CA and IBM are adding predictive analytics to their systems management products to make them more proactive. CA tried this before, with neural networking technology called Neugents, without much success. Back then, there was a disconnect between the technology and customer expectations, says CA chief architect Vince Re. It turned out that many IT problems were random events, such as power failures or software bugs, that Neugents couldn't predict.
What's changed is that new technologies anticipate the need for IT resource allocation and dynamic system partitioning. Unicenter, the CleverPath portal, and other CA products use pattern recognition and other predictive capabilities to monitor IT system and network performance levels.
IBM has built basic predictive analytics technology into several of its Tivoli systems management tools. Tivoli Intelligent Orchestrator has models of a company's data center, resources, and workload capacity. It monitors operations and reconfigures workloads or brings new equipment online when its predictive algorithms conclude that service-level thresholds might be exceeded.
IBM's TotalStorage Productivity Center for Data monitors storage infrastructure and, by analyzing utilization data, predicts bottlenecks. The Tivoli Application Dependency Discovery manager monitors composite applications and issues warnings when changes cause the system to stray from its optimal configuration.
Illustration by Steven Noble
Businesses Mine Data To Predict What Happens Next
Mining For Love In Myriad Places