Data volumes, both structured and unstructured, are increasing at a very rapid pace. Traditional storage systems are ill-suited to this evolution, becoming complex in proportion to the quantity to be managed. Adapted to this new data environment, object storage is emerging as a "universal storage" model.
The explosion in data volumes will reshuffle the cards in the storage market. The changing context is impressive. According to a 2017 study by Forrester Research, 70% of the world's population will be connected to mobile devices by 2022, compared with 21% in 2013. The data associated with digital video with Ultra HD definition (3840 x 2160 pixels), streaming and VOD is pushing traditional storage to the limits of what's possible.
All companies, whatever their activity, but also end-users want fast access to content. Instant and unlimited access to files is becoming the rule.
Finally, the use of artificial intelligence (AI) and its derivatives, such as machine learning and deep learning, represents a decisive asset for companies. And here too, access to large volumes of data is decisive.
A unique "object
This context exposes the limitations of traditional storage technologies (block or network protocol layers such as SCSI commands). They either lack the flexibility needed to operate at petabyte scale (as in the case of NAS), or are costly to scale up, due to complex, proprietary architectures (as in the case of SAN).
Object storage is currently the most suitable solution. And yet it's not a recent development, dating back to the mid-90s. But the arrival of NAS and SAN around 2005 consigned object storage to the dustbin of technological innovation.
The data explosion has brought it back into fashion. Object storage is based on a different method of storage, organization and access. It doesn't use a file system, in which file metadata is stored separately from file data. Instead, it stores file data and metadata in the form of a single "element". This is referred to as an object, hence the name.
The advantages of object storage
- Easier queries: as each "element" has a unique identifier, it can be located precisely and quickly, even if it is not in the same physical location as the requester. This is not the case with traditional solutions, which are unsuitable for geographically distributed access or in the cloud.
- Reduced costs: the company only pays for the storage space used. There are no additional costs for multiple copies of objects. Some providers do not charge for transfers or incoming/outgoing bandwidth usage. Others charge for HTTP requests to send files to the Internet.
Offers Cloud can reduce your storage costs. However, we found that this was not always the case! It's essential to read and understand suppliers' offers, and to estimate the number of operations in your environments. To save on storage costs, some applications require private environments. We'll come back to this in a future article.
- Flexibility in managing a network of resources: they can be distributed geographically and extended worldwide. Object storage volume is dynamically allocated as you add or remove files.
- Simplified access: the storage architecture and each object can be accessed by software applications over the Internet, or a private network, using a REST (Representational State Transfer) API and a secure HTTP communication protocol.
Main use cases
The various advantages of object storage are of interest in a number of cases, the main ones being :
Archiving and backup
Archive and backup data is rarely modified. Companies that need to keep this type of data for a very long time represent a typical use case.
They benefit not only from the solution's attractive cost, but also from its redundancy system, which ensures the durability and availability of data over time.
Indeed, object storage providers ensure redundancy by storing copies of the same object on several nodes. When an object is created, it is created on a single node, before being copied to one or more other nodes, depending on the rules in force.
Managing a website's static pages
Object Storage lets you create and modify a static website (unlike dynamic pages, which can present information in different ways depending on visitor interaction, static pages don't) directly from the graphical user interface or via API requests.
The Internet of Things (IoT), Artificial Intelligence (AI) and machine learning
According to analysts such as Gartner and IDC, the growing use of these trend-setting technologies will massively increase the amount of data to be stored! Companies will be looking for low-cost, secure and reliable storage. Object storage will therefore be the most appropriate solution to accompany this revolution.
Using the S3 protocol
The success of the AWS storage offering has boosted the popularity of the S3 protocol. Its ease of use for developers has made it a must-have protocol. Many software publishers now use this protocol in their applications to read and write data.
Hybrid Cloud deployment
Many companies wish to outsource the storage of certain data outside the company. Object storage is the right technology for this purpose. It enables companies to extend their storage capacity without any financial investment.
Customizable metadata
It's the management of metadata that complicates the task for companies. By adding information, they enable a more sophisticated approach to analysis. And yet, object storage makes it possible to benefit from customizable metadata. Much more identifying information can be entered for each item.
It should also be remembered that object storage is not suitable for transactional data such as databases.
Object storage is an attractive solution for many companies. Even small and medium-sized businesses can benefit from object storage, enabling them to analyze their information assets in greater detail, while enhancing security through redundancy.
