What is Storage as a Service?

Storage as a Service, or STaaS, is a vendor provided service that is synonymous with cloud storage. While this is mainly the case, there are STaaS models where vendors may manage the on-premise storage of clients, like tape drives, appearing more like an in-house consultant than a vendor. However, this is becoming obsolete due to digital transformation of smaller and enterprise companies moving into the cloud.

STaaS models are highly useful for businesses of any size. For small businesses, vendors provide the expertise, and the hardware for them to operate in competition with much bigger players. The ability to leverage the same technical capabilities, such as storage scaling, disaster recovery management, and security protection helps to level the playing field for many businesses, as well as lower barriers of entry for start-ups.

For enterprises, STaaS can be used to augment their data storage strategy. A typical case would be to store long term data with a provider, while maintaining operational/transactional data on on-premise infrastructure. But this configuration can also serve as a stepping stone for enterprises to shed their on-premise infrastructure and move to the cloud.

Storage as a service is ubiquitous. OneDrive, DropBox, iCloud, and Google Drive are some of the most familiar names for consumers. They offer easy, often free plans, to users that can be upgraded when more storage is needed. And they are accessible through the web, and other native computer apps making them easy to adopt.

For businesses, there are two popular options: renting storage based on the quantity or forming a service level agreement (SLA) between the business and storage provider. SLAs are more common the more a business needs to ensure specific conditions of the service. SLA will typically outline the responsibilities of each party, usually the vendor assumes responsibility for data as it enters their domain, including protection, integrity, and security. It also provides technical clarity by defining service expectations, such as uptime/downtime allocations (data centers are tiered 1-4, 4 being the top, with an uptime guarantee of 99.995%, or about 26 minutes of downtime a year), as well as other service expectations, such as repairs, like defining mean-time-to-repair (MTTR) thresholds, and other technicals, like read/write speed averages.

SLAs are defined in detail because data is intended for warm or hot use: it will be accessed regularly. Less demanding is the use of cold data storage, which may need an SLA, but also may be purchased on quantity. Pricing by quantity tends to be less costly, and may suit consumers who don’t need to access their data often.

The overarching benefit of STaaS is to offload labor and costs. By offloading some or all storage management to a vendor, effectively that branch of effort is reduced to a line item on the IT department’s budget. A very close secondary benefit, and perhaps more valuable to small businesses, is that the storage provider will assume much of the protection and security of data, something that may be difficult for smaller IT teams.

Other advantages include:

• Storage subscription models optimize costs
• STaaS solutions are inherently scalable, storage will expand and contract to fit consumer demands
• Data files are synced frequently so storage is always fresh
• Storage costs tend to be lower than on-premise infrastructure, highly beneficial for most businesses
• Top class STaaS also provides resiliency services, like disaster recovery to mitigate against worst possible outcomes and get servers and your data back up
• Security tends to be stronger for top class providers because storage is their primary business (however, always ensure security matches company needs and requirements)

Most vendors have made uploading data to their infrastructure easy. However, the main concern with using vendors is vendor lock-in. Vendor lock-in is when a business cannot easily, if at all, leave a vendor because it is impractical for any number of reasons, usually boiling down to being too costly. As easy as it was to upload data, it may not be easy to retrieve. It is always advisable to look at the data access and retrieval capabilities allowed by a cloud vendor, especially any costs. Data storage formats are also a concern, even if data is easily retrieved, sometimes uploaded data may be stored in a different format by the provider.

Other disadvantages include:

• Exceeding your storage allocation will almost always require extra costs
• Downtime, controlled mainly by the vendor, can pose a problem by disrupting service
• Consumers must adhere to the vendors configurations and may grant limited customizability

For many who work with storage, they tend to view it as files and folders, but this is not the only type of storage format in the cloud. The three main types are block storage, file storage, and object storage.

• Block Storage — Block storage, or block-based storage, is a popular option for storage formatting, and is often put in contrast to file storage, and object storage. Block storage, as the name implies, chunks out data into blocks, each with a unique identifier. These fixed-sized blocks then can be individually saved to any number of storage devices, wherever the system finds it most convenient. Block storage can conveniently spread data across multiple environments, enabling systems to take advantage of systems that best serve the needs of storage and data. Some blocks can be stored in a Windows system, while others on a Linux box, to be later assembled when a data request is made.
• File Storage — Very common in the minds of computer users, data is stored in files and subsequently organized into folders.
• Object Storage — Object storage is an approach to data storage meant to overcome the limitations that other file systems have in the face of the voluminous data generated by the growing number of users and devices in the world today. This storage system differs from others by organizing data using metadata rather than using a file hierarchy. Object storage is used chiefly in data lakes.

Data storage capacity has traditionally been a capital expense. The purchase and installation of on-premise servers to store long term data is always a continuous business challenge. On-premise infrastructure must be maintained, secured, and upgraded. As new technology is innovated to supplant older storage systems and meet user demands, data must be migrated to more capable units and outdated storage shuffled off. Utilizing software-defined storage software, which virtually manages multiple disparate storage units as if it were one, can breathe new life into old on-premise storage by linking them together and adding functionality found in newer units.

But, more than not, companies are moving to the cloud, because in the cloud, storage is inherently set up for scaling. Scaling requires more expense when done in-house. In the cloud, because storage and its costs are spread out among many users, the scaling effect can be achieved more readily. Enterprises can still avail of this advantage and have devoted storage capacity. And when storage reaches astronomical levels, Google and Apache both have created their own storage solutions for super high capacity data centers.

With top of class storage vendors present in the market, there are few advantages for maintaining on-premise infrastructure. A compelling reason is compliance. Personally Identifiable Information (PII), by regulation, must be maintained in the country in which it was produced. If you are an American, law dictates that your PII must reside on a server within the U.S. and not in say China. While it is possible to strike a deal with a vendor to ensure that data resides on a server in the appropriate country, many companies simply don’t take the chance, and instead divide data into two categories: operational but inconsequential data that can go to the cloud (like streaming content), and data that must be guarded at the highest level, such as personal, health and financial information, that can be retained on premise.

In the end, moving to the cloud tends to reduce what was a capital expense to an operational expense.

Three classes of storage as a service have stood out in the modern cloud. Each innovates upon basic cloud storage capabilities, and new developments will continue to evolve vendor offerings, subsequently creating new classes. Generally, these classes can be referred to as IaaS-based products, specialized STaaS providers, and capacity-on-demand offerings.

• IaaS-based products — Traditional cloud storage offers have evolved offering more sophisticated capabilities. Many now offer both object storage and block type storage, as well as file sharing and added features.
• Specialized STaaS providers — Specialized STaaS providers will offer hybrid products that bridge on-premise infrastructure and cloud spaces using a unified managed storage service.
• Capacity-on-demand offerings — This class charges consumers based on usage. Like IaaS-based products, these providers assume operations and own the infrastructure.