Concepts

Cloud storage is a scalable solution for storing and distributing unstructured data of any size and format. As an object storage solution, it is suitable for storing large datasets with robust data integrity mechanisms. Its key characteristics are described below.

Object storage

Cloud storage, as an object storage solution, includes the following key features:

• Automatically adjusts storage capacity and resources to handle growing workloads

• Uses metadata to manage data independently, as object storage does not provide direct access to the operating system

• Supports distributed storage, ensures data integrity, and reduces complexity in managing large datasets

Data structure

The organization of data in object storage includes the following key features:

• Organized with unique identifiers and metadata, simplifying data management and enabling instant access regardless of location

• Objects are stored in a flat structure without the nesting that is typical of hierarchical file systems. However, graphical interfaces may simulate hierarchies using path-like identifiers, such as folder1/file.txt, to represent objects within folders

• Triple data replication across three different servers ensures high reliability

• Specialized mechanisms detect and resolve discrepancies among object copies, maintaining data consistency

• Multipart upload is supported for objects over 100 MB, enabling efficient independent uploading of segments, better utilization of network resources, and resumption of interrupted uploads

Logical organization

Objects are grouped into containers (also known as buckets), which act as independent logical structures. Containers provide the following capabilities for managing data and metadata:

• Containers can be configured with access control settings to be either private or public.

• Containers store both system and user-defined metadata describing objects, their properties, and settings, such as deletion policies or version control

• Containers support functionalities such as restoring deleted files, regional synchronization, object versioning, and partial object updates

• Containers can be deployed in various regions to optimize access speed and performance based on user location

Access and managing cloud storage objects

Cloud storage objects have a direct integration with web applications and can be managed in various ways:

• Using the customer portal

• Using APIs, such as S3 or OpenStack Swift

• Via third-party applications that support Swift (e.g. Cyberduck)

• File Transfer Protocol (FTP)

Cloud storage containers (buckets) provide two types of access:

• Private access restricts container contents from being accessible via the internet. Commonly used for backups, confidential data storage, and other tasks requiring limited access

• Public access makes container contents accessible over the internet through public links via HTTP or HTTPS protocols. Ideal for hosting and distributing files, such as static web content (e.g., images, documents, videos) or publicly shared materials

For more detailed information on managing objects and configuration, refer to the article How to access data in cloud storage.

Additional features

Here is a list of additional features you can apply to your cloud storage:

• Versions storage allows you to designate a container for storing old copies of modified objects

• Custom domain enables the association of a custom domain with a container, allowing the use of custom URLs for publicly shared materials

• Container synchronization ensures objects are synchronized across containers in different regions, maintaining data consistency

Suggested Articles