February 1, 2017 | Written by: Avi Alkalay, IBM Cloud Advisor
Share this post:
Object storage is a relatively new option for data storage, optimized for general binary or unstructured data, often multimedia. It has gained a lot of importance in the last few years due to the exponential growth of audio, video and images on the web, and thanks to the huge growth of mobility and social networks.
When developers create new apps, they must decide where and how to store data. Structured data (such as name, date, ID, and so on) will still be stored in regular SQL or NoSQL databases.
Images and other binary data that pass through the app will have a better home with object storage, thanks to the following benefits:
- Ease of use. Each object gets a unique ID and a HTTP URL that can be publicly accessible. Data read and write operations are very simple and may be performed directly by user’s browser, via representational stat transfer (REST), without having to go through the control of the server app. The app is released from the rigid structure of database tables and file system hierarchy.
- Scalability. Unlike classical storage using files and tables, the infrastructure to store objects doesn’t grow in complexity when data grows. Object storage can grow quickly, without limits.
- Agility. File systems and databases are complex and require constant care by the sysadmin or database administrator. Thanks to the simplicity of objects, the developer or app owner doesn’t have to depend on these professionals, which eliminates bottlenecks in the app’s evolution. A developer has more freedom to change an app without the help or blessing of the infrastructure team. This agility aspect is what makes object storage so attractive for modern apps.
Object Storage is not a substitute for older storage methods such as file systems or databases. Rather, it complements them with new features.
There are also inaccurate comparisons between object storage and block storage. These are very different things that solve different problems. Objects are used by programmers at the application level, while block storage is a concern of the infrastructure architect. There is no relation between the two. To compare object and block storage is like comparing cars and tires.
Object storage is being used in place of physical tapes and libraries in backup solutions. Maintaining the physical integrity of tapes and robots and transporting them to other locations requires a lot of logistical work.
It is precisely the elimination of these logistics that make object storage-based backup so attractive. You don’t even have to change your backup solution. Object storage can be easily integrated and plugged to what you already have.
Object storage with IBM
IBM Cloud Object Storage focuses on hybrid agility, which means you can have your objects in your own data center, protected by your own firewall, in the public cloud or a mix of both.
In the public cloud, object storage that uses S3 and Swift protocols can be activated in the IBM Bluemix catalog.
Behind the firewall in one’s own data center, there are software options that can be used with an organization’s own, low-cost hardware. There are also integrated options with high-performance hardware.
It also enables transparent hybrid architectures to help organizations find the best cost and benefit balance between their own data centers and public or private clouds.
Here’s a short list of takeaways about object storage:
- It does not replace more traditional methods of storage, but complements them with a lot of agility.
- It can be used to simplify and reduce the cost of existing and new backup solutions.
- IBM Cloud Object Storage helps organizations have consistently integrated data in their own data centers, the public cloud or both.
Learn more about IBM Cloud Object Storage.
Shutterfly has nearly 150 Petabytes of images within its site. Hear how they use IBM Cloud Object Storage to manage—and continue to grow—live at IBM InterConnect.