What should I use to build a horizontally scalable enterprise storage for backups and archives?
I’m planning to create a horizontally scalable enterprise storage for backups and archives. Currently, all my infrastructure runs on Synology. I’m considering switching to TrueNAS, but I don’t fully understand how to set up multiple servers when the capacity of a single device is exceeded. Could you share your experience on what technological foundation would be best for building such a solution?
TrueNAS SCALE is the optimal solution for building a horizontally scalable enterprise backup storage system thanks to its Linux-based architecture, support for clustering via GlusterFS, and the ability to combine up to 100 nodes into a single system with a total capacity of up to 2 EB. This platform is specifically designed for horizontal scaling, ensuring high availability, fault tolerance, and the ability to seamlessly add new servers as data volumes grow.
Table of Contents
- Horizontal Scaling Technologies for Backup
- TrueNAS SCALE Architecture
- Comparison with Other Solutions
- Practical Implementation and Recommendations
- Performance Optimization for Backups
Horizontal Scaling Technologies for Backup
Horizontally scalable solutions for enterprise backup are based on several key technologies, each offering unique capabilities for distributed data storage.
TrueNAS SCALE uses GlusterFS to create a distributed file system, allowing multiple servers to be combined into a single storage repository. This technology provides:
- Scaling up to 100 nodes in a single cluster
- Capacity up to 2 EB (exabytes) of distributed storage
- Support for various replication and distributed storage modes
Azure Blob Storage is a cloud solution from Microsoft designed to handle massive volumes of unstructured data. As noted in research by Airbyte, Azure Blob Storage supports multimedia content, documents, and backups with distributed access capabilities.
Veeam Data Platform provides unified protection for cloud, virtual, physical, SaaS, and Kubernetes environments. According to Gartner, this platform guarantees constant application and data availability.
TrueNAS SCALE Architecture
TrueNAS SCALE is a specialized solution for horizontal scaling developed by iXsystems. Key architectural features include:
Scalability and Performance
TrueNAS SCALE supports impressive scales:
- Up to 64 CPU cores
- Up to 1 TB of RAM
- Up to 1200 disks of 18 TB each (over 20 PB raw storage)
As noted in the official TrueNAS blog, these characteristics make the system one of the most powerful on the market for enterprise needs.
Cluster Architecture
TrueNAS SCALE clusters allow three or more nodes to be combined into a single system. As explained in the XDA Developers article, clustering provides the ability to use the storage and computing capabilities of multiple servers as a single entity.
Storage Modes
TrueNAS SCALE supports various data storage modes:
- Replicated volumes - automatic data replication between nodes
- Distributed/Erasure-coded volumes - distributed storage with redundancy
- Hybrid configurations - combination of SSD and HDD for optimal performance
It’s important to note that TrueNAS SCALE uses a shared storage architecture, although it also supports a “shared nothing” approach for distributed clusters, as indicated in Storage Gaga.
Comparison with Other Solutions
TrueNAS SCALE vs TrueNAS Core
Key differences between versions:
| Feature | TrueNAS Core | TrueNAS SCALE |
|---|---|---|
| Operating System | FreeBSD | Linux |
| Clustering Support | No | GlusterFS |
| ZFS Version | 2.1.14 | 2.2.2 |
| Scaling | Vertical (scale-up) | Horizontal (scale-out) |
| Fleet Management | Limited | Full cluster management |
As shown in the Lawrence Systems comparison, SCALE is specifically designed for horizontal scaling, making it the preferred choice for growing infrastructures.
TrueNAS SCALE vs Synology
While Synology offers reliable solutions for small and medium businesses, TrueNAS SCALE provides significant advantages for enterprise needs:
- Scalability: TrueNAS SCALE supports up to 100 nodes compared to Synology’s limited node count
- Openness: Support for Docker and Kubernetes for advanced scenarios
- Performance: Higher performance for large data volumes
- Flexibility: Ability to use any compatible hardware
Other Enterprise Solutions
Caringo Swarm offers private cloud solutions without proprietary hardware lock-in, as mentioned in CIO Influence.
Synology C2 Object Storage provides cloud object storage compatible with S3 API, allowing flexible data storage, archiving, and recovery.
Practical Implementation and Recommendations
Infrastructure Planning
When migrating from Synology to TrueNAS SCALE, consider the following aspects:
-
Network Performance: For optimal cluster operation, high-speed network equipment is recommended, as demonstrated by community configurations.
-
Storage Subsystem Optimization: For enterprise backups, it’s recommended to use:
- Fast SSDs for cache (SLOG)
- High-performance SSDs for active data
- Large HDDs for archives and long-term storage
-
Redundancy: Implement mirrored configurations for critical data
Step-by-Step Migration
- Current Infrastructure Audit: Assess data volume and performance of existing Synology system
- Pilot Deployment: Install TrueNAS SCALE on test equipment
- Data Migration: Gradual migration of backups while preserving versions
- Integration: Configure integration with existing backup systems
- Monitoring: Implement performance and availability monitoring systems
Hardware Configuration
Example enterprise configuration for TrueNAS SCALE:
- Nodes: 3-5 servers with Intel Xeon processors
- Memory: 256-512 GB ECC RAM per node
- Network: 10/25GbE for inter-node connections
- Disks:
- NVMe SSDs for cache and active data
- SAS/SATA HDDs for archives
- Separation into pools for different purposes (BigPool, SSDPool, AppPool)
Performance Optimization for Backups
ZFS Configuration for Backups
TrueNAS SCALE uses ZFS for data management, providing exceptional optimization capabilities:
- Snapshots: Automatic creation of recovery points
- Compression: Reduction of stored data volume
- Deduplication: Space savings when duplicate data exists
- Caching: Optimization of read/write performance
Automation and Monitoring
TrueCommand provides centralized management for TrueNAS SCALE clusters. As noted in StorageReview, TrueCommand includes wizard assistants for creating SMB clusters and managing storage.
Integration with Backup Systems
To ensure seamless operation with existing systems:
- Configure S3-compatible APIs for integration with backup systems
- Configure SMB/NFS protocols for client system access
- Automate backup creation processes through scripts
It’s important to note that TrueNAS 24.04.1.1 brought significant improvements, as demonstrated by Lawrence Systems, including new clustering and high availability features.
Sources
- TrueNAS SCALE Clustering Overview
- CIO Influence - Top 20 Enterprise Data Storage Providers
- Airbyte - Big Data Storage Solutions
- Gartner - Enterprise Backup and Recovery Software Solutions
- Storage Gaga - TrueNAS SCALE Clustered SMB
- StorageReview - iXsystems Updates TrueNAS Scale
- XDA Developers - TrueNAS Scale vs Core
- Lawrence Systems Forums - TrueNAS Core vs Scale 2024
- Synology - Enterprise Backup Storage
- TrueNAS Community - Build 2024 Questions
Conclusion
TrueNAS SCALE represents the optimal solution for building a horizontally scalable enterprise backup storage system thanks to its specialized architecture for distributed storage. Key advantages include the ability to combine up to 100 nodes into a single cluster, support for various data storage modes, and high performance for processing large data volumes.
For a successful transition from Synology, it’s recommended to start with a pilot deployment, gradual data migration, and careful network infrastructure planning. Integration with existing backup systems through S3 API and SMB/NFS protocols will ensure seamless operation in a mixed environment.
For further study of the topic, it’s recommended to review the TrueNAS SCALE documentation, participate in the user community, and conduct hardware testing before full-scale implementation.