Comparison of File-, Block- and Object storage
Traditional File- and Block Storage have now been complemented with Object Storage, and we get questions when to use what and why. We will in this article try to answer that.
The Goldilocks of storage. Good for you excel and word documents.
A file is typically stored in a file system with a hierarchal organized structure, directories. You will be able to find the file by describing the path to the file and metadata will be stored locally together with the data. A great way to store file are on a NAS device, Network Attached Storage, attached on your LAN. Watch out for data growth, managing too many NAS devices can be complex and using NAS in a WAN is not optimal. File storage is fast when used in its right environment.
Small and fast, perfect for databases and applications.
Block Storage is similar to File storage, but working with a chunk of data, which makes its a block. The block has an address and applications retrieve a block by making a call to the address. With block storage there is no other metadata, it´s up to the application to combine and access the data. Under the right circumstances, block storage extracts the best performance from any given storage, but you need to think about the latency. At long distances, block storage will face latency problems. Using it for the right application, block storage has been the best choice for performance-centric applications, e.g. databases with great performance and almost no metadata.
Block storage is the building block for file storage.
Slow but price aggressive and very scalable, perfect for data that is not frequently accessed.
An object is defined as data, typically a file with all its metadata. All objects are given an ID, much like a URL. this ID is what the application is asking for when retrieving the data. Object storage is a large storage-pool with flat structure and may be geographically separated. A large benefit of object storage is that you can define metadata in a way not possible before which makes analytics, data protection, scalability and managing extremely good. Object storage was invented because of growing data volumes, inside the concept of web 2.0, the target was to store data at a low cost. The simplicity is an advantage tougher with the low cost, the only drawback is performance, of the three this one is the slowest.
To find out more about our solutions for storage, please visit: Cristie Data Storage >
Hope this article answered some of your questions!