Object storage support overview

IBM Spectrum Scale™ for object storage combines the benefits of IBM Spectrum Scale with the most widely used open source object store, OpenStack Swift.

Data centers are currently struggling to efficiently and cost-effectively store and manage vast amounts of data. The increasing number of application domains, such as analytics, online transaction processing (OLTP), and high-performance computing (HPC), have created silos of storage within data centers. With each new application, a new storage system can be required, forcing system administrators to become experts in numerous storage management tools.

In addition, the set of applications that includes mobile and web-based applications, archiving, backup, and cloud storage has recently created yet another type of storage system for the system administrator to manage: object storage. Objects cannot be updated after they are created (although they can be replaced, versioned, or removed), and in many cases the objects are accessible in an eventually consistent manner. These types of semantics are well suited for images, videos, text documents, virtual machine (VM) images, and other similar files.

IBM Spectrum Scale for object storage combines the benefits of IBM Spectrum Scale with the most widely used open source object store today, OpenStack Swift. In IBM Spectrum Scale for object storage, data is managed as objects and it can be accessed over the network by using RESTful HTTP-based APIs. This object storage system uses a distributed architecture with no central point of control, providing greater scalability and redundancy. IBM Spectrum Scale for object storage organizes data in the following hierarchy:

1. Account

   An account is used to group or isolate resources. Each object user is part of an account. Object users are mapped to an account and it can access only the objects that reside within the account. Each user needs to be defined with the set of user rights and privileges to perform a specific set of operations on the resources of the account to which it belongs. Users can be part of multiple accounts but it is mandatory that a user must be associated with at least one account. You must create at least one account before adding users. Account contains a list of containers in the object storage. You can also define quota at the account level.

2. Container

   Container contains objects and it lists object in the specified container. It provides a namespace for the objects. You can create any number of containers within an account.

3. Object

   Objects store data. You can create, modify, and delete objects. Accounts have containers, and containers store objects. Containers logically reside within the accounts. So, a container named “Documents” in two different storage accounts are two distinct containers within the cluster. Each object is accessed through a unique URL. The URLs in the API contain the storage account, container name, and object name. You can define quota both at the account and container levels.

IBM Spectrum Scale for object storage also offers the following features.