Product Description Object Storage
Object Storage — Servercore's service for storing and distributing unlimited amounts of unstructured and semi-structured data in one of the pools.
Object storage can be connected in one of the pools.
User types and roles and projects are supported in the product.
How to work with object storage
It is possible to work with the repository:
- via S3 API, Swift API and Servercore Storage API. You can use various tools (Rclone, AWS CLI, s3cmd and others) to work through the API;
- via the FTP protocol. It uses Swift API to translate requests into object storage;
- through the Servercore Control Panel, which performs queries to object storage via the Swift API and Servercore Storage API. The control panel has limitations for working with a large number of objects and with objects of large size;
- using Terraform.
Tasks to be solved
Object storage is versatile, it can be used for:
- static data distribution (site and application data), data streaming, etc...;
- to store backups and archives. You can set up scheduled server backups to move important data to the vault;
- big data storage for machine learning and analytics;
- hosting static sites.
Principle of operation
The object storage is based on a software-defined distributed data storage system. Data in the storage are stored as objects and replicated to three independent servers located in different racks.
Objects are stored in containers and inherit their settings.
The object store has a flat address structure — there are no folders or hierarchy in the store, so you can quickly access objects over HTTP.
Each object in the repository contains:
- data;
- a unique identifier that is used by the hash function to determine the location of the object;
- metadata. You can add your own metadata to realize your own data storage and processing system.