The History of Object Storage
Object storage was a redundant and scalable storage system that provided a generalized data container with an index and metadata layers for the objects. Objects stored in the system were accessed via the index, and information pertaining to these objects was contained in the metadata. Files of all types and sizes could be stored as objects. Whether you were dealing with an office document, video clip, or complex data structures, this capability made it ideal for storing petabytes of data and billions of objects into an architecture designed to safeguard the information.
Back in the day, there were no standards for accessing data in an object store, so most of the implementations were proprietary in nature. As a result, the vendor had to provide the APIs. The adoption of 3rd party software vendors writing APIs was slower than expected, so object storage did not become mainstream.
Nevertheless, object storage was primarily used as an archive target for email and file systems. This type of storage made it easy to retrieve an object, simply by looking up its address in the index and fetching the object.
The Revival of Object Storage
With the new focus on Big Data and Open Source, object storage is making a major comeback. The advent of Amazon Web Services (AWS) and their Simple Storage Service (S3), which are both based on object storage, provide a new interface allowing for the larger adoption of this familiar storage solution. A number of vendors have written their software to work with the S3 interface, as well as hardware (object store) vendors who have implemented the S3 APIs into their offerings.
Meanwhile, the OpenStack community has been backed by several of the largest storage vendors to help create the OpenStack Swift object store. Complete with open APIs, this type of open source software initially found its way into research communities, and is now being adopted by a number of enterprise organizations as a way to deal with the data deluge they are facing.
Furthermore, a file system gateway has been created in order to access this storage in a more traditional fashion. Many of the objects that end up in an object store are created as unstructured data in a file system, and as a result, this type of store has become more desirable for use in data centers. Currently, there are several file based access and caching technologies available through object data stores.
OpenStack includes an open source version of a file based interface, and there are a number of hardware vendors that have also created similar solutions that allow this type of access. These solutions can be designed to grant disparate, remote users access to the same back-end data for collaboration. Based on the intelligence applied to data access and configurable policies, these caching applications can keep frequently used data in a local store to provide faster access for the user.
The Future of Online Storage
So what does all of this mean? Basically, as data continues to grow at exponential rates, additional solutions are being designed to facilitate online storage. Users will now have access to these storage methods from anywhere in the world. Although object storage is being introduced in a new and improved way, the application is still being utilized in a familiar fashion that provides all the benefits of scale, security, and availability.
At OnX, we believe that today’s storage needs have surpassed the capabilities of traditional architecture, and that most businesses and organizations need a multi-tiered approach to managing their data. In addition to understanding these technologies, we also know how to apply them to solve today’s business problems.
We’re here to provide our clients with simplified solutions in order to reduce operating costs and facilitate the management of online storage. Please feel free to share your questions, comments, and any additional feedback about this post.