Intelligent Data Management in a Unified Object Storage for Active Archiving

Continuous data growth remains one of the most urgent problems for most companies. The increasing data volumes must be stored reliably and securely. Archiving too is a particular challenge here: because on the one hand, data must be stored safely and cost-effectively over the long term, but at the same time, seamless and transparent access must remain possible at all times.

Object storage is becoming increasingly popular – with good reasons: Since it is easily scalable, it is predestined for the task of accommodating constantly growing volumes of data. This also makes it particularly suitable for the long-term storage and archiving of large volumes of cold, inactive data.

Object storage systems are mainly hard disk-based. However, increasing data volumes in particular raise the question of storage hardware costs – and of possibilities for optimizing the infrastructure so that it can grow with future data volumes without costs exploding.

This is the starting point of intelligent data management software solutions that identify and transparently move cold and inactive data from expensive primary storage systems to secondary archive storage systems. This concept of tiering focuses on storing data in the optimal location within a multi-tiered storage architecture. This allows users to benefit from the specific properties and advantages of the different storage media.

The optimal solution is a Unified Object Storage that combines different storage classes under a single namespace according to their respective advantages. In order to achieve maximum flexibility for optimization and in the selection of storage hardware, two aspects are crucial:

• Vendor Independence
  The storage software should be designed to support storage systems and technologies regardless of the manufacturer. This allows companies to continue using existing infrastructure and to retain maximum flexibility in selecting new acquisitions.
• Standardization
  To ensure vendor independence and protect investments, the software should rely on standardized interfaces and formats.

Due to the large amounts of data that are typically stored on object storage systems, considerable optimization can be achieved with the integration of a tape storage class. Tape is optimally used for the long-term storage of data that is rarely or never used. Advanced software-defined object storage, for example, also integrates a tape storage class via standardized interfaces and uses intelligent tiering mechanisms to select the optimum storage location for the data respectively.

A tiering policy that monitors the data lifecycle from the creation of the data to its deletion in accordance with the regulatory requirements may then, for example, be as follows:

• “Move data older than X days from disk storage class to tape storage class.”
  Cold, inactive data is moved to more cost-effective secondary or archive storage, and storage space is freed up on the fast, disk-based primary storage systems.
• “Archive the data on the tape storage class for the duration of X months.”
  For archiving, the software provides special functionalities, such as protection against modification or deletion of the data.
• “Delete the data after the expiration of the specified retention period.”
  This is how the data management software ensures compliance with legal and corporate rules for long-term archiving of data.

For the user, the deployment of Unified Object Storage remains completely transparent, i.e. access to archived data is performed seamlessly. In the same way, the Unified Object Storage can retrieve archived data that is used more frequently again back to disk – and after a specified time, transfer it back to tape.

The concept of an active archive for large and constantly increasing data volumes can thus be implemented optimally and with maximum flexibility with a Unified Object Storage that integrates different storage classes under a single namespace and with intelligent tiering mechanisms.