What's This?

The easiest way to think of object storage may be to compare it to the containers we’ve been discussing in this blog series. This will also introduce the first of many advantages of object storage, which is its ability to manage data in context.

The Major Components

File storage consists of data in raw form -- ultimately zeroes and ones arranged to express specific characters or other entities such as pixels.  Block storage organizes the files into fixed-sized blocks for easier management and transport.

Object storage consists of three key components:

1. The data, just as file storage handles it. It could be as small as an email or as large as video and other rich media files.
2. Metadata. Information about the data that characterizes it and provides context regarding what the data is, what it should be used for, privacy and confidentiality information, and other relevant information as defined by the creator of the object.
3. A globally unique identifier. Being globally unique allows each object to be found across distributed systems without the user needing to know the actual physical location.

Distributed Systems: The Fourth Component

Many of the most important values of object storage are derived from the fact that it leverages the advantages of distributed, scale-out architecture, which expands capacity by simply adding nodes. No overhead is added in this process, so object storage performance tends to scale directly with capacity in a distributed environment.

This also adds tremendous resilience as data objects are stored redundantly across multiple systems. Should one or more nodes fail, the data and its context can be readily accessed from alternative nodes. Note that we said “accessed” and not “recovered.” High availability is achieved through the redundant storage of objects across a minimum of three nodes in most cases. Users remain unaware that nodes have failed.

Where Object Storage Fits

While object storage scales far better than block storage because of its distributed nature and scale-out architecture, block storage offers more interactivity with the data more readily and can be accessed directly by the operating system with no performance degradation. This often makes block storage a better choice for highly transactional environments and use cases where the data must frequently be changed and updated.

Object storage responds directly to the meteoric growth of big data entities that must be stored and interrogated occasionally. As some data stores reach petabyte and even exabyte scale, the superior scalability of object storage becomes indispensable. Ideal use cases include web content, especially with the exponential growth of smaller public cloud providers that is prevalent in the market today. Data backup and archival use cases, including eDiscovery, are finding object storage to be far more responsive, resilient, durable, and scalable than other alternatives. Large file requirements such as those used in music and video recording and medical imaging are also ideal uses cases for object storage.

Superior Solution At Reduced Cost

For use cases requiring petabyte-scale storage and beyond, object storage solutions from Red Hat are the ideal way to deliver cost-effective results. Since Red Hat Ceph Storage is software-defined storage, it may be deployed using less expensive industry-standard drive technology instead of high-priced appliances with integrated storage intelligence.

Red Hat provides the guidance and support you need to put object storage to work with your growing content, archival, or big data assets.