Njini manages unstructured data

By Heidi Biggar

— At the Storage World Conference this week (June 7–9 in Long Beach, CA), Njini introduced the first modules in its Information Asset Management (IAM) software suite. The start-up joins a growing list of vendors that are trying to help users not only better manage growing stockpiles of unstructured data, but also—and perhaps more importantly—extract business value from the data. Potential competitors include vendors such as Arkivio, Kazeon, and StoredIQ (see Kazeon manages unstructured data and Deepfile reinvents itself, narrows focus ).

Explains Brad O'Neill, a senior analyst with the Taneja Group consulting firm, "Njini is a new entrant in an emerging category we call 'information classification and management' [ICM]. Players in this space leverage proactive indexing schemes for metadata and content attributes in order to create a metadata repository, which can then be used for searching and retrieving [as well as for automating a variety of processes]."

The Taneja Group expects ICM to take off in three major areas over the next several years: compliance, security, and information life-cycle management (ILM).

The first module in the IAM suite, njiniENCOUNT, prevents unnecessary duplication of unstructured data objects. Hierarchical storage management (HSM) and compliance tools are expected in the next six to nine months. All IAM modules run on top of the njiniENGINE, which sits in the data path on dedicated Linux or Solaris server blades and automatically categorizes data objects at their point of origin based on the business value of the data (see table).

Njini believes the fact that its software's being in the data path, versus out-of-band, will be a key differentiator, allowing the company to build "proactive" software tools that are capable of taking automated policy-based actions based on information about the data as it is created, not after it reaches the storage device. The software works with any storage device.

"We have a different take on ILM," says Phil Tee, co-founder and chief technology officer at Njini. "We look at unstructured data before it ends up on the storage to see what the content is, who's using the data, etc."

The core engine identifies data at its origin, indexes the metadata, categorizes the data based on its business value, and then manages changes in this value throughout the data life-cycle, eventually storing data objects on the most appropriate storage tier or within an Oracle database or Exchange server, as determined by users.

Tee says it is wrong to base the value of data on its age, as many storage management software applications do today. "For example, a spreadsheet that was created six months ago could all of a sudden become very important if it is needed for a lawsuit," says Tee.

The njiniENCOUNT module helps users reclaim storage resources and consequently drive down storage costs by minimizing the number of duplicate copies of data that are stored.

Pricing depends on the amount of data being managed; specific pricing information was not available at press time. The software is initially targeted at Fortune 1000 organizations.

Information Asset Management Suite

Tool Function Description Availability
ENGINE Core data management layer Automatically identifies data objects (both textual and bit-mapped) at their point of origin, categorizes the data based on its business value, and then manages the changes in this value throughout the data life cycle. June 2005
ENCOUNT De-doping Ensures that the correct number is kept of each data object, according to its business value, which could include replicating or de-duplicating via ancillary modules. June 2005
ENFORCE Compliance Tracks access and lineage of objects through their life cycle, ensuring compliance with corporate retention and deletion policies. Early 2006
ENROLL Dynamic HSM/ILM Decides the correct place to store the data object (e.g., storage tier, Oracle database, Exchange server), according to the rules set by the user. Year-end 2005
ENQUIRE Search/retrieval Aids productivity/efficiency by searching and grouping like data objects, making all relevant information easily accessible. NA

This article was originally published on June 07, 2005