A SAN (storage area network) is a network of data storage devices. By taking storage devices and storage traffic off the Local Area Network (LAN), another network is created specifically for Storage Data. SAN storage solutions can range from a few servers accessing a central pool of storage devices to thousands of servers accessing TBs or more of storage.
In a SAN, data is presented from storage devices to a host so that the storage looks like it is locally attached. This is achieved through various types of data virtualization. SAN storage, then, is a high-speed network that provides network access to Storage. In some cases, SANs can be so large that they span multiple sites, as well as internal data centers and the cloud.
Storage area Networks differ from Direct Attached Storage (DAS) . In DAS, the data is directly attached to one server. A SAN, on the other hand, presents storage devices to a host such that the storage appears to be locally attached. Presentation of storage to a host is accomplished through the use of different methods of virtualization.
SANs are also different from Network Attached Storage (NAS). While NAS also takes storage devices away from the server to create a central pool of data, NAS storage connects directly to the network (LAN). In SAN storage, capacity is pooled and provided With a dedicated network. This enables faster communication over faster media.
Advantages of SAN storage
There are many SANs, so SANs are very popular in the enterprise. The advantages of SAN include:
- Eliminate bandwidth bottlenecks associated with LAN-based server storage
- SCSI bus-based implementations have no scalability limitations.
- High availability
- Higher fault tolerance
- Centralized storage management.
- Faster backup
- Global file system
- Fast data migration
- Fault tolerance
- Improve data security
- Improve storage utilization
- Higher scalability
- Improve application availability, such as multiple data paths
- Enhance application performance by offloading storage or isolating the network.
- Better data protection and disaster recovery (DR)
For SAN advantages, consider storage migration. If the data is on many servers, it is a laborious process for the storage administrator to shut down each server and transfer it to a new home. This may involve steps such as using the storage to unmount the file system, unplugging the device, moving the device, then connect it to another host and starting the file system on the new computer. In a SAN storage solution, moving an entire large storage array from one host to another is a simple matter. All you have to do is unmount the file system, quickly reconfigure the SAN, and then bring up the data on the new host. This saves a lot of time for storage administrators.
As storage grows, this architecture becomes more and more important. Trying to manage multiple terabytes of data on a server by the server is too much trouble. It requires a storage network to do the heavy lifting and eliminate the chore. For example, if you need to add storage from another array to a server, the architecture of a SAN connection allows you to assign logical unit numbers (LUNs) from multiple arrays to that server.
What is a LUN? A LUN is a unique identifier that specifies a single storage device or a set of physical or virtual storage elements to execute I/O commands with a host. The identified logical unit may be a capacity block on the storage drive, the entire drive, or a portion of a plurality of hard drives, SSDs or tape storage residing on one or more storage systems. Therefore, a LUN might refer to the entire RAID set, a single hard drive, or a partition on a drive.
The storage area network combines a range of storage devices for faster communication.
File and block
There is another key difference between the data storage on the server (or NAS box) and the way data is stored in the SAN. The former uses file-level storage and the latter uses block-level storage.
File-level storage is located on hard drives and NAS systems. Storage disks are configured with protocols such as NFS or CIFS for bulk storage and access to files, for example, file-by-file storage and access. This method is simple and easy.
Block-level storage creates raw storage volumes, where each data block is controlled by the operating system as if it were a single hard drive. These data blocks are independent of a particular file. The block then stores the management LUN instead of a single file managed in the NAS system.
The advantages of block-level storage include:
· Management flexibility
· Easy to store database management
· Faster, more reliable data transfer
• The ability to treat each storage volume as a separate disk drive controlled by an external server operating system.
· Access and control permissions are easier to manage.
It used to be a / or proposition – block storage or file storage. Recently, however, storage systems have been developed that can handle file storage and block levels within a single device. These unified storage systems and hyper-converged systems are becoming more and more common.
Storage area network technology
Many SANs use the Fibre Channel (FC) standard, a high-performance data communication technology that supports very fast data rates. FC SAN switches are used to connect devices in a SAN to create a so-called SAN fabric. These switches are somewhat similar to switches running on regular Ethernet because they act as a connection point to the network. By using FC SAN switch technology, a dedicated path is established between devices in the fabric to take advantage of high bandwidth.
A SAN usually consists of the following elements:
· Disk or solid state drive (SSD) array
· Disk array (or flash) controller
Host Bus Adapter (HBA). The HBA is basically an I/O adapter located between the host computer bus and the FC fabric. It is used to manage information transfer and reduce the impact of the SAN on host processor performance.
·FC switch
Fibre Channel networks typically have the core and edge switches. A core switch is often referred to as a Director switch. They are usually rack-mountable and have no single point of failure. Edge switches are smaller, simpler, and often do not have as much redundancy as possible.
Of course, the problem with a large network is that a small problem can affect the entire network. SANs can solve this problem by creating smaller structures in larger SANs. This is done using a variety of routing methods, some of which are vendor specific. For example, SAN switches typically use Inter-Switch Link (ISL) to transfer data from one switch to another.
In addition to FC, some SANs also utilize Fibre Channel over Ethernet (FCoE). This allows FC traffic to move across high-speed Ethernet infrastructure. The advantages of this approach include the ability to fuse storage and IP protocols onto a single cable.
iSCSI SAN
In addition to FC SANs, there is the Internet Small Computing System Interface (iSCSI) SAN (sometimes called IP SAN). iSCSI storage enables data to be transferred to or from a storage device over an IP network by serializing traffic from the SCSI connection. iSCSI SANs are often used in small and medium-sized businesses, and as a cheaper alternative to FC, it has become a prominent growth point in the past decade. This is because FC SANs have a complex, unmanageable reputation and require highly trained (and lucrative) experts who are generally expensive. By using Ethernet, iSCSI SANs can transfer SCSI commands in IP packets, so FC connections are no longer needed.
Advantages: No need to learn, build and manage FC networks; use the same cable to connect Ethernet-based LANs and storage.
Disadvantages: There is too much traffic blocking the possibility of a LAN – so small and medium-sized organizations use iSCSI storage more.
On the other hand, FC SAN is primarily used by large organizations and organizations that need distributed applications with fast local network performance. By providing multiple data paths, FC SANs provide better application performance and offload storage from IP networks.
Whether your organization uses an iSCSI SAN or an FC SAN, you can increase storage availability and utilization because administrators can consolidate resources and build tiered storage:
· Top-of-the-line ultra-fast storage for smaller data sets and mission-critical applications.
• The storage speed that follows is slower and the capacity is higher.
Of course, SAN software also needs to organize servers, storage devices, and networks to implement functions such as data transfer. To move LUNs and add storage to an existing file system, you need to use volume management software. RAID is also used in SANs (such as software-level RAID) to provide RAID 0 stripes.
SAN evolution
SAN technology has grown rapidly since the 1990s and early 2000s.
Unified storage: They are not separate files and block storage islands, but are integrated into one storage device.
Virtual SAN: A virtual SAN or VSAN is the result of software-defined storage. It is implemented with virtualization software such as VMware’s Hyper-V hypervisor. For those in a large virtualized server environment, VSAN provides simplified management and greater scalability.
SANs have been limited by the number of storage arrays that can be connected together in a single physical data center. However, a virtual SAN with a simple hypervisor-converged storage design provides a way to set up storage for a large number of virtual machines. This simplifies storage configuration and management in a virtual server environment.
Virtual SANs can also be clustered together to provide enterprises with greater scalability. By creating larger virtual SANs, it is relatively easy to centrally manage a large number of centrally managed storage. In addition, the latest virtual SANs are moving away from proprietary hardware, using industry-standard server components to reduce storage resources and providing vendor-independent architecture.
Enterprises are using virtual SANs to:
- Pool internal disk for the virtual server environment.
- Extend your storage network in data centers that span metropolitan distances.
- Synchronize replication data between two geographically distinct sites to provide a recovery point objective (RPO) of a few minutes.
- Set up each VM policy and configure it automatically.
- Create software-defined data centers that extend on-premise storage and management services across different public clouds to deliver a more consistent experience.
Original Article Source from https://www.enterprisestorageforum.com/storage-networking/storage-area-networks-in-the-enterprise.html