Enterprise data storage allows businesses to store and access large volumes of company information. The size of data that businesses can store depends on the storage type they use. For most large companies, a good data storage platform is essential to security and success: the demands of twenty-first century business requires that huge amounts of data be stored safely but also be easily accessible.
Before listing the main types of business storage, we note that some in the storage industry consider some of these solutions irrelevant, ineffective, or soon to be. Though the argument could also be made that some of these solutions aren’t true “enterprise-level” storage options, many businesses still use them, and they’re still considered to belong in the broad range of enterprise storage.
Direct-attached storage (DAS) is a form of data storage accessed through an external drive’s direct connection to a computer. Examples include external HDDs, which have to be connected to a desktop or laptop through a cable, and SSDs, which use a cable or an M.2 port in the motherboard. Direct-attached storage is a quick way for users to access their storage on a computer.
DAS is one of the less flexible methods of data storage. Hard drive data can only be accessed on the computer it’s plugged into, unlike network-based storage, which can be accessed from multiple locations. It’s not as scalable, because you can only connect as many drives as the computer has ports. Depending on the drive used, storage space can be limited.
NVMe, a protocol for solid-state drives, connects storage to computer memory using PCI Express buses. It’s a flash protocol that enables high sequential read/write speeds for stored data. While HDDs may be considered too slow for many enterprises, all-flash storage is in a class of its own. Many storage companies are offering flash arrays.
NVMe over Fabrics is largely a combination of DAS and storage area networks: NVMe devices connect to a computer, but the protocol is used over a network fabric through Ethernet, InfiniBand, or Fibre Channel. It’s a relatively new technology that powers NVMe connections across a whole network. NVMe-oF doesn’t have all of its kinks worked out yet, but it’s extremely fast. Variants of DAS aren’t obsolete, even for enterprises that require large capacities and high performance for their storage.
Network-attached storage (NAS) is a storage system that makes files stored in one device available to users in different locations, as long as they can access the network. NAS devices hold hard drives with stored files, such as documents, videos, music, and photos. A couple of the best NAS devices even support M.2 SSDs. For businesses that want plentiful backup for their hard drives, network-attached storage features Random Array of Independent Disks (RAID), which can copy data from one of the hard drives to another in case of disk failure.
NAS is a good choice for businesses that want lower storage costs than what cloud providers will offer. A good NAS device will cost money up front, but the lack of regular monthly payments can save buyers in the long run. NAS devices can also allow users to create access controls, so that they can manage employee file access—an important feature for enterprise security.
NAS isn’t typically advertised for enterprises by the NAS sellers themselves. They target their devices toward home users, small businesses, and occasionally mid-sized businesses. But NAS is also included in most overviews of enterprise storage, with good reason. NAS is a good choice for smaller enterprises that need on-premises backups.
Storage Area Networks
Storage area networks are high-speed data storage networks created between a group of specified devices. These devices can be in multiple locations. Storage area networks, or SANs, often use Fibre Channel technology to transmit signals across either copper or fiber cables. Switches direct network traffic between locations. Fibre Channel is a good protocol for SANs; it’s fast, and it doesn’t drop packets, even when the network’s bandwidth is stretched.
SANs are popular with businesses that need to store important applications, such as databases, and also need fast connection to those applications. SANs are also a good choice for enterprises because the size of the network is scalable: many devices and servers can join a storage area network. It can be big or small as needed.
Software-defined storage is a method of managing data storage services through software rather than hardware. The software does run on hardware, but it’s not bound to it: it’s an abstracted layer, and it can be moved between different pieces of hardware. Software-defined storage (SDS) is a flexible storage option, especially for companies with data centers, because it allows more hardware options. SDS infrastructure can run on a variety of hardware.
Because software-defined storage can run on multiple pieces of hardware, it can be much less susceptible to vendor lock-in (being trapped with one vendor because of legal, monetary, or necessity reasons). SDS also offers multiple options for storing data and can run in different environments. SDS architectures can automatically move stored data to different storage locations depending on how often it’s accessed. Moving data that isn’t accessed frequently, for example, to an archive platform could save a business more money because archive storage is less expensive.
Cloud storage makes stored data accessible from a distance in such a way that it’s backed up and still available if hardware fails. Cloud storage providers will often store multiple copies of the data in multiple locations, so that if one data center has a power outage, for example, the data is still safe in another one.
Before cloud storage, enterprises stored their data on premises on servers. This made the data quickly accessible, but enterprises ran the risk of losing their data when they didn’t have backups in another physical location. Cloud storage is more scalable than on-premises server storage. It’s available from more than one location and backs up the data, providing more safety in case of disaster.
Businesses have three choices for cloud storage environments: public cloud, private cloud, and hybrid cloud. A public cloud environment is completely managed by a cloud service provider. A private cloud is managed by the business, often on premises and on company servers. Hybrid cloud combines cloud environments in whatever way a business needs.
Hybrid cloud is a competitive offering from cloud providers: it offers businesses more flexibility for storing their data. Hybrid cloud is a good choice for businesses that have many different types of data and many workloads: some data might need to stay on premises, but some data might be best in an object pool in a public cloud, for example. A hybrid environment provides more options.