EMC Unveils Celerra Unified Storage Systems

12:10 PM -- Unified storage systems enable the integration of IP network-attached storage (NAS) for file serving with block-based storage applications that use either iSCSI or Fibre Channel. A unified storage system is basically stand-alone (where the storage is included as part of the system), but a gateway (where a choice of backend storage is permitted) can deliver equivalent functionality.

Organizations that are heavy NAS users have found that they can use such systems to consolidate their block-based storage requirements with their NAS requirements, thus gaining all the benefits that entails -- cost savings through better storage utilization, improved manageability, a common management interface, and simplified backup processes. That message has resonated well in the market as both EMC and its chief competitors, NetApp and IBM, have discovered. EMC reports that its NAS Celerra system sales are growing 40 percent year-over-year, which is good in normal times but outstanding considering the current economic environment.

EMC recently announced three new models of its unified storage platform -- the Celerra NS-120, NS-480, and NS-960. Each replaces a model in the previous Celerra generation and doubles the raw capacity of the system it replaces -- raw capacity is expressed in the new model numbers, i.e., 120 TB, 480 TB, and 960 TB, respectively. These new models use the latest generation of CLARiiON technology, the CX4 storage systems, so they inherit all the properties and characteristics that those systems enjoy. CLARiiON is EMC's midrange storage platform brand. The current generation that was introduced last year, the CX4, is based upon a new architecture, which EMC calls its UltraFlex technology. The UltraFlex technology enables the use of dual protocols (FC and iSCSI) and offers greater expandability.

The new Celerra systems target customers with Windows/Linux file server environments that also use Exchange and/or SQL Server. In addition, most of these target environments are likely to use VMware for server virtualization and many are adopting Oracle for database-based applications.

Along with the trio of new systems, EMC announced its NS-G8 Celerra gateway platform, which replaces a previous model and adds NAS and iSCSI capability to an existing Fibre Channel SAN. As such, the gateway connects to a CLARiiON and/or Symmetrix SAN. The target environments are Windows/Linux servers that may also run VMware or Oracle, plus specific functions that can include media, software development, or engineering.With these new solutions, EMC now offers file system de-duplication for Celerra. The approach is different than for backup data de-duplication, which takes advantage of the fact that although many generations of data are preserved over time, much of that data is really the same. As a result, block-based data de-duplication can get rid of the extensive redundancy that is present in all those extra copies of the data.

Data de-duplication on Celerra solves a different problem. Remember, for backup there are many copies of the same data and so getting the redundancy water out is easy. Within a Celerra system, you can only squeeze what is already in the storage box and frankly you can only squeeze so far.

EMC takes a straightforward approach to de-duplicating Celerra file system data. First, it separates active data from inactive data using an internal policy engine (leveraging EMC's Avamar intellectual property) that sets and applies rules for classifying active data as inactive. Active data is not subject to de-duplication since you don't want to have to reverse the process every time that it is used. However, EMC claims that in most file systems as much as 80 percent of the data are good de-duplication candidates.

After the active/inactive classification is complete, EMC applies a file-level data reduction technique called single instancing, which gets rid of redundant file copies. EMC then applies compression to the remaining inactive data using a compression engine based on its RecoverPoint intellectual property. Compression is a standard data reduction technology that has been used for a long time (notably on almost all magnetic tape cartridges), i.e., before data de-duplication was a hot topic. However, though applying compression to disk is uncommon it is a tried and true technology.

It is arguable that block-based data de-duplication reduces greater volumes of redundant data -- for example, in cases where multiple versions of a file differ only by a little -- but it is questionable how much more cost-effective that would be for many or even most companies. EMC states that Celerra's combination of single instancing and compression gives about 50 percent savings in storage. Since the data de-duplication software is free, that seems like a good deal in these cost-conscious times for businesses that want to be able to get additional use of their storage assets.Adding data de-duplication to Celerra is the big Kahuna in this announcement, but there are several other key points to note:

The continued growth in sales for NAS and unified storage systems seems to fly in the face of the difficult economic times, but a closer look reveals compelling cost and manageability benefits. With the new Celerra NS-120, NS-480, and NS-960, and the NS-G8 Gateway, EMC makes the economics even more attractive:

Note that Celerra offers both primary and archiving storage capability. While EMC's Centera is used for heavy-duty archiving requirements (including the long-term retention of information), Celerra is used for more general file system archiving.While predictions are difficult any time and more so in turbulent economic times, EMC is doing everything its power to keep the good times rolling on unified storage. The new products -- the NS-120, NS-480, and NS-960, and NS-G8 -- offer a range of capabilities that deliver extra value to customers.

David Hill is principal of Mesabi Group LLC, which focuses on helping organizations make their complex storage, storage management, and interrelated IT infrastructure decisions easier by making the choices simpler and clearer to understand.