Proxmox Hybrid Cloud Backup and Disaster Recovery (Proxmox + PBS + Cloud DR)

🔰 Introduction

In enterprise IT infrastructure, availability and data durability are two of the most critical pillars.
Even with a high-availability (HA) cluster, without a proper offsite backup and disaster recovery (DR) plan,
a single data center outage, hardware failure, or ransomware attack can still cause a complete operational shutdown.

The introduction of Proxmox Backup Server (PBS) has transformed backup operations
from traditional “disk image copies” into an intelligent framework based on incremental backups, deduplication, encryption, and cloud synchronization.

This article explains how to design a hybrid cloud DR architecture using Proxmox VE + PBS + Cloud Storage, including:
1️⃣ Overall hybrid backup architecture
2️⃣ Cross-site replication and cloud synchronization setup
3️⃣ Automated recovery process and orchestration examples

🧩 1. Proxmox Hybrid Cloud Backup Architecture

Architecture Overview

        ┌──────────────────────────────┐
        │     Local Proxmox Cluster     │
        │     (Taiwan Data Center)      │
        └────────────┬──────────────────┘
                      │
                Incremental Backup (PBS)
                      │
        ┌──────────────────────────────┐
        │   Remote Proxmox Cluster      │
        │ (Malaysia / Vietnam Site)     │
        └────────────┬──────────────────┘
                      │
             Cloud Backup / DR Storage
         (AWS S3 / Azure Blob / Wasabi)

This architecture includes:

Local backups (Primary PBS)
Cross-region replication (Remote PBS Sync)
Cloud-based long-term backups (S3 / Blob)
Automated recovery (Failover & Restore)

🧠 2. Key Features of Proxmox Backup Server (PBS)

Feature	Description
Incremental Backups	Transfers only changed blocks, minimizing bandwidth usage.
Deduplication	Identical data is stored only once, saving storage space.
Compression & Encryption	Uses ZSTD compression and AES-256 encryption for data protection.
Remote Sync Jobs	Automatically replicates backups to remote PBS nodes.
S3-Compatible Storage	Supports AWS, Wasabi, Backblaze, or self-hosted MinIO.
Snapshot-Based Recovery	Instantly restores entire VMs or individual files.

⚙️ 3. Cross-Site Backup and Replication Setup

1️⃣ Local Backup Configuration (Primary PBS)

Create a local datastore:

mkdir /mnt/pbsdata
proxmox-backup-manager datastore create local-pbs /mnt/pbsdata

Schedule daily incremental backups:

proxmox-backup-manager backup-job create \
--store local-pbs --schedule "daily" \
--notes "Daily VM backups"

2️⃣ Remote PBS Synchronization

Create a datastore on the remote PBS node:

/mnt/pbsremote

Set up the sync job:

proxmox-backup-manager sync-job create \
--source "local-pbs@10.0.0.11:8007" \
--remote "remote-pbs" \
--store "remote-pbs"

This ensures the remote PBS automatically replicates incremental backup data
from the primary PBS, enabling offsite disaster recovery (DR).

3️⃣ Cloud Storage (S3 / Blob) Integration

PBS 9.x natively supports S3-compatible storage:

proxmox-backup-manager datastore create cloud-backup s3://backup-bucket
proxmox-backup-manager remote create cloud --fingerprint auto

Set credentials:

export AWS_ACCESS_KEY_ID=xxxxx
export AWS_SECRET_ACCESS_KEY=yyyyy

Supported cloud destinations:

AWS S3
Azure Blob Storage
Wasabi / Backblaze
MinIO (on-prem S3-compatible platform)

🔄 4. Disaster Recovery Workflow

1️⃣ Detect and Trigger DR Mode

When a primary site or node failure is detected,
the remote PBS can trigger automated alerts and recovery workflows
through Prometheus, Grafana Alertmanager, or custom scripts.

2️⃣ Import Remote Backups into a DR Node

Mount the remote PBS datastore on the DR Proxmox node:

pvesh create /storage \
--storage dr-pbs --type pbs \
--server 10.0.1.50 --datastore remote-pbs

3️⃣ Restore Virtual Machines

qmrestore pbs:remote-pbs/vm-300 300 --unique 1

Or via Web UI:

Datacenter → Storage → remote-pbs → Backup → Restore

4️⃣ Automated Verification and Testing

Schedule regular verification of backup integrity:

proxmox-backup-manager verify-job create \
--store remote-pbs --schedule "weekly"

This ensures backup data is complete and restorable at any time.

☁️ 5. Proxmox + Cloud DR Best Practices

Category	Recommendation
Backup Frequency	Daily incremental + weekly verification
Remote Sync Frequency	Every 6–24 hours
Encryption & Compression	Always enable AES-256 + ZSTD
Recommended Bandwidth	Minimum 50 Mbps upstream
Disaster Recovery Test	Perform full DR drill at least quarterly
Remote PBS Capacity	~120% of total primary backup size

🧭 6. Real-World Example: Taiwan ↔ Malaysia Dual-Site DR

Environment Overview:

Taiwan Data Center: Main compute cluster (Proxmox + Ceph + PBS)
Malaysia Office: Remote DR site (Proxmox + PBS)
Cloud Storage: AWS S3 for cold data archiving

Workflow:

1️⃣ Taiwan PBS runs daily VM backups
2️⃣ Syncs incrementally to Malaysia PBS overnight
3️⃣ Malaysia PBS uploads copies to AWS S3
4️⃣ In case of outage, DR nodes in Malaysia restore VMs from local PBS

Architecture Diagram

[Proxmox Cluster - Taiwan]
   │
   ▼
[Local PBS TW]  --sync-->  [Remote PBS MY]  --S3-->  [AWS Cloud Storage]
   │
   ▼
[Local Restore / Instant Recovery]

🧮 7. Automated DR Orchestration

You can automate full failover workflows using Proxmox API, Ansible, or Terraform.

Example (Ansible Task):

- name: Restore VM from remote PBS
  command: >
    qmrestore pbs:remote-pbs/vm-200 200 --unique 1

Combine with:

Grafana Alertmanager for automated triggers
Slack / Telegram Bot for notifications
Ansible Tower / AWX for multi-step orchestration

✅ Conclusion

By integrating Proxmox VE + PBS + Cloud DR,
organizations can build a resilient and cost-effective disaster recovery system
that ensures business continuity, data integrity, and fast recovery across multiple sites.

This hybrid architecture allows enterprises to:

Instantly recover workloads after outages
Encrypt and protect data during transit and storage
Combine local performance with cloud-based durability

💬 In the next article, we’ll explore
“Proxmox Automated Cloud DR and Orchestration Implementation,”
focusing on practical automation with API, Terraform, and Ansible
for fully autonomous multi-region recovery workflows.