One year after launching its Neoview operational business intelligence suite, Hewlett Packard is taking the concept further, with hardware and software to get closer to reality.
Central to its latest release is technology capable of handling massive volumes of data and a hardware platform geared for fault tolerance and mission critical operations.
HP worldwide lead on Neoview John Miller explains that, on the software side, key components include Skewbuster, adaptive segmentation, publish and subscribe, and workload management.
“These are new to business intelligence,” he says. “They mean that applications, tools and processes can now be extended to knowledge workers, so they can analyse not just historical, but real time trends, using their transactional databases.”
He points out that achieving BI for the masses demands serious mission-critical capabilities, way beyond those required for elite business analysts.
“The workload is different, simply because there are a lot more knowledge workers. So systems must be able to handle much more data, without batch windows and with much higher availability,” says Miller.
“We’re trying to move BI into mainstream business, so it becomes more about delivering decision support in real time – for example, embedding analytics into supply chain applications,” adds Dan Holle, HP’s CTO for BI solutions.
“Conventionally, there are problems with this, because of the high volumes of queries and the complexity of enquiries. So Skewbuster is about balancing workloads across lots of processors; and adaptive segmentation is about tuning enquiries to the right amount of parallelism.”
Miller cites aerospace industrial coatings manufacturer Sermatech as an early adopter, but adds that HP itself is also deploying one of the largest enterprise data warehouses using NeoView, with links to its SAP system, to provide better supply chain visibility, for its own knowledge workers.
|