What Is a Storage Area Network (SAN)? SANs Explained
Sometimes, computers and local storage resources fail to provide enough storage, data protection, scalable access, or high performance for mission-critical applications. Therefore, many enterprises choose to use SANs, or storage area networks, to meet their storage needs. In this article, learn what SAN is, how one works, and how they're used in computing environments.
What is SAN?
A SAN (storage area network) is a specialized high-speed network or subnetwork that stores and provides access to large amounts of data. These networks consist of multiple disk arrays or tape libraries, servers, and switches.
SANs play a crucial role in modern data storage. They interconnect multiple servers to shared pools of storage devices. A SAN is what enables multiple users to access consolidated block-level storage over a dedicated network that is separate from the main local area network (LAN). These networks operate at the block level, allowing servers to access storage blocks directly.
While both the availability and accessibility of storage remain a concern for enterprising computing, a SAN does offer solutions. Traditionally, you could only attach a limited number of storage resources to a server. This limited a network's storage capacity.
Administrators developed storage area networks to boast networking flexibility, enabling a server or multiple servers across numerous data centers to share a common storage utility. Even better, a SAN eliminates traditional bandwidth bottlenecks since storage devices appear to serve as attached drives.
Operating systems recognize storage area networks as locally attached hard drives rather than shared network drives. Any organizations that need to make reliable data available to many users can utilize a SAN due to its enhanced storage reliability and availability.
How does a SAN work?
A storage area network aims to connect servers with storage, making it easy to move data between servers and share it with users. Therefore, enterprise or company environments commonly use them, as they need to store large amounts of data and make it accessible to multiple users.
SAN networks use fiber channel or iSCSI protocols to connect servers to storage devices. This allows servers to access individual blocks of data on the storage devices efficiently.
Consider the two types of SAN solutions: fiber channels and the Internet Small Computer System Interface (iSCSI).
- Fiber channel refers to a high-speed network technology used mainly for storage. Fiber channel is a standard for storage area networks because it's fast for high-demanding applications that require constant data access. It offers data rates between 2 Gbps and 128 Gbps across metropolitan area distances, supporting high throughput and low latency.
- The iSCSI is a protocol that allows block-level storage devices to be accessed over a TCP/IP network. This type of SAN network can run at speeds up to 100 Gbps. It's cheaper than a fiber channel. In most cases, iSCSI creates virtual SANs (VSANs).
The connections between servers and storage devices create a network known as the SAN fabric, which operates independently from the regular data network. The key to the storage area network's functionality is its block-level data access.
When a server needs to read data, it sends a request over the storage area network fabric to the storage device. Instead of fetching entire files, SANs work at the block level where data divides into fixed-size blocks. This allows for extremely fast data access. The storage devices, often arranged in disk arrays or other configurations, retrieve and transmit the requested blocks back to the server.
Three layers of SAN configuration
SAN configuration is perceived to comprise three distinct levels, each of which has its own characteristics. They include the host layer, the fabric layer, and the storage layer.
Host layer
This layer represents the servers or computer systems that need to access the storage area network. These servers primarily run enterprise workloads, such as databases, to ensure efficient storage.
The servers have host bus adapters (HBAs) or network interface cards (NICs) specifically designed for SAN connectivity. These components serve as the initiators that generate read and write requests.
Initiator software running on the servers manages the communication along with the SAN. It initiates data transfer requests and communicates with the fabric layer through the initiators.
Fabric layer
The fabric layer contains the network switches, gateways, and routers. These devices create the network infrastructure that interconnects the servers and storage devices. They direct data traffic within the SAN. A SAN fabric encompasses all hardware that connects servers to storage devices through the use of fiber channel technology.
The fabric layer is all about how the storage connects to the user, such as via cables and network devices. This connectivity is mostly through either fiber channel or Ethernet, but both take pressure off LAN by associated data traffic to their own high-speed network.
This layer also involves zoning techniques that control which servers can access specific SAN devices.
Storage layer
The storage layer consists of physical data storage resources such as traditional magnetic HDDs. Other storage devices may include disk arrays, tape libraries, and various storage systems.
Even SSDs, along with optical drives, such as CD and DVD drives, reside there. A storage controller connects these devices to the fabric layer, which has responsibility for managing the storage devices. All the SAN components are logically tied together through these layers so that information can be transferred from the host layer to the fabric layer and eventually to the devices in the storage layer.
What is the purpose of a storage area network?
The purpose of a storage area network is to provide centralized data storage. Raw storage is treated as a pool of resources that can be centrally managed and allocated on an as-needed basis. A SAN, however, improves storage efficiency while simplifying storage management. This architecture is optimized for data-intensive tasks like database management where consistent high-speed access is essential.
If you're looking for a way to improve the performance of your data storage infrastructure, a SAN may be the right solution.
Characteristics of SANs
Storage area networks (SANs) have several key characteristics that make them a valuable tool for organizations of all sizes. These key characteristics are:
- Block-level access. Block-level access means that the systems store, access, and manage data at the level of data blocks, or fixed-size units of data. Blocks are smaller yet more efficient than file-level access for many applications.
- Performance. High performance is a hallmark of SANs. SANs are engineered for high performance, therefore making them suitable for demanding data-intensive applications. The block-level access contributes to their superior performance.
- High scalability. SANs are designed to grow along with an organization's data storage needs. Administrators can add new storage capacity to the network if needed with minimal disruptions.
- Multiple protocols. These dedicated networks support a variety of protocols, including fiber channel, Ethernet, and iSCSI. For that reason, SANs are compatible with a wide range of servers.
Components of storage area networks
A storage area network contains several key components that work together in order to provide reliable data storage. These components include hosts, node ports, cables, storage arrays, SAN switches, and SAN management software.
Host
The host is the server or computer that accesses the storage resources on the SAN. They need HBAs that support the SAN's protocol in order to communicate with the SAN fabric and storage devices.
Node ports
Node ports are the connection points or physical interfaces on devices within the SAN. They can exist on servers, storage arrays, or even other SAN components. Node ports serve as the entry and exit points for data in the SAN. All the fiber channel devices are node ports.
Storage arrays
Storage arrays are central storage devices that house the data in the SAN. The arrays are equipped with multiple HDD drives, SSD drives, and other storage media to provide reliable storage for data.
SAN switches
At the heart of storage area networks is the SAN switch. The switches, hubs, and directors connect hosts and storage arrays together, enabling communication between them. They can also be used to zone the storage area network.
Cabling
Cables are they physical connections that link nodes on the storage area network. The type of cable used depends on the SAN's protocol. Fiber optic cables and copper cables are commonly used to transmit data between servers, storage devices, and SAN switches.
Storage area network management software
SAN management software provides a centralized interface for administrators to configure, monitor, and manage the SAN. The software is for zoning, masking, and provisioning.
What are SANs used for?
SANs work well in a variety of instances. They help with:
- Simple centralized management. SANs can centralize storage resources in a data center, making it easier to manage and access data.
- Improving performance. SANs can improve the performance of applications that require high-bandwidth access to data, such as database servers and video editing software.
- Providing disaster recovery and backup capabilities. You can use a storage area network to replicate data to remote locations, providing disaster recovery and backup capabilities.
- Extending the reach of LANs. Admins can use SANs to extend the reach of networks to remote locations, such as branch offices.
- Securing data. SANs typically have built-in security measures to protect data from unauthorized access, keeping sensitive data safe.
- Enhancing data redundancy and availability. SANs often incorporate features like RAID (Redundant Array of Independent Disks) to improve data availability and redundancy. This helps protect against data loss due to hardware failures.
Limitations of storage area networks
Despite the benefits of using storage area networks, these infrastructures also present challenges. IT experts should consider the following potential disadvantages before deploying a storage area network.
- Complexity. Storage area networks can be complex to design, implement, and manage because of the various components that go into creating one. The protocols involved also require extensive care.
- Cost and administration requirements. These networks come at a high cost. Setting up storage area networks creates a strain on budgets, particularly for smaller organizations.
- Management. Managing a storage area network can be challenging, even for experienced administrators. These networks are complex systems with many moving parts. When problems occur, troubleshooting the network can prove difficult.
Though these limitations exist, many find that the benefits of using a storage area network outweigh the negative aspects.
Storage area networks vs. network-attached storage
Network-attached storage (NAS) and storage area networks frequently work together, although they represent different storage technologies. Both are methods for managing storage, but they work differently to achieve that goal. Consider the chart below to understand the difference between the two.
Aspect | SAN (Storage Area Network) | NAS (Network-Attached Storage) |
Full name | Storage area network | Network-attached storage |
Data access method | Provides block-level access | Offers file-level access |
Standard components | Storage devices, switches, HBAs, servers | Specialized NAS appliances, Ethernet network, file servers |
Ease of administration | Requires more IT administration | Easier to manage |
Data storage | Data is accessed as if it were a local hard drive | Servers access data as if it were a network-attached drive |
Network type | Dedicated network for data storage | Integrated into existing Ethernet network |
Cost | A storage area network is costlier, but extremely fast | Cheaper than storage area networks, but not fast |
Scalability | Highly scalable | Low-end; not highly scalable |
Frequently asked questions
What is a SAN fabric?
A SAN fabric is the network infrastructure that interconnects the storage devices and servers. The fabric ensures that data can be accessed reliably.
What are HBAs?
HBAs, or Host Bus Adapters, are specialized network interface cards installed in servers. They act as the communication bridge between the servers and the SAN fabric.
What are other storage area network protocols?
Other storage area network protocols include Fibre Channel over Ethernet (FCoE), InfiniBand, and NVMe over Fabrics (NVMe-oF). Each protocol has its own advantages and disadvantages.
What are alternative storage area network approaches?
Alternatives to traditional storage area networks include virtual storage area networks, unified storage area networks, converged storage area networks, and hyper-converged infrastructure (HCI).