With the new 4.0 release of Data Palette, Stratavia is striving to break out of the DBA automation niche that it has carved through its product and managed service offerings. The new edition takes more of a process automation focus, repositioning the company to compete in the broader IT process or run book automation space.

Other new features, including predictive analytics and “intelligent” rules engines, are meant to elevate this to more of a closed-loop process. By that, Stratavia claims that it has a more thorough process automation solution that player associated with run book automation, such as Opalis, iConclude (now part of HP/Opsware), and RealOps (now part of BMC).

Of course, Stratavia’s rivals will claim that they have rules engines driving their IT process automation approaches as well. Stratavia claimed that its approach is different in that it is more dynamic. Instead of setting a static threshold that when a specific resource is 80% utilized, the rule will kick in only after pinging the status of other resources which are interdependent.

That admittedly opens a Pandora’s box, because then you get into the dilemma of being faced with infinite permutations when you govern various combinations of servers, OSs, apps, and other elements. That's where another new feature of Data Palette 4.0, the predictive analytics engine, kicks in. Based on preset rules, it predicts when out of threshold conditions will occur based on the interdependencies of various elements and past usage patterns.

All this is based on an embedded metadata engine that acts like, but isn't, a full-fledged CMDB (configuration management database). Stratavia said that its repository will interoperate with CMDBs through a published API.

Stratavia correctly recognizes that the root cause analysis of an outage often boils down to more than just a single exceeded threshold. In fact, one server operating at 90% utilization might prove a weak leak, but not as vulnerable as a server at 80% utilization with an SAP application where a month-end closing of the books is scheduled during the next hour.

But when you start factoring multiple variables, at some point you have to draw the line. We’d expect that future enhancements of Data Palette might include some pattern matching techniques drawn from the word of complex event processing to, paradoxically, cut analysis of the problem down to size.

The new release comes on the heels of the company’s securing $6.25m B series financing last summer. As we reported a year ago, the company is gradually diversifying from its roots as a managed service provider. Its Data Palette tooling was rooted from its MSP days.

According to Mike Puterbaugh, vice president of marketing, the original service provider business still comprises a narrow majority of the business. But the company is transitioning to a products company, with Puterbaugh predicting that next year, products should account for most of the revenue mix.

---