Cloud server storage is a service model in which the digital data is stored in logical pools across multiple virtual servers. It is normally located in a data center managed by a Cloud provider. Data can be easily accessed by the user from any location, any device, and at any time.
Cloud storage thus houses the customers’ data by transferring it via the internet to the Cloud server which stores the information and retrieves as per the requirement through a web-based interface. The customer can access, manipulate or delete the data as per his requirement similar to a hard disk. The Cloud is a virtual storage basically.
Storing multiple copies of same data on different servers facilitate redundancy. This ensures that the customer can retrieve the data even at the time of any Cloud upgrade or even if an individual system crashes.
General architecture of cloud server storage
Cloud Storage architecture primarily consists of a front-end, a middleware and a back-end storage panel. The front-end of the Cloud architecture exports an API (Application Programming Interface) to access the storage. This storage logic middle layer takes care of a variety of features like data de-duplication and distribution/replication of data across the various geographical regions. Finally, the backend layer implements the physical storage required for data.
Cloud data storage is normally built on object-storage architecture. The APIs used here is on the basis of REST principles. A Restful API uses HTTP requests to manage stored data (i.e. to carry out operations like GET, PUT, POST, and DELETE data). Using this Restful API, an object storage protocol is triggered. An object storage protocol maintains the blocks of data that make a file together.
It adds all its associated metadata (A metadata includes the basic information such as author, date created, date modified, and file size which makes it easier to locate a specific document/file) to that file and stores it as a single object. This eliminates the need for a hierarchical data structure and places everything into a flat address space called the storage pool. A unique identifier is then assigned to this object. When this content needs to be retrieved, the user presents this unique identifier to the system and the content is generated with all its metadata, authentication and security.
For the sake of protection and reliability, data is stored in the object storage method in two ways namely Replication and Erasure Coding method. In Replication method, data is copied to x number of nodes (a group of servers) with multiple levels of sophistication on where the different nodes might be physically located. In Erasure Coding method, an object is segmented into parts and then those parts are distributed across multiple discreet nodes.
Types of Cloud server data storage
Public Cloud Storage: In Public cloud server storage, data is stored in global data centers with storage data distributed across the different geographic regions. Here a third-party provider makes computing resources available to the public over the internet.
Public Cloud Storage Service generally provides a multi-tenant storage environment. A single instance of a software application serves several customers. This is the principle used in multi-tenancy storage architecture. Each customer is referred to as a tenant. Tenants may be given the access to customize some parts of the application such as the color of the User Interface, but they cannot customize the application’s code. A SaaS provider runs one instance of its application on one instance of a database and provides web access to several customers. In such cases, each tenant’s data will remain isolated and invisible to other tenants.
Multi-tenancy storage architecture is mostly suited for unstructured data (textual such as email messages, PowerPoint presentations, Word documents, instant messages, and non-textual such as JPEG images, MP3 audio files, flash video files, etc.). Examples include Amazon Simple Storage Service (S3), Amazon Glazier for cold storage (storing of inactive data for backup, archival or disaster recovery purposes), Google Cloud Storage, Google Cloud Storage Nearline for cold data, Microsoft Azure, and so on.
Private Cloud Storage: Private cloud storage services or on-premises storage services provide a dedicated environment to the requirements/goals of a single organization. Here an organization uses a proprietary architecture that runs on Cloud servers within its own data centers. It is apt for customers who need customization and more control over the data. Examples include HPE, VMware, DELL EMC, IBM, Red Hat, Microsoft, OpenStack, etc.
Hybrid Cloud Storage: Hybrid Cloud Data Storage is a combination of a private and public cloud storage service with orchestration (A programming technology developed to manage the interconnections between private and public workloads) between different platforms. Hybrid Cloud storage has become most commonly used storage type in recent times. An organization stores the actively used and critical structured data in a Private Cloud and unstructured, archival data in Public Cloud.