Hitachi claims virtualised storage performance boost

HDS has increased the performance of its high-end storage with what it claims is a breakthrough in virtualisation

Hitachi Data Systems (HDS) is claiming a 25 percent performance boost to its high-end TagmaStore universal storage platform and says it is now capable of delivering up to 2.5 million input/output operations per second (IOPS).

According to the company, this latest version of its Universal Storage Platform (USP) has an embedded virtualisation layer "capable of managing up to 32 petabytes of internal and external storage". This is considerably beyond the USP's past capacity.

The USP also has impressive multi-protocol consolidation, and supports iSCSI, Fibre Channel (FC), ESCON, FICON and NAS within a single storage services platform.

David Floyer, chief technology officer of ITCentrix, an independent software and services firm, has conducted testing on the USP. "The USP significantly reduces the total cost of storage. Our analysis of the enhanced USP shows that in a large configuration of 200 terabytes with 800TB of managed storage, Hitachi is projected to reduce the cost of storage and storage management by as much as 52 percent, compared with competitive solutions."

In order to get the maximum performance level of 2.5 million IOPS, existing users of TagmaStore need to "upgrade to the latest USP microcode level and new 4GB/s Fibre Channel adapters," a representative of HDS told ZDNet UK. "A 32-port 4GB/s Fibre Channel board is approximately $100,000 (£54,000)."

The USP has software updates including enhancements to the HDS Universal Replicator for out-of-region operations that now include support for 64K data volumes distributed across up to four USP systems.

The Delta Resynch capability, which allows a customer to recover to a local (or intermediate) site when the primary site has failed in a three data centre configuration, now will copy only the delta data (data that has changed) from the intermediate site to the remote disaster recovery site during recovery instead of copying all the data. This should speed up recovery.

The ShadowImage In-System Replication facility, which allows data replication within any HDS storage system, has been improved to support greater data volumes, while data replication performance has been increased by up to 300 percent to more than 1GB/s, according to HDS.