HP Technology at Work

The must-read IT business eNewsletter
Subscribe

Object storage: what is it, how does it work and is it right for you?

The object of object storage

February 2014

Data is beginning to define our world, as expressed in current trends like our growing dependence on mobile apps, the emergence of Big Data, and the ubiquity of cloud computing. Guess what data needs? Storage—more and more of it. However, traditional storage systems based on files and directories become unwieldy when handling Big Data–sized volumes. Object storage is emerging as one possible strategy for handling some large data workloads.
 
What is object storage?
Object storage, also called object-oriented storage, is a specific approach to storage that bundles logically related data into discrete units called objects. Object storage is different than traditional file storage in that the objects are not organized in a hierarchy, but instead exist in a flat address space called a storage pool.
 
Objects can vary in size and include more metadata than traditional files are assigned, such as security details or the object’s physical location, as well as a unique identifier so that the object can be easily found. Object storage management relies on metadata to locate and manipulate data, rather than by the data’s location in a file directory.
 
The problems with traditional storage
Traditional storage systems house data in fixed-size blocks contained in hierarchical folder and file structures. But these structures have a limit as to how many files they can accommodate. Today’s data, particularly the unstructured types created by multimedia and social media, are growing too fast for rigid storage structures to handle.
 
As data volumes grow, backup and recovery becomes more difficult, time consuming, and expensive. Hierarchical storage systems that exist today have to work much harder, making them more susceptible to mechanical drive failure and forcing administrators to spend time copying data to more systems to guarantee availability. Furthermore, traditional data-replication and backup approaches become more time consuming and costly to perform.
 
Searching traditional file hierarchies requires analyzing the entire index to find a single piece of information, which can be time consuming and CPU-intensive. And applications and end users must access information via a large central index, which can act as a bottleneck during periods of peak demand.
 
How object storage works
Object storage on the other hand, searches for data contained within objects by using the unique identifiers and/or metadata. Searching is faster because applications aren’t navigating through a predefined organizational structure or having to locate the physical location of a file. Applications are simply scanning a flat data pool for matching metadata or identifiers.
 
Object storage systems offer a range of efficiencies that can save organizations money. They are simpler and scalable, which makes them less expensive to set up and operate. Additionally, they use the same types of hardware systems as traditional storage, so no major hardware investment is required to get started. And because object storage uses a simpler architecture that speeds up data access, it can maintain performance on less costly drives.
 
Object storage can simplify data protection and management as well. The ability to include more metadata with objects allows organizations to apply security and management policies directly to the data they impact.
 
Object storage use cases
Because of its scalability and ease of access to data, object storage is currently best suited for housing large unstructured files, such as multimedia. It is mostly being adopted for storage in cloud-based environments and for applications that provide online file storage and sharing. Common use cases include:

  • Data accessibility: Create a file library that can be accessed, uploaded, or downloaded by employees in any location, at any time
  • Data archiving and backup: Store data that needs to be readily, but not constantly, available
  • Data compliance: Apply automation and data management and security policies, and retain data safely for audits and reporting
  • Social media and human information: Manage unstructured, high-growth or fluctuating data quickly and cost effectively

Even though object storage has existed since 1994, adoption remains somewhat limited. That’s because until recently, we didn’t widely share a need for alternate storage approaches. Organizations who are already struggling with managing their data, or who anticipate data challenges in the near future, should consider object storage among their menu of possible solutions.
 
You might also like
HP Cloud Object Storage
HP File and Object Storage
Around the Storage Block blog

Subscribe

Popular tags

Most read articles

HP Technology at Work

Contact Us
Search archive
Customize your content