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 DR (ABDR) provides enterprise-grade backup and DR capabilities for workloads running on Arcfra Cloud Operating System (ACOS). Featuring Backup and Recovery as well as Replication and Recovery, ABDR not only meets the requirement of “RPO ≥ 15 minutes, RTO in minutes”. It also applies to three-data-center DR, establishing a centralized DR center, rack room migration, anti-ransomware, and DR drill with lower investment and simpler architecture.

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 and Recovery

ABDR’s Replication and Recovery feature supports asynchronous replication (with a minimum RPO of 15 minutes), failover, test failover, and failback 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 and Recovery 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 Three-Data-Center DR

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. On one hand, the stretched active-active clusters can achieve real-time synchronization between available zones through data replication. In case of an available zone failure, VMs are automatically brought up in another available zone, achieving the data protection level of “RPO = 0 and RTO in minutes.” Off-site clusters can achieve periodic data replication via ABDR, and in the event of a failure of the entire active-active cluster, business operations can be switched to the DR cluster, ensuring a DR level of “RPO ≥ 15 minutes, RTO at the minute.”

#2 Centralized DR 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.

#3 Mutual DR Between Single-Center Production Clusters

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.

#4 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.

#5 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.

#6 DR Drill*

* To be launched by early Q2 2025.

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 and DR and read Documentation | Arcfra Enterprise Cloud Platform.