EMC adds de-duplication to disk libraries

By Kevin Komiega

—EMC pulled back the curtain at this week's EMC World user conference to reveal a family of de-duplicating disk libraries, new disk drive spin-down capabilities, and low-power drives for its high-end DL 4000 virtual tape library (VTL), all in an effort to lower the cost of disk-based backup versus tape technologies.

Leading off EMC's product parade are the EMC Disk Library 3D 1500 and 3000, a pair of LAN-based backup-to-disk systems featuring policy-based data de-duplication and IP replication.

The DL3D 1500 provides up to 36TB of capacity, while the DL3D 3000 provides up to 148TB. Both systems use 1TB SATA drives with RAID-6 protection and include optional Fibre Channel ports for SAN connectivity.

Next up is a series of power and cooling-conscious enhancements to the enterprise-class EMC Disk Library 4000. Along with the addition of de-duplication, the DL 4000 now features a disk spin-down option, which powers down idle drives, as well as the latest high-capacity, low-power disk drives.

The drive spin-down option for the DL 4000 series is free and, according to EMC, can save approximately 19% on power and cooling costs by putting idle drives in sleep mode. New and existing DL 4000 customers can use low-power 5,400rpm 1TB SATA drives that require 32% less energy than 7,200 rpm drives. EMC claims that combining the spin- down option with the low-power drives in a DL 4000 will consume up to 47% less energy.

EMC also upgraded its data-protection software with new versions of EMC Avamar Data Store Gen 2 and EMC Avamar 4.0, featuring global source-based data de-duplication and a doubling of backup capacity.

The Avamar Data Store Gen 2 is an integrated product consisting of the Avamar 4.0 source-based data de-duplication backup-and-recovery software running on a pre-configured hardware platform. Avamar Data Store Gen 2 is available in several models, ranging from single node to multi-node configurations. The solution provides twice the capacity per server node of the previous generation. A fully configured Avamar Data Store Gen 2 can provide enough capacity to store the equivalent of 1PB of non-de-duplicated, traditional backup capacity.

In addition, the Avamar software now supports 64-bit Red Hat Enterprise Linux 4 and offers expanded client support for Microsoft SQL Server 2005, Vista, Native Netware client backup and restore, and Novell Storage Servers (NSS) volumes on Novell Open Enterprise Server (OES) SLES 10.

EMC is also making EMC NetWorker available in a new package for medium-sized businesses in the form of NetWorker Fast Start, which is designed to reduce deployment and installation times by decreasing the number of manual steps necessary to configure and implement NetWorker.

EMC is OEMing data de-duplication software from Quantum to expand the amount of backup data users can retain on disk by 90% or more. Less data requires less disk capacity and creates a trickle-down effect of cost savings in the areas of hardware and media management. De-duplication also reduces network bandwidth requirements for disaster recovery by reducing the amount of data moving across the WAN.

All data de-duplication is policy-based and can take place during the backup process or be shut off for a defined time period to speed ingest rates, allowing users to customize de-duplication policies based on application requirements.

Quantum's approach to de-duplication is unique in that the technology gives users the choice of applying de-duplication inline or post-process, depending on performance requirements. Enterprise Strategy Group (ESG) analyst Lauren Whitehouse says this level of flexibility gives EMC the ability to satisfy multiple use cases for the application of the technology, but she says the question of performance is a tricky one.

"The flexibility to choose inline or post-process is a nice option. It allows users to weigh performance vs. capacity requirements and fine-tune their systems," says Whitehouse. "There are a few factors that contribute to how fast or slow data de-duplication operates. In addition to network performance, the rate at which the data can be ingested depends on how the solution identifies duplicates and how that de-duplication process workload is distributed."

According to Whitehouse, ESG Lab validated Quantum's performance claims about a year ago and found that their appliance sustained an impressive 800GB-per-hour aggregate backup throughput.

Pricing for the EMC Disk Library 1500 starts at $115,000, while the higher-capacity Disk Library 3000 is priced from $230,000. Both prices include hardware, software, a three-year warranty, installation, and professional services.

Pricing for a new EMC Disk Library 4000 starts at $200,000 and is based on total capacity.

On the software side, EMC Avamar 4.0 licensing is based on de-duplicated backup capacity. List pricing starts at $17,000 for 1TB of de-duplicated disk capacity.

EMC Avamar Data Store Gen 2 starts at $30,000 per 1TB node. A typical entry-level, single-node system has a list price of approximately $60,000. The multi-node version starts at approximately $150,000.

EMC NetWorker Fast Start is priced at $18,500.

This article was originally published on May 20, 2008