Why Object data storage is chosen over File level storage?

By CIOReview | Monday, August 8, 2016

The digital world is formed from the ocean of binaries and enveloped in mesh of networks. The stream of data is continuously flowing. Accumulation of continuously moving data or information introduced the term ‘Big data’ in the present IT market that hatched new data storage technologies like NAS, DAS and SAN. These data storage devices are empowered by traditional methods of File level storage and Block level storage, which leverages different protocols for file access. As information started flowing through the internet, Object level storage system became more prominent. Instead of asking vendors, “Which is the best storage system—Object level or File level?” a CIO needs to understand the function of each storage method himself before making the crucial decision.

A hawk’s eye view of File storage and Object storage 

File Storage System: A File storage system is also called as ‘file-level’ or ‘file-based’ storage system. Data in this system is stored in the form of files and folders following a hierarchical structure.  These data can be presented in both, the storing system and the retrieving system in the same format. In Unix or Linux, Network File System (NFS) protocol is used for accessing data whereas, in Microsoft Windows Server Message Block (SMB) protocol is used for data access.

NFS allows the client to store and view files on a server as if they were on the client’s system. SMB offers its clients, shared access to files, printer, and serial ports by establishing miscellaneous communication between nodes on a network.  

Object Storage System: Like files, objects consist of data which are not organized in hierarchy. Every object is placed at a same level onto a flat address space called a storage pool where one object cannot be placed over another object. In this data storage system each object is assigned a unique identifier, enabling the user or client to retrieve the data without the need to know the physical location of it. Object storage is useful in optimizing cloud computing by automating and streamlining the whole data storage system. 

Providing an analogy of valet parking, Mark Goros, CEO and Co-Founder, Carnigo says, “With object storage, you drop off your car, and when you want it back all you need to do is present a ticket. You don’t care where or how your car is parked just as long as you get it back quickly in good condition.” Mark concludes saying that today’s enterprises are focusing mainly on agility of the stored data.

How Object storage subdues File level storage?

Knowing what the File storage system and Object storage system does, the next most cardinal step for a CIO is comparison. CIOs can ease this process of evaluation by harvesting attributes and weighing them to the enterprise needs. This can help in choosing the right system and portray its utility for the organization. 

Scalability: The file level storage system is based on hierarchical file structures having limited namespace and more restriction. The utility in file level storage does not offer geographical access limiting it to internal networks. The Object storage system has an infinitely scalable flat structure that's only limited by the number of bits of the object ID. 

Metadata: Metadata provides contextual information about the data. In case of a file based storage system, metadata is limited to File attributes whereas, in Object based system metadata can be enriched with any custom attributes. For customizing attributes in a file level storage the system needs application to control any additional information related to files. Terri McClure, Senior Analyst, Enterprise Strategy Group says, “Custom metadata allows building rich, self-contained file objects that can be stored away in the object store, enabling building massive unstructured data stores with reduced administration overhead.”    

Redundancy: Object storage features redundancy and high availability by storing copies of the same object on multiple nodes; the nodes can be within the same data centre or dispersed geographically, reducing the complexity. For file level storage system, managing files across multiple instances is a tiring work and is directly proportional to complexity. 

Protocol support: File level storage supports protocols which work well within the data centre where latency is not an issue. But file level deployment fails where data needs to be scaled geographically and cloud access is the necessity. Object storage, on the other hand can be accessed through REST API over HTTP where users can control the object through commands sent over HTTP—put, get, read, delete, and list.

Cost Efficiency: As Object storage enhances cloud-based storage system, potential costs such as costs incurred on network administrators and scalability of the system capacity is purged; while limitations in scaling through internet and multiple complexities has antiquated file level storage.