Fujitsu : Releases ETERNUS CD10000 for Optimized Hyper Scale Storage

Fujitsu Limited today announced the release of the FUJITSU Storage ETERNUS CD10000, the ideal storage system for storing hyper scale data volumes, which is now available globally.

The storage capacity of the ETERNUS CD10000 can be scaled out to 56 petabytes just by adding storage nodes, which combine disks with controllers. This makes it possible to meet the storage needs for data volumes that, in recent years, have been growing explosively in the form of documents, images, and video, mainly reference material. Moreover, in addition to ensuring redundancy of data among nodes for high availability, the system automatically distributes data so that processing loads are not concentrated on any given node. As a result, the design work required for redundancy when using conventional RAID() is no longer needed, making it easy to store petabyte-scale volumes of data.

In recent years, the volume of data used online and in data distribution services with documents, images and video, mainly for reference, has been multiplying at nearly exponential rates. To store this massive amount of material, there is a need for storage with high scalability that can keep pace with the explosive growth in data volumes, and that has superior operability to efficiently store large volumes of data.

To meet these needs, Fujitsu is now offering the ETERNUS CD10000, the largest storage capacity system in the FUJITSU Storage ETERNUS lineup. This new system features high scalability and operability, and meets the storage requirements for petabyte-scale volumes of data.

1. Scalable up to 56 petabytes

   The system's capacity and performance can be scaled out by adding storage nodes, which combine disks with the controllers that control data input-output. Up to 224 nodes can be configured, delivering a massive level of storage capacity of up to 56 petabytes. Moreover, nodes can be added without shutting down the system, enabling the capacity of the system to be scaled up in a flexible and timely way as processing levels increase.

2. Superior availability and operability

   The system employs the Ceph() open source distributed file system software, which, in addition to ensuring redundancy of data among nodes, automatically distributes data so that processing loads are not concentrated on any given node. As a result, the design work required for redundancy when using conventional RAID is no longer needed, and high availability is achieved. In addition, with the Ceph distributed file system, the movement of data required when changing the system's configuration is kept to a minimum. Accordingly, compared to ordinary distributed file systems, in which everything needs to be reconfigured, the impact on performance is significantly reduced.

3. Easy migration

   When exchanging ETERNUS CD10000 hardware, just by swapping out old nodes for new nodes in sequence, data is transferred to the new nodes without downtime. When changing the configuration of nodes, the controller automatically restores data redundancy and automatically adjusts the load distributions. As a result, in contrast to conventional storage systems, there is no longer a need to shut down the system for long periods of time or to conduct cumbersome data transfer operations.

4. Flexible access

   This is Fujitsu's first storage product to support the object access method, which assigns a specific ID to content and manages it along with its attribute information. This enables flexibility in content management based on the attribute information, such as the type of data, storage time, and replication policies. In addition, the system supports the block access method, which reads and writes data in block units of a specified size, making it possible to select the access method in accordance with the application.

   Furthermore, it supports the OpenStack open source software, so it can be used as the storage system for private cloud platforms that were built on, and run on, OpenStack.