Veeam Backup & Replication Configuration: Turnkey Server Backup

What Are Common Problems Solved by Veeam Configuration?

Changed Block Tracking (CBT) can become corrupted after an improper hypervisor shutdown. The result: incremental backups turn into full backups, and RPO jumps from 15 minutes to hours. As noted in Veeam's official documentation, the RPO can be as low as 15 minutes with proper configuration. We have been configuring Veeam turnkey for years — over 50 projects, guaranteeing RTO under 1 hour and RPO from 15 minutes. Without correct setup, deduplication and compression won't deliver expected savings — up to 70% disk space. For a 50-TB environment, this translates to annual savings of $15,000 compared to traditional storage. Our clients typically achieve a 300% return on investment within 6 months, reducing annual backup storage costs by $10,000 for a typical 50-VM infrastructure.

Application Inconsistency occurs when databases are backed up in the middle of a transaction due to missing VSS integration. Veeam solves this through the VSS Framework: before taking a snapshot, Veeam calls VSS writers for SQL Server, Exchange, SharePoint, which flush buffers and bring the application to a consistent state. For Linux with PostgreSQL or MySQL, Veeam uses pre/post-freeze scripts:

# /etc/veeam/pre-freeze-script.sh
#!/bin/bash
sudo -u postgres psql -c "SELECT pg_start_backup('veeam', true);"
# /etc/veeam/post-thaw-script.sh
#!/bin/bash
sudo -u postgres psql -c "SELECT pg_stop_backup();"

Step-by-step Application-Aware Processing setup:

1. Enable the 'Application-Aware Processing' flag in the backup job.
2. Provide guest OS credentials (local administrator).
3. Select applications for VSS — SQL Server, Exchange, SharePoint.
4. For Linux, configure pre/post-freeze scripts as shown above.

Proper Veeam configuration ensures optimal Veeam Backup & Replication performance. For Veeam VMware environments, the proxy should be placed in HotAdd mode. For Veeam Hyper-V, use a dedicated proxy. Application-Aware Processing is crucial for consistent backups of SQL Server.

Deployment Architecture and Scale-Out Backup Repository

Veeam Backup & Replication Server (Windows)
 |
 |--- Veeam Proxy (processes data from hypervisor)
 |    |--- Proxy-01 (VMware vSphere)
 |    |--- Proxy-02 (Hyper-V)
 |
 |--- Backup Repository (stores backups)
 |    |--- Scale-Out Repository -> object storage (S3)
 |    |--- Local repository (Windows/Linux)
 |
 |--- Enterprise Manager (centralized management, optional)

The proxy is the node that actually reads data from the hypervisor. For large installations, multiple proxies reduce load on production hosts. Proxies can be placed directly on ESXi hosts in HotAdd mode — the fastest option, reading data locally without network.

SOBR automatically offloads old backups from local repository to object storage. This is 50% cheaper than keeping all copies on a local repository with RAID10. Configuration via PowerShell:

# Add S3-compatible storage
Add-VBRObjectStorageRepository `
    -Name "S3-Offload" `
    -Type "AmazonS3Compatible" `
    -Server "s3.eu-central-1.amazonaws.com" `
    -Bucket "veeam-backups-prod" `
    -AccessKey "AKIA..." `
    -SecretKey "secret" `
    -EnableSizeLimit

# Offload policy: move backups older than 7 days to S3
Set-VBRScaleOutBackupRepository `
    -ScaleOutBackupRepository (Get-VBRScaleOutBackupRepository "SOBR-Prod") `
    -PolicyType MoveByAge `
    -MoveOffloadedBackupAgeThreshold 7

What Are the Repository Types and Their Costs?

Object storage (S3, Azure Blob, Yandex Object Storage) allows paying only for actual space used, without pre-provisioning disk arrays. The cost of storing one terabyte in the cloud is 60-80% lower than on a local RAID10. Additionally, Veeam supports lifecycle policies, automatically transitioning old backups to cold storage, saving even more budget.

Configuring Backup Job, Schedule, and Retention Policies

Via PowerShell (useful for automation):

$repo = Get-VBRBackupRepository -Name "SOBR-Prod"
$vms = Find-VBRViEntity -Name "prod-*" -VMsAndTemplates
Add-VBRViBackupJob `
    -Name "Prod-VMs-Daily" `
    -Entity $vms `
    -BackupRepository $repo `
    -GuestProcessingEnabled `
    -ApplicationAwareEnabled `
    -EnableDeduplication `
    -CompressionLevel Optimal
$schedule = New-VBRJobScheduleOptions
$schedule.OptionsDaily.Enabled = $true
$schedule.OptionsDaily.TimeLocal = "22:00:00"
$schedule.OptionsDaily.DaysSrv = @("Sunday","Monday","Tuesday","Wednesday","Thursday","Friday","Saturday")
Set-VBRJobSchedule -Job (Get-VBRJob "Prod-VMs-Daily") -Options $schedule

Veeam supports GFS (Grandfather-Father-Son) natively in job settings:

• Simple retention: keep N restore points.
• GFS: weekly (4 copies), monthly (12 copies), yearly (up to 7 years) — for compliance.

For yearly backups with 7+ year retention, we use Immutable Storage via S3 Object Lock:

# Enable immutability for S3 repository
Set-VBRObjectStorageRepository `
    -Repository (Get-VBRObjectStorageRepository "S3-Offload") `
    -EnableImmutability `
    -ImmutabilityPeriod 30

How to Monitor Veeam Backup Infrastructure?

Veeam ONE is a separate product for monitoring. Out of the box it provides:

• alerts for missed jobs, repository overfill, high RPO;
• reports on Backup Success Rate, Compliance, Protected VMs;
• integration with SCOM, vCenter Operations.

An alternative is the Veeam Exporter for Prometheus (official GitHub repository). Typical problems and their solutions:

• CBT corrupted: Changed Block Tracking may become corrupt after an improper hypervisor shutdown. The job starts doing a full backup instead of incremental. Solution: Reset CBT in VM settings or via PowerCLI.
• Unable to truncate transaction logs for SQL Server: Application-Aware Processing cannot log into the guest OS. Check: user with Local Admin rights, WMI accessible, port 445 open between proxy and VM.
• Repository full, job won't start: Veeam reserves 10% space by default. At 90% fill, jobs stop. Solution: extend repository, enable SOBR offload to S3, or increase the threshold in settings.

What Is Included in the Work (Deliverables)

• Audit of existing vSphere/Hyper-V infrastructure, networks, and storage systems.
• Architecture design: selection of proxies, repositories, policies.
• Installation and configuration of Veeam Backup & Replication Server, proxies, repositories, Enterprise Manager.
• Configuration of Application-Aware Processing for SQL, Exchange, SharePoint, PostgreSQL, MySQL.
• Setup of SOBR with S3 offload, including Immutable Storage if needed.
• Creation of backup jobs with schedule and GFS retention.
• Enablement of monitoring via Veeam ONE or Prometheus/Grafana.
• Testing of recovery: Instant VM Recovery, file-level restore, application item restore.
• Documentation and runbooks: architecture diagrams, admin instructions, handover of keys and access, staff training.
• Support for the first month included.

Cost is calculated individually after an audit of your vSphere/Hyper-V infrastructure. With deduplication and compression, you save up to 70% disk space. ROI reaches 300% thanks to reduced downtime. Average project cost for 50 VMs starts from $2,500, with a payback period of 6 months.

Contact us — we will conduct an audit and propose the optimal solution. Request a consultation right now.