What Is a Storage Area Network (SAN)?

Data volumes keep growing at astronomical rates. Corporations need to supply users with fast, simple access to vital information. Traditionally, users have had to go through a server to reach their data. Providing direct access over a storage area network (SAN) boosts performance, simplifies management, and enhances reliability.

Legacy storage options

The amount of data generated has been increasing worldwide, from 64.2 zettabytes in 2020 to an estimated 181 zettabytes in 2025. The world is on pace to create 230–240 zettabytes of data in 2026.

Organisations are managing unprecedented volumes of data. As data volumes have grown, the way they store and provide users with access has changed.

Legacy systems relied on direct attached storage (DAS) to deliver information. A server managed interactions with the data storage systems connected to it. Bottlenecks arose because the speed at which information moved from one system to another was constrained by the physical connection between the two.

Network-based storage offered a better option. This approach pooled data and allowed more than one computer to access it through a network, improving data sharing and collaboration.

Two types of network storage emerged. Network attached storage (NAS) collects data in a single device made up of redundant storage containers or a redundant array of independent disks (RAID).

This technique is easy to set up. Users typically access the data over an IP network. However, this design does not scale well because it's constrained by the device's processing power.

What is a storage area network?

A storage area network (SAN) is a dedicated network of storage devices that provide a pool of shared storage servers that multiple computers and servers can access. Storing data in a centralized shared storage architecture like a SAN allows organisations to manage storage from a collective place and apply consistent policies for security, data protection, and disaster recovery.

Unlike file-level storage systems, SANs deliver block-level storage access. This means servers interact with SAN storage as if it were locally attached disk drives, enabling faster performance and lower latency for demanding applications.

SANs network multiple devices so they offer more storage space and faster speeds than a NAS. They work well with large data sets, lots of users, and complex workloads. Another plus is a SAN can eliminate single points of failure. As a result, they improve reliability and system availability.

Storage area network architecture

A SAN comprises three distinct architectural layers that work together to deliver centralized, high-performance block storage:

Host layer

The host layer consists of servers and applications that require storage access. These hosts connect to the SAN fabric through host bus adapters (HBAs), which are specialized network interface cards designed specifically for storage connectivity.

HBAs translate server I/O commands into SAN-compatible protocols, enabling operating systems and applications to treat networked storage as locally attached disks. This abstraction allows databases, virtual machines, and enterprise applications to access shared storage resources seamlessly, without requiring application-level modifications.

Each host typically employs multiple HBAs to prevent single points of failure. If one adapter fails, the server continues accessing storage through alternate paths, maintaining uninterrupted operations.

Fabric layer

The fabric layer forms the network infrastructure that interconnects SAN hosts and storage devices. This layer comprises SAN switches, directors, routers, and high-speed cabling that create the data pathways between servers and storage.

The fabric layer provides redundancy through multiple alternate paths. SAN fabric architecture ensures that if one path fails due to cable damage, switch failure, or planned maintenance, traffic automatically routes through alternative connections. This multipath capability is fundamental to achieving the high availability requirements of enterprise applications.

The fabric layer also enables advanced features like zoning and LUN masking, which control which servers can access specific storage resources. These security mechanisms prevent unauthorized access and isolate storage traffic based on application requirements or organizational policies.

Storage layer

The storage layer encompasses the physical storage devices where data resides. This includes storage arrays, disk systems, solid-state drives, and tape libraries. Modern SAN storage layers typically deploy enterprise storage arrays configured with RAID for data protection, performance optimisation, and capacity management.

Storage devices present logical unit numbers (LUNs) to the SAN fabric. A LUN is a unique identifier for a storage volume that hosts a mount and format like a local disk. Administrators use LUNs to partition storage capacity, assign specific volumes to individual servers or applications, and implement access controls.

The separation of these three layers enables efficient storage management, allows for independent scaling of compute and storage resources, and simplifies troubleshooting when issues arise.

3 components in a SAN

A SAN includes three pieces of equipment that work together to deliver data to users:

Network interface card (NIC)

A network interface card (NIC) is a specialized circuit board that connects storage systems to a network, either wired (Ethernet) or wireless (Wi-Fi). This hardware component ensures communication between storage devices and the rest of the network, facilitating data delivery to users.

In SAN environments, host bus adapters (HBAs) serve a similar function but are specifically designed for storage protocols like Fibre Channel or iSCSI. HBAs offload storage processing from the server's main CPU, improving overall system performance. Each HBA contains specialized firmware and processors optimised for handling high-volume storage I/O operations with minimal latency.

Storage devices

Depending on application requirements or company needs, various storage devices are deployed within a SAN. These devices house the information and can include hard-disk drives (HDDs), solid-state drives (SSDs), flash storage, and hybrid storage options. The choice of storage technology impacts the speed, capacity, and efficiency of the SAN.

Modern enterprise SANs increasingly deploy all-flash storage arrays to meet the performance demands of virtualised environments, databases, and real-time analytics. Flash-based storage delivers significantly lower latency and higher input/output operations per second (IOPS) compared to traditional spinning disks.

Storage arrays within a SAN typically include built-in data protection features such as RAID configurations, snapshots, replication, and encryption. These capabilities ensure data durability and enable rapid recovery in the event of hardware failures or data corruption.

SAN switches

SAN switches play a critical role in connecting servers to storage devices and managing the data flow within the network. This hardware can include hubs, switches, gateways, directors, and routers. They work in tandem with SAN management software, which monitors and optimises the performance of the entire storage network.

SAN switches differ from traditional network switches in their specialized capabilities. They support storage-specific protocols, implement advanced traffic management for storage workloads, and provide features like zoning that segment the SAN fabric into logical groups for security and performance isolation.

Directors are high-end SAN switches designed for enterprise data centers. They offer greater port density, enhanced redundancy with hot-swappable components, and superior throughput capacity compared to standard SAN switches.

How does a SAN work?

The components of SAN include cabling, host bus adapters, and SAN switches attached to storage arrays and servers. SANs use block-based storage and high-speed architecture to connect servers to logical disk units (LUNs), a range of block storage from a pool of shared storage, and appear to the server as a logical disk.

When a server needs to read or write data, the process follows these steps:

1. The application sends an I/O request to the server's operating system.
2. The operating system forwards the request to the HBA.
3. The HBA translates the request into the appropriate SAN protocol (Fibre Channel, iSCSI, etc.).
4. The SAN switch receives the request and routes it to the appropriate storage device based on LUN mappings.
5. The storage array processes the request and returns the requested data through the fabric.
6. The HBA receives the response and delivers it to the server's operating system.
7. The application receives the data and continues processing.

This block-level access provides significantly better performance than file-level protocols because it eliminates the overhead of file system operations. The server's operating system manages the file system directly on the SAN volume, just as it would with a locally attached disk.

Multipath I/O software running on servers enables them to use multiple simultaneous paths to the same storage device. If one path experiences congestion or failure, I/O automatically flows through alternative paths, ensuring consistent performance and continuous availability.

SAN protocols

A storage area network protocol determines how devices and switches communicate with each other. A SAN can use one protocol or many because multiprotocol routers and switches move information from place to place in different ways. SAN technologies support multiple protocols that allow the layers, applications, and operating systems to communicate. The most common protocol used is the Fibre Channel Protocol (FCP), which is based on Fibre Channel (FC) technology. Internet Small Computing System Interface (iSCSI), a less expensive alternative to FC, is commonly used by small and medium-sized organisations. Let's take a closer look at the different types of SAN connections.

Internet Small Computer System Interface (iSCSI)

The Internet Small Computer System Interface (iSCSI) is an IP-based standard that links data storage devices over a network. Familiarity is an advantage. Enterprises use the same networking protocols for storage, storage management, and data networks, which simplifies system management.

iSCSI encapsulates SCSI commands within TCP/IP packets, allowing organisations to implement SAN storage over their existing Ethernet infrastructure. This approach significantly reduces deployment costs compared to Fibre Channel, which requires specialized cabling and switches.

Modern iSCSI implementations can achieve speeds up to 100Gbps on high-performance Ethernet networks. Organisations commonly deploy iSCSI for disaster recovery, remote office connectivity, and cloud storage integration, where the flexibility of IP networking provides advantages over dedicated Fibre Channel infrastructure.

Fibre Channel Protocol (FCP)

The Fibre Channel Protocol (FCP) is a gigabit-speed network technology primarily used for storage networking. The protocol was developed for supercomputers but became a common standard in enterprise data centers.

Fibre Channel delivers exceptional performance with speeds ranging from 8Gbps in older implementations to 128Gbps in the latest 128GFC specifications. The protocol provides extremely low latency and deterministic performance, making it the preferred choice for mission-critical databases, enterprise resource planning (ERP) systems, and high-transaction workloads.

Fibre Channel SANs support distances up to 10 kilometers with optical cabling, enabling storage replication between geographically separated data centers for disaster recovery purposes.

Fibre Channel over Ethernet (FCoE)

Fibre Channel over Ethernet (FCoE) is a protocol to route FC packets over Ethernet networks. This approach simplifies management because the enterprise LAN and SAN share a common network infrastructure.

FCoE enables data centre consolidation by converging storage and data networks onto a unified Ethernet fabric. This reduces cabling complexity, lowers infrastructure costs, and simplifies network management. However, FCoE requires specialized converged network adapters (CNAs) and lossless Ethernet infrastructure to maintain the reliability characteristics of Fibre Channel.

Fibre Channel over IP (FCIP)

Fibre Channel over IP (FCIP) is a tunneling approach. Here, Fibre Channel storage information is wrapped in TCP/IP network protocol. Because many organisations already have IP infrastructure in place, they can connect geographically dispersed SANs at a relatively low cost.

FCIP is commonly used for SAN extension across wide area networks, enabling data replication between data centers separated by long distances. Unlike iSCSI, which operates at the block level, FCIP tunnels entire Fibre Channel frames across IP networks, preserving the native Fibre Channel protocol end to end.

SAN vs. NAS vs. DAS: Understanding storage architectures

Organisations evaluating storage solutions encounter three primary architectures, each with distinct characteristics and optimal use cases:

Block-level access vs. file-level access

• Storage area networks (SANs) provide block-level storage, treating storage devices as raw disk volumes that servers can partition, format, and manage. This approach delivers maximum performance for structured data and transactional workloads like databases, virtual machine storage, and enterprise applications. The server's operating system directly manages the file system on SAN volumes.
• Network attached storage (NAS) operates at the file level, presenting pre-formatted file systems accessed through standard network protocols like Network File System (NFS) or Server Message Block (SMB). NAS excels at sharing unstructured data such as documents, media files, user home directories, and collaborative workspaces. Users access NAS as shared network drives.
• Direct attached storage (DAS) connects storage devices directly to individual servers via SCSI, SATA, or SAS interfaces. While simple and cost-effective, DAS lacks the sharing capabilities and scalability of networked storage approaches.

Organisations frequently deploy hybrid approaches, using SANs for performance-critical structured data and NAS for file sharing and archival storage. This architecture optimisation balances cost and performance based on workload requirements.

Benefits of deploying a SAN

SANs are quite popular. Worldwide sales were $19.4 billion in 2022 and are expected to increase to $52.3 billion in 2032, exhibiting a CAGR of 10.7%. The popularity stems from SAN's numerous benefits.

Simpler administration

A SAN centralizes storage resources, which can help reduce administrative overhead and lower total cost of ownership.

Storage consolidation enables administrators to manage terabytes or petabytes of capacity from a centralized management console. Instead of configuring dozens or hundreds of individual server disk subsystems, IT teams provision storage from centralized pools, implement organisation-wide policies, and automate routine management tasks.

Improved application availability

Storage exists independently of applications and servers. Since it's accessible through multiple paths, reliability increases, and availability improves.

SAN architectures eliminate single points of failure through redundant components at every layer. Dual HBAs in servers, redundant fabric paths through multiple switches, and dual storage controllers in arrays help ensure that component failures don't interrupt application access to data.

Better application performance

SANs offload storage processing from servers onto separate high-speed networks. Consequently, information retrieval occurs faster.

By dedicating network bandwidth exclusively to storage traffic, SANs prevent storage I/O from competing with general network traffic. This separation ensures consistent, predictable performance for latency-sensitive applications. Modern all-flash SANs deliver sub-millisecond response times and millions of IOPS, enabling real-time analytics and high-frequency transaction processing.

Enhanced scalability

SANs make it simpler for companies to boost storage as needed. As a result, they can scale more easily as the business grows.

Organisations can add storage capacity to existing SAN infrastructure without disrupting running applications. Storage arrays support non-disruptive capacity expansion through hot-swappable drive installation. The SAN fabric scales through additional switch ports or switch expansion, accommodating growth from dozens to thousands of connected servers.

Simplified backup and disaster recovery

SANs enable sophisticated data protection strategies that would be impractical with server-attached storage. Storage arrays provide built-in snapshot capabilities, creating point-in-time copies of data without impacting application performance.

LAN-free backup allows data to move directly from storage arrays to backup targets across the SAN fabric, eliminating backup traffic from production networks. This approach dramatically reduces backup windows and minimizes performance impact on production systems.

SAN-based replication synchronizes data between geographically separated storage arrays, providing disaster recovery capabilities with recovery time objectives (RTOs) measured in minutes rather than hours.

Use cases for SAN

Data drives corporate decision-making. Companies rely on SANs to provide data to employees, partners, and customers. This approach also enables them to protect information and ensure it's available when needed. The technology supports all applications. A few common ones include:

Database hosting

Enterprise databases generate intense I/O demands that benefit from SAN storage performance. Relational database management systems (RDBMS) like Oracle, Microsoft SQL Server, and PostgreSQL leverage SAN block storage for transaction logs and data files. The low latency and high throughput of SAN storage reduce query response times and support higher transaction volumes.

Server virtualisation

Virtual machine environments depend on shared storage to enable features like live migration, high availability, and resource pooling. VMware vSphere, Microsoft Hyper-V, and other hypervisors use SAN storage to store virtual machine disk files (VMDKs/VHDXs) that multiple host servers can access. This shared storage model allows virtual machines to move seamlessly between physical servers for load balancing or hardware maintenance without downtime.

Data consolidation and access

Historically, corporations managed data in a fragmented manner. Information was housed in autonomous applications. This approach led to a great deal of duplication and waste. A SAN collects information in a single place, which increases efficiency and the likelihood that users can access needed information.

Centralized SAN storage enables consistent data protection policies, simplifies capacity management, and reduces storage costs through higher utilization rates. Organisations can eliminate islands of underutilized storage scattered across individual servers.

Enterprise applications

Business-critical applications, including enterprise resource planning (ERP), customer relationship management (CRM), and financial systems, require the performance, availability, and data protection capabilities that SANs provide. These applications often serve hundreds or thousands of concurrent users with strict service level agreements for responsiveness and uptime.

Remote site data transfer and vaulting

Companies need to protect sensitive information. SANs enable them to create a remote copy of their data, which will be available if problems arise with the central system.

SAN-based replication provides robust disaster recovery by maintaining synchronized copies of production data at secondary locations. Synchronous replication ensures zero data loss for critical applications, while asynchronous replication balances data protection with performance over longer distances.

Best practices for SAN management

Implementing and maintaining a storage area network requires adherence to proven practices that ensure optimal performance, reliability, and security:

• Design with redundancy: Eliminate single points of failure by implementing dual HBAs in all servers, deploying redundant SAN switches, and ensuring storage arrays have dual controllers. Separate fabric paths should use independent switches to prevent a single switch failure from disrupting storage access.
• Implement proper zoning: Use fabric zoning to segment the SAN into logical groups based on application requirements, business units, or security policies. Zoning prevents unauthorized access, reduces troubleshooting complexity, and limits the scope of security vulnerabilities.
• Monitor performance proactively: Track key performance indicators including IOPS, throughput, latency, and capacity utilization. Establish baseline performance metrics and configure alerts for deviations that could indicate impending issues. Regular monitoring can help prevent performance degradation from becoming critical problems.
• Maintain comprehensive documentation: Create and maintain detailed SAN diagrams showing physical connections, zoning configurations, LUN assignments, and multipath settings. Documentation can accelerate troubleshooting, prevent configuration errors during changes, and ensure knowledge continuity when staff transitions occur.
• Keep firmware current: Regularly update firmware on HBAs, SAN switches, and storage arrays to address security vulnerabilities, compatibility issues, and performance improvements. Test firmware updates in non-production environments before deploying to production systems.
• Test failover mechanisms: Periodically validate that redundancy mechanisms work as designed through controlled failover tests. Disconnect cables, power off switches, and simulate component failures to verify that multipath software, alternate fabric paths, and storage array failover operate correctly.
• Plan for capacity growth: Monitor storage capacity trends and forecast future requirements. Provision additional capacity before existing resources reach 80% utilization to help prevent performance degradation and ensure headroom for unexpected growth.

Accelerate your SAN with all-flash solutions from Everpure

SANs are an important part of a company's technology infrastructure because they provide quick access to needed information. Large corporations rely on SANs to deliver information to users. Increasingly, small and medium-sized businesses are deploying these devices to house their corporate data.

Everpure delivers enterprise-grade SAN storage that combines the performance of all-flash architecture with simplified management and predictable economics. The FlashArray™ family provides organisations with robust, cost-effective block and file storage that eliminates the complexity traditional SANs require.

Everpure offers multiple FlashArray models designed to meet different performance, capacity, and budget needs:

FlashArray//X™: High-performance all-flash storage for mission-critical workloads that require low latency, consistent performance, and enterprise-grade resiliency. Ideal for Tier-1 databases, virtualisation, cloud-native applications, and real-time business services.

FlashArray//XL™: The highest-performance and most scalable platform in the FlashArray family, designed for large enterprises running the most demanding, data-intensive workloads and large-scale consolidation initiatives. Ideal for large databases, high-transaction environments, and platinum-tier storage services.

FlashArray//C™: Capacity-optimised all-flash storage that balances performance, scale, and cost for workload consolidation, virtualisation, backup and recovery, and file services, and Tier-2 workloads.

FlashArray//E™: All-flash storage engineered to deliver economics comparable to or better than disk, making it ideal for data protection, active archive, disaster recovery, and other large-capacity workloads traditionally deployed on HDD-based storage.

Key advantages of Everpure SAN solutions

All FlashArray platforms include the Purity operating environment, which provides built-in data reduction averaging 5:1, eliminating the need to purchase excess capacity. Evergreen® architecture ensures non-disruptive upgrades without forklift replacements, protecting your technology investment.

SafeMode™ Snapshots deliver immutable recovery points that protect against ransomware attacks, while Pure1® AI-powered management provides predictive analytics that prevent performance issues before they impact applications.

For more information about how Everpure can help your company address its data needs, get your storage evaluation kit.