.

All selected items appear in the Selected members list. After selecting the servers or databases you want to protect, select Next.
On the Select Data Protection Method blade, provide a name for the protection group and select the I want online Protection checkbox.
On the Specify Short-Term Goals blade, include the necessary inputs to create backup points to disk, and select Next.

In the example, Retention range is 5 days, Synchronization frequency is once every 15 minutes, which is the backup frequency. Express Full Backup is set to 8:00 P.M.

Observação

In the example shown, at 8:00 PM every day a backup point is created by transferring the modified data from the previous day’s 8:00 PM backup point. Este processo é chamado Express Full Backup. Transaction logs are synchronized every 15 minutes. If you need to recover the database at 9:00 PM, the point is created from the logs from the last express full backup point (8PM in this case).
On the Review disk allocation blade, verify the overall storage space available, and the potential disk space. Selecione Seguinte.
On the Choose Replica Creation Method, choose how to create your first recovery point. You can transfer the initial backup manually (off network) to avoid bandwidth congestion or over the network. If you choose to wait to transfer the first backup, you can specify the time for the initial transfer. Selecione Seguinte.

The initial backup copy requires transferring the entire data source (SQL Server database) from production server (SQL Server computer) to Azure Backup Server. This data might be large, and transferring the data over the network could exceed bandwidth. For this reason, you can choose to transfer the initial backup: Manually (using removable media) to avoid bandwidth congestion, or Automatically over the network (at a specified time).

Once the initial backup is complete, the rest of the backups are incremental backups on the initial backup copy. Os backups incrementais tendem a ser pequenos e são facilmente transferidos pela rede.
Choose when you want the consistency check to run and select Next.

Azure Backup Server performs a consistency check on the integrity of the backup point. Azure Backup Server calculates the checksum of the backup file on the production server (SQL Server computer in this scenario) and the backed-up data for that file. If there's a conflict, it's assumed the backed-up file on Azure Backup Server is corrupt. Azure Backup Server rectifies the backed-up data by sending the blocks corresponding to the checksum mismatch. Because consistency checks are performance-intensive, you can schedule the consistency check or run it automatically.
To specify online protection of the datasources, select the databases to be protected to Azure and select Next.
Choose backup schedules and retention policies that suit the organization policies.

In this example, backups are taken once a day at 12:00 PM and 8 PM.

Observação

It’s a good practice to have a few short-term recovery points on disk, for quick recovery. Esses pontos de recuperação são usados para recuperação operacional. Azure serves as a good offsite location with higher SLAs and guaranteed availability.

Best Practice: If you schedule backups to Azure to start after the local disk backups complete, the latest disk backups are always copied to Azure.
Choose the retention policy schedule. The details on how the retention policy works are provided at Use Azure Backup to replace your tape infrastructure article.

Neste exemplo:
- Backups are taken once a day at 12:00 PM and 8 PM and are retained for 180 days.
- The backup on Saturday at 12:00 P.M. is retained for 104 weeks
- The backup on Last Saturday at 12:00 P.M. is retained for 60 months
- The backup on Last Saturday of March at 12:00 P.M. is retained for 10 years
Select Next and select the appropriate option for transferring the initial backup copy to Azure. You can choose Automatically over the network
Once you review the policy details in the Summary blade, select Create group to complete the workflow. You can select Close and monitor the job progress in Monitoring workspace.

Executar uma cópia de segurança a pedido

A recovery point is created only when the first backup occurs. After creating a backup policy, you can trigger the creation of a recovery point manually, rather than waiting for the scheduler to take the backup.

To run an on-demand backup of a SQL Server database, follow these steps:

Wait until the protection group status shows OK for the database before creating the recovery point.
Right-click the database and select Create Recovery Point.
Choose Online Protection in the drop-down menu and select OK to start creation of a recovery point in Azure.
View the job progress in the Monitoring workspace.

Recover the database from Azure

To recover a protected entity (SQL Server database) from Azure, follow these steps:

Open the Azure Backup Server Management Console. Navigate to Recovery workspace where you can see the protected servers. Browse the required database (in this case ReportServer$MSDPM2012). Select a Recovery from time that's specified as an Online point.
Clique com o botão direito do mouse no nome do banco de dados e selecione Recuperar.
MABS shows the details of the recovery point. Selecione Seguinte. Para substituir o banco de dados, selecione o tipo de recuperação Recuperar para a instância original do SQL Server. Selecione Seguinte.

In this example, MABS recovers the database to another SQL Server instance, or to a standalone network folder.
On the Specify Recovery options blade, you can select the recovery options like Network bandwidth usage throttling to throttle the bandwidth used by recovery. Selecione Seguinte.
On the Summary blade, you see all the recovery configurations provided so far. Selecione Recuperar.

The Recovery status shows the database being recovered. You can select Close to close the wizard and view the progress in the Monitoring workspace.

Once the recovery is completed, the restored database is application consistent.

Partilhar via