Azure Solutions Architecture — Business Continuity: Backup, Site Recovery & HA Design

Azure Solutions Architecture · AZ-305 · Networking & Business Continuity · by Raushan Ranjan, MCT

A solid business continuity and disaster recovery (BCDR) strategy is essential for protecting your data and applications from unexpected disruptions. Azure provides two key services for this: Azure Backup and Azure Site Recovery.

Design for Backup and Recovery

Azure Backup is a cloud-native service that provides a simple, secure, and cost-effective solution for backing up and restoring data. It centralizes backup management for various workloads, both on-premises and in Azure.

Azure Blob Backup and Recovery

• Operational Backup: For short-term retention and fast recovery, Azure Backup provides an operational backup solution for Block Blobs. This uses built-in Blob Storage features like point-in-time restore, soft delete, and versioning to enable quick recovery of data within the same storage account. This is ideal for recovering from accidental deletions or data corruption.
• Vaulted Backup: For long-term retention, you can use a vaulted backup. This transfers data from the storage account to a Recovery Services vault, providing a separate, isolated copy of the data that can be retained for up to 10 years. This protects against a full storage account or regional outage.

Azure Files Backup and Recovery

• Azure Backup integrates directly with Azure Files to provide a managed backup solution.
• It uses file share snapshots to create point-in-time recovery points. The snapshots are stored locally with the file share for fast, operational recovery.
• For long-term retention, a vaulted backup can be configured to transfer the snapshot data to a Recovery Services vault.

Azure Virtual Machine Backup and Recovery

• Azure Backup is the primary service for protecting Azure VMs. You can configure a backup policy that runs on a set schedule.
• When a backup runs, an extension on the VM coordinates with the guest OS (via VSS for Windows or fsfreeze for Linux) to take an application-consistent snapshot of the VM's disks.
• The snapshot data is then transferred incrementally to the Recovery Services vault, ensuring only changed blocks are sent. This process is optimized for efficiency and minimal impact on the VM's performance.

Azure SQL Backup and Recovery

• Automated Backups: Azure SQL Database provides built-in, automated backups. It performs a full backup weekly, differential backups every 12-24 hours, and transaction log backups every 5-10 minutes. These backups are stored in geo-redundant storage.
• Point-in-Time Restore (PITR): You can restore an Azure SQL database to any point in time within the defined retention period (up to 35 days). This is the standard method for recovering from accidental data corruption.
• Long-Term Retention (LTR): For longer retention periods, you can use LTR policies to keep backups for up to 10 years in Azure Blob Storage. This is useful for compliance and auditing requirements.

Design for Azure Site Recovery

Azure Site Recovery (ASR) is a disaster recovery as a service (DRaaS) solution that ensures business continuity by replicating workloads from a primary location to a secondary one. This is about minimizing downtime, not just data loss.

• How it Works: ASR continuously replicates your VMs and physical servers from a primary site (on-premises or another Azure region) to a secondary Azure region.
• Orchestration: A Recovery Plan is a key feature of ASR. It is a defined sequence of steps to automate the failover process. For a multi-tier application, a Recovery Plan can ensure that the database server fails over first, followed by the web servers, and so on, to ensure the application starts in the correct order.
• Failover and Failback: In the event of a disaster at the primary site, you can perform an unplanned failover to the secondary region. This brings up the replicated VMs with minimal downtime. Once the primary site is restored, you can fail back to return to normal operations. You can also perform a test failover at any time without disrupting your production environment.
• Benefits: ASR provides very low Recovery Time Objective (RTO) and Recovery Point Objective (RPO). It can provide an RPO as low as 30 seconds and an RTO of minutes for some workloads, making it a powerful tool for business continuity.