It’s the storage format used by direct-attached storage (DAS) and network-attached storage (NAS) systems.
That means even if your storage system is a hybrid of cloud service providers, Linux servers, and Windows servers, your SAN will be able to quickly reassemble and retrieve your data when you need it.įile Storage is the storage format most people are familiar with-data is stored in files you can interact with in folders within a hierarchical file directory. Block storage decouples data from its storage environment, allowing the storage area network to store data where it's most convenient regardless of the underlying operating system. Block storage allows the underlying storage system to retrieve it no matter where it gets stored. Object storage supports HTTP and REST, the application programming interface (API) architecture used by most websites and software-as-a-service (SaaS) apps.īlock Storage breaks data up into separate pieces of fixed-sized blocks of data that each get a unique identifier. It’s similar to a photograph, for example, which might contain metadata describing the camera settings used to take a picture along with the time and place it was taken.
Objects function as modular units, each acting as self-contained repositories complete with metadata that describes details like permissions, privacies, securities, contingencies, and other information. Each unit has a unique identifier or key, which allows them to be found no matter where they’re stored on a distributed system. Object Storage is a storage format in which data is stored in discrete units called objects. Let’s take a look at the main differences between them: In the world of data storage, there are three main approaches to storing your data: file, block, and object. Flat, API-friendly, and highly scalable, object storage is the format of choice for public cloud storage services like Amazon S3 and on-premises solutions like Pure Storage® FlashBlade®. Each object contains a key, data, and optional metadata.
Some of the better known are the SNIA’s Cloud Data Management Interface (CDMI) and Amazon’s S3 that operate through the HTTP protocol this makes object storage attractive for Web and cloud use.įor more information, we recommend that you take a look at the informative SNIA webcast, “ Object Storage: What, How and Why.Object storage is a data storage architecture in which data is stored and managed as self-contained units called objects. Regardless of how they are stored, all object storage systems support simple APIs to create, read and delete objects, and to search meta-data. Many systems use file or block-based storage some are specifically designed with devices that directly support objects. How objects are stored is not important to the application. They can be simple sets of data, files, entire file systems, videos, virtual machines or containers, databases, the list is endless, since the storage system simply sees and manages the object as an object ID that is associated with a chunk of data. What objects contain is not important to the storage system. These permit collections of objects, individually addressable by their object ID, to be searched, filtered and read in groups without needing to know the specific object IDs.
It is often expressed as an attribute-value pair for instance, an attribute of COLOR in our collection of objects may have the value RED for some objects and BLUE for others. This is data that is part of the object, but that is in addition to the object ID and the data. Most object storage allows for objects to be deleted.
There is often no capability to update to the object or parts of the object the entire object is usually required to be re-written. When reading from object storage, you can read either the whole object, or ask to read parts of it. When you create an object on this type of storage, the entire set of data is handled and processed without regard to what sub-parts it may have. There is no structure corresponding to a hierarchy of directories in a file system each object is uniquely identified in the system by a unique object identifier. It applies to any forms of data that can be wrapped up and managed as an object. Object Storage is a method of storing and subsequently retrieving sets of data as collections of single, uniquely identifiable indivisible items or objects.