This episode focuses on backing up mission-critical applications: application-aware processing with VSS for SQL, Exchange, AD, and Oracle consistency, Application Groups (SQL + AD + file server), and agent jobs for databases via Veeam Agent and plugins. You will understand the difference between crash-consistent and application-consistent backups.

In episode 5 we touched on application-aware processing; in episode 8 we touched on Explorers for application item restores. In episode 10 we combine both thoroughly: database application backups — SQL Server, Exchange, Active Directory, and Oracle — from the perspective of consistency, Application Group design, and agent jobs.
Databases are the heart of business, and backing them up wrong means inviting disaster. This episode builds a proper understanding of what makes a database backup restorable correctly.
As covered in episode 5, a hypervisor snapshot produces a crash-consistent backup — the disk files are intact but the application is not. For databases this is dangerous: uncommitted transactions can corrupt the database during recovery. Application-aware processing (VSS) solves this problem.
ALTER TABLESPACE ... BEGIN BACKUP for Oracle) and post-thaw scripts so the database is consistent without VSS.Application-aware processing requires valid guest credentials. For SQL and Exchange, Veeam also needs access to the application (e.g. the SQL instance) to perform log truncation and restore-aware logging. Configure it in the job wizard: Guest Processing → Application-aware processing → Edit → enable per application.
Important
The crash-consistent vs application-consistent difference determines whether a database can be restored. A crash-consistent backup of a SQL Server might be recovered by SQL recovery, but it can also fail or lose the last transactions — and you will not know until you try. Always use application-aware processing for production databases.
An Application Group is a set of VMs that are backed up and tested as a single ordered entity — aware of dependencies between applications. A classic example: a file server holds a database attach from SQL, and both are used by AD. An Application Group ensures:
From the console: Backup Infrastructure → Application Groups → Add. Select the VMs, then give each VM test scripts (e.g. ping, check port 1433 for SQL). This group then serves as the basis for a SureBackup job — full verification that the entire application stack can come to life from backup.
For physical servers or cases where agent-level is more appropriate, Veeam Agent for Windows supports application plugins:
This pattern gives stricter RPO (transaction log backups can run frequently) without touching the hypervisor.
Veeam uses SQL log backups for point-in-time restores (PITR) in the restore step: after restoring the full backup, roll forward the transaction logs to the desired minute. To take advantage of this, enable Log backup in the SQL job:
Full backup ──▶ Log1 ──▶ Log2 ──▶ Log3 ──▶ TARGET TIMEOnce application-consistent backups are available, granular restores use the Explorers:
A quick flow via PowerShell:
$rp = Get-VBRRestorePoint -Name "SQL-01" | Select-Object -First 1
Start-VBRSQLRestore -RestorePoint $rpThen select the needed items in the Explorer GUI.
Tip
Test application restores at least once a month. Backing up SQL with VSS is useless if you have never proven that Explorer for SQL actually returns a database to an instance. Make "application restore drills" part of your monthly SOP — episode 20 connects them to reports and monitoring.
After this episode, make sure:
Key takeaways:
In the next episode, episode 11, we will cover M365 & Salesforce backups — Veeam Backup for Microsoft 365 (Exchange Online, SharePoint, OneDrive, Teams) with per-user licensing, and Veeam Backup for Salesforce for metadata & data, whose platform scope expanded in v13. Your SaaS data will finally be protected!