Nowadays IT applications run enterprises, almost invariably using data of the structured variety held within databases and flat files. Structured data is much easier to manipulate than unstructured in terms of corporate governance, security, SLAs and compliance policies—since a fixed format means the location of specific information in each record can be known.
The same factors mean information could, in theory, be properly classified according to content. This leaves the way open to better decisions about placement of data within the tiered storage pool, archival and eventual destruction. In other words the full information life cycle (ILM) is covered.
That's the good news. But few enterprise-level data management products focus on managing ‘only’ the structured data-pool, let alone doing this in concert with their applications; yet this makes sense. Storage management vendors commonly focus unselectively on storing, backup, retrieval and archiving all data—both structured and unstructured. Systems management solutions likewise are all-encompassing and, of course, unstructured data needs handling.
However, one company, Princeton Softech, has made a virtue out of focusing primarily on enterprise applications and their structured data. Yet if Princeton Softech and its Optim software (now wholly owned by IBM) only concentrated on storage and archival of structured data storage it would probably have little appeal; it is the application aspect that changes the enterprise data management (EDM) game.
Enterprise infrastructures struggle to meet service level agreement (SLA) criteria, especially for high performance real-time applications. Performance degradation comes from ever-increasing demand, data typically doubling each year and, in an increasingly virtualised environment, greater difficulty pinpointing unexpected faults. Application-focused infrastructure management is gaining credibility, but it needs to be agile to quickly respond to or even predict status or performance changes.
Princeton's Optim wraps up a single software solution to manage enterprise applications and (hence) the structured information they access throughout their lifecycles—true ILM. This means assessing and classifying the information from the outset (although not yet to the finest degree of granularity as standard). From this, business and corporate governance rules can be applied to the stored information to automate the process. This includes applying decisions for storing, retrieving and archiving of data, for ‘subsetting’ of specific database rows and columns, for security protecting the individual data items by applying privacy rules, and finally (when applicable) for destroying the information with proof that this happened.
The software's ability to ‘see’ the stored application data also means, for instance, that a rule covering information identified as purely ‘referential’ (not to be updated) can say that if it is not accessed for 90 days, for instance, it can be moved to lower-cost storage. This reduces power consumption and typically frees up 60–70% of the tier one total storage pool, according to Paul Winn, Princeton Softech's CEO, who added, "This capability results in huge corporate savings alongside performance improvements”. It is not hard to see why.
Absolutely crucial to its viability as a labour-saving management tool is Optim's relational engine; this self-discovers and configures the databases and applications across the enterprise. Optim is also truly heterogeneous, supporting all the major databases plus operating systems which include Windows, Linux and UNIX flavours as well as IBM zSeries mainframes (of especial interest to IBM and its mainframe users); custom applications are also supported.
Mergers and acquisitions (M&A) plus historical application purchase decisions mean that most enterprises use a mix of databases and operating systems. So Optim's heterogeneity to (usually) cover the lot with a single solution is an attractive proposition, setting it apart from key competitors. Winn sees this gap widening in 2008; this is because Princeton is busy adding more advanced support for some widely-used enterprise applications such as SAP and PeopleSoft. This will also help counter an argument that the individual database vendors are always the best ones to deal with management related to their own databases.
Yet without rich functionality and other features Optim might still not have had market pull. One valuable feature is its ability to merge disparate databases, for instance a DB2 database merged with an Oracle one into Oracle format. Reducing the variety and number of databases can be a huge deal, as it simplifies management and maintenance, improves overall performance and reduces licensing costs. Optim's knowledge of what applications and databases are where, and how each will be affected by any changes, also facilitates streamlining database upgrades as well as reducing the risk of faults when changes occur; most system faults result directly or indirectly from infrastructure changes.
I would concur with Winn when he says the software's core capabilities potentially offer an enterprise huge savings, sufficient for the software to perhaps pay for itself within six months. There is also more to come in 2008 such as enhanced content classification with greater granularity.
For now Princeton is being allowed to operate separately from the rest of IBM, which is important to many of its clients who use other infrastructure management software from IBM's major competitors. On the other hand IBM sees Optim as strategic going forward; it is retaining and already using the brand name. So 2008 enhancements include extra IBM-oriented facilities, including increased integration with IBM's Tivoli infrastructure management and an Optim wrapper for IBM's Data Studio due for beta shortly; Winn says this common touch user interface "will be very strategic". (Meanwhile, IBM has separate software to offer to cover unstructured information management).
Optim currently holds a unique position in its EDM market niche—with the capability to make ILM a reality for structured data. Princeton Softech, even before being bought by IBM, was extremely profitable. The IBM factor now gives it extra clout at enterprise board level.
The danger is that, longer term, IBM swallows up Optim into a larger portfolio where it loses focus and momentum—and favour with current users who have competitive infrastructure management software. IBM recognises this. It has actually made the Princeton facility a separate IBM development centre and supports Princeton's rapid expansion plans that include 70% more staff this year. 2008 looks to be a big year for Optim.