Many enterprises deploy core applications and databases on virtualization platforms. These applications usually involve diverse workloads with various demands in data protection strategies. Additionally, some users face more complex disaster recovery (DR) challenges, such as DR drills and ransomware protection. As a result, seeking a comprehensive, reliable, user-friendly, efficient, and cost-effective IT infrastructure solution for DR sites has become a top priority for many organizations.

As the native and agentless data protection service of Arcfra Enterprise Cloud Platform, Arcfra Backup and Disaster Recovery (ABDR) provides enterprise-grade backup and DR capabilities for workloads running on Arcfra Cloud Operating System (ACOS). ABDR offers a versatile, cost-effective architecture that spans standard backup, asynchronous replication (RPO ≥ 15 min), and synchronous replication (RPO = 0), enabling robust business continuity across the following scenarios:

• Real-time DR Center for Same-city Deployment
• Minute-level RPO DR for General Workloads
• Unlocking the Value of DR Data
• Secure Networking with Everoute

ABDR: Reliable, Comprehensive, Efficient, and Easy to Use

Backup and Recovery

ABDR’s Backup and Recovery feature enables the backup of Arcfra Virtualization Engine (AVE) configuration data and virtual volume data to external NAS servers (via NFS3/NFS4 protocols). The backup service supports both full and incremental backups, allowing flexible configuration of backup schedules and recovery point retention policies to meet diverse data protection needs. Users can restore virtual machines (VMs) either to their original state or by reconstructing a new VM using backup files from the backup repository.

• Enhanced technologies to improve backup speed: Backup data volume is reduced through optimizations such as data compression, breakpoint transmission, and skipping unallocated, all-zero, or unmodified data blocks. ABDR also supports parallel backup of multiple VMs and concurrent data transfer for multiple volumes within a single VM, improving data transfer speed.
• Flexible backup strategy to meet custom requirements: Users can configure full/incremental backup cycles at varying granularities to comply with industry-specific data backup standards and regulatory requirements. Additionally, backup time windows, task quantities, and transmission rates can be adjusted to minimize the impact of backup tasks on other operations during peak periods.
• Dual recovery methods to meet diverse needs: ABDR supports both source VM recovery and the rebuilding of new VMs, catering to various needs such as business continuity and development/testing.

Replication

ABDR’s Replication feature supports both synchronous replication (RPO = 0) and asynchronous replication (with a minimum RPO of 15 minutes).  It provides failover, test failover, failback, parmanent failover within a single site and across two sites. In case of a disaster, the running business can be quickly switched to the DR cluster to resume business operations. The Replication feature not only ensures efficient data synchronization and flexible replication policies but also offers the following advantages:

• Support for data protection and DR across different sites: ABDR enables data synchronization, failover, and DR of replica objects between multiple clusters, whether within the same site or across different sites, achieving cluster-level DR and ensuring business continuity.
• Selecting optimal fault-handling strategies: Support emergency failover and planned failover to deal with unexpected emergencies and anticipated switchovers like DR drills. After the failover, users can choose from failback and permanent failover; depending on the recovery status of the source object, users can choose to synchronize any data changes from the replica VM during failover back to the source VM, or let the replica VM permanently take over from the source VM.
• Multiple strategies to reduce downtime after VM failure: ABDR supports the configuration of comprehensive fault-handling strategies via a unified management interface. Users can pre-define VM recovery strategies and network settings for replica VMs, such as batch failover operations with customizable VM boot orders and delays to ensure an orderly recovery process. These features enable fast DR of application systems across different clusters.

Use Cases

#1 Minute-level RPO DR for General Workloads

Centralized Disaster Center

Users can build a centralized DR center based on AECP, providing DR protection for multiple data centers running business systems.

• Unified DR: In the event of a disaster at any data center, the DR center can take over operations and restore data, enabling rapid business switchover.
• Centralized management: With centralized management of data replicas, users can analyze and mine data without disrupting production operations, maximizing the value of the data.
• Reduced total cost: A centralized DR center eliminates the need for costly, dispersed DR clusters across multiple locations, which significantly improves the cost-effectiveness and resource utilization of the overall DR infrastructure.

Dual-Cluster Mutual Disaster Recovery Solution

With AECP, users can build DR resource pools separately in the two production clusters at the same data center to realize the mutual DR between dual production clusters. Each production cluster supports both the business VMs and VM replicas and can backup data from the other cluster.

• Significant cost reduction for DR deployment: Eliminate the need for dedicated DR appliances — only partial node expansion in production clusters is required to accommodate DR resources. Production clusters simultaneously handle business operations, backup storage, and replica data pools, maximizing hardware utilization.
• Streamlined operations and management: Avoid deploying multiple independent systems, reducing maintenance complexity caused by heterogeneous storage environments. AOC centralizes backup and replication management, minimizing cross-system coordination overhead.
• Flexible and versatile DR resource pools: DR resource capacity can be dynamically scaled to prevent overprovisioning, lowering Total Cost of Ownership (TCO). File storage adapts seamlessly to business growth, delivering high-efficiency storage support for diverse workloads.

#2 Real-time DR for Critical Workloads

Users can build active-active clusters in the same city and cross-city DR clusters based on Arcfra Enterprise Cloud Platform (AECP), addressing center-level and city-level failure scenarios.

Real-time DR Center for Same-city Deployment

Achieve real-time disaster recovery for mission-critical workloads within the same metro area. This solution ensures zero data loss (RPO=0) and low RTO without requiring application changes.

• Real-time Protection for Mission-Critical Business: Achieve real-time data sync between metro-area sites with RPO=0. This solution offers VM-level protection without requiring application refactoring, ensuring a seamless fit for your existing architecture and workloads with low RTO requirements.
• Strong Consistency & Rapid Recovery: Primary writes are mirrored to the standby site in real-time to guarantee absolute data consistency. In the event of a failure, services can be restored at the disaster recovery site within minutes, significantly boosting business continuity and system availability.
• High Adaptability & Simplified Management: Deploy rapidly across diverse network environments with zero application changes. Centralized management ensures the safety of your core business while reducing operational complexity and maintenance costs.
  

Two-Site Three-Center Deployment

Implement a two-site, three-center DR architecture combining real-time synchronous replication for same-city sites with asynchronous replication for offsite disaster recovery, ensuring multi-level protection for critical workloads.

Same-City DR Center (Real-time, RPO=0)

• SMTX Active-Active Stretch Cluster: Data replicas are synchronized in real time between same-city sites with RPO=0 via active-active stretch clustering. Automatic takeover by the other zone when one availability zone fails.
• Synchronous Replication: Real-time VM-level replication between same-city sites with RPO = 0. In case of a VM failure, workloads can be quickly recovered manually at the local DR center, ensuring zero-data-loss protection for critical applications.

Offsite DR Center (Periodic, RPO Configurable)

Asynchronous replication syncs data to a remote disaster recovery center every 15 minutes (at the minimum cycle). Safe failover to the offsite disaster recovery cluster if the active-active cluster fails. Achieves multi-level data protection. In case of a regional disaster, business switches to the local DR center. For large-scale disasters affecting both zones, the disaster recovery center takes over and restores operations.

#3 Rack Room Migration

Users can leverage ABDR to reduce potential risks of rack room migration by replicating data from the production site to the DR site and then migrating the data to the clusters in the new rack room.

• Ensuring migration efficiency alongside data security: With failover features, business VMs can be migrated in batches according to customized sequences, which enables rapid service recovery. Additionally, using Backup and Recovery features, users can create multiple data copies across sites to enhance data security.
• Cross-cluster VM migration: ABDR supports cross-cluster VM migration between different physical servers, offering various migration methods to facilitate flexible VM scheduling while ensuring business continuity.
• Decoupling software from hardware to reduce migration risks: As business data is synchronized and stored in the DR facility’s storage systems, business services can run normally even if the hardware fails during the rack room relocation. Once data migration is complete, the relocation of physical devices can be performed at any time.

#4 Anti-Ransomware

By integrating the network traffic visualization feature of Arcfra Operation Center (AOC), the software-defined networking and security components of Arcfra Network Service (ANS), and ABDR, users can leverage AECP to protect their systems from ransomware through a three-stage strategy: actively detect ransomware attacks, isolate infected VMs, and recover data in a secure environment.

• Abnormal traffic detection: Visualized network traffic size and direction as well as proactive alerts for abnormal traffic allow users to quickly identify anomalies like network attacks.
• One-click isolation of abnormal VMs: ANS distributed firewall can automatically block suspicious traffic while allowing admins to investigate and eliminate threats. This feature effectively helps to contain virus transmission and suppress the spread of risks.
• Reliable data recovery: With ABDR, users can rapidly resume services with backup files or replica VMs, ensuring business continuity and minimizing data losses.

#5 DR Drill

By integrating ABDR’s test failover feature and ANS VPC networking feature, users can replicate business VMs within the existing clusters to conduct DR drills without impacting the normal business operations.

• No need for a separate DR drill environment: Users can leverage VPC to create an isolated zone within the current environment for DR drills.
• No impact on business VMs: The source VMs’ operations remain uninterrupted, with no downtime. The test VMs operate independently of the source VMs, ensuring that the testing process does not interfere with the data of the source VMs.
• No impact on business services: During the drills, any fluctuations in network traffic or the injection of test data are confined to the isolated zone, ensuring that these changes do not affect the business operations within the production environment.

Having questions? Learn more about Arcfra Backup & Disaster Recovery and read Documentation | Arcfra Enterprise Cloud Platform.