Doing More with Less: Predictive Analysis

By Julie Craig

(Back to article)

The past and present drive the future, and IT executives are hyper-aware that 2009 will force change on individuals, corporations and companies large and small. As an industry, we are likely to see a repeat of the flat budgets of the early 2000s, and all eyes will be on hoped for improvements in the economy, which may or may not materialize. At the same time, the show must go on, and business as usual is the order of the day. 


One bright spot on the horizon is a rising potential to reduce support costs with a new generation of intelligent enterprise management products. Nipping at the heels of the technology innovations of the past five years—including the maturing of Web services standards, the introduction of service oriented architecture (SOA) and agile development, and the spread of virtualization—management products are finally starting to catch up. And although we are still seeing just the tip of the iceberg, it’s possible to catch a glimpse of a leaner future. This evolution is multi-faceted, but products are a key player and the promise of automated, autonomic, and self-learning management solutions are starting to materialize.


I’ll discuss these products later, but I want to start with a caveat first—research is showing that, in 2008, IT organizations were actually less proactive and more reactive than they were in 2006. The number of application-related problems detected by users (versus by management products and/or IT personnel) has risen by 10% in two years.


So, it appears IT is between a rock and a hard place. The rock being rapid innovation, and the hard place being the directives of the business. This positioning is unlikely to change. As a matter of fact, given the economic challenges of 2009, it is likely that this paradox will continue to squeeze CIOs in a seemingly inescapable vise.




Putting the rose-colored glasses back on, it is also important to understand that this has been a time of accelerated innovation in the application management space. While systems management products have been available for years, products capable of managing applications—specifically, today’s virtualized, SOA-enabled, standards-based business services—have, in many cases, not kept pace. However, 2008 has seen a spate of intriguing advancements in capabilities that go beyond managing systems to managing “systems of diverse systems,” or applications.


Better yet, these capabilities are timely in that most IT managers will be unable to turn to the old standby, headcount, to solve business service delivery problems in 2009. Since headcount growth will not be an option for many organizations, turning to automation to bridge this gap makes good business sense. Investments in management solutions are among the most cost effective IT-related investments possible in today’s economic climate.


Predictive Analysis


Product announcements from large and small vendors alike offer examples of why this is the case. For example, one of the areas of innovation is predictive analysis. Why is this important? Customer satisfaction, either internal or external, depends on availability and performance of business services. Predictive analysis enables IT organizations to become more proactive in terms of managing these services; in many cases “fixing” application problems before they impact users. And the best way to become proactive is automation.

Some of today’s tools observe and understand the “normal behavior” of an application, and, over time, “learn” the problem profiles that have led to performance and delivery problems in the past. These tools not only anticipate problems, but some can take automated action to avoid them, as well. This is, of course, the Holy Grail of application managers, and many IT organizations would be content even to be able to do this manually. However, today’s applications drive such high transaction through-put; are so mission-critical; revenue-intensive; and complex that automated tools can actually manage them much more efficiently than highly-skilled personnel. People just aren’t fast enough to correlate information from multiple sources in the same way that automation can.


Vendor Announcements


Recently, IBM announced the introduction of predictive analytics across the Tivoli stack, including base lining, dynamic thresholding, predictive alerting and automated actions. BMC’s ProactiveNet Analytics solution has similar capabilities, and products from ASG, CA Wily, Aternity, Symphoniq, HP, Nastel, Integrien, and OpNet have predictive capabilities, as well. Such products could go a long way towards reducing the gap between proactive and reactive IT, as well as in reducing the diagnostic and remediation backlogs of overburdened support personnel.


OpTier introduced an interesting variation on this theme with its CoreFirst 3.0 announcement earlier this Fall. While the product has done predictive alerting and remediation for some time, CoreFirst 3.0 introduced predictive analysis of changes to the IT infrastructure. In other words, the product predicts the effect of a change on a transaction service level agreement (SLA), notifying IT personnel whether a given change had a positive or negative effect on the business service. Since between 25% and 75% of infrastructure changes drive application-related problems (depending on the maturity level of IT change processes), this capability can potentially eliminate a hefty portion of IT’s support load.


Interest in configuration management and its close cousin, application discovery and dependency mapping (ADM) continues to rise. To that end, Tideway made some interesting announcements recently. In addition to discovering and mapping business applications “on the fly,” as they execute across diverse infrastructure, Tideway now offers “hardware reference data”—information on the rack space consumption, heat output and power consumption of discovered assets. This information can be used to compute an applica­tion’s overall carbon footprint, helping to automate data center rationalization and cost optimization assessments.


At the same time, Tideway introduced a “Dynamic Discovery Extension” that adds exploration and documentation of relational database management systems utilization—data utilization and interactions—to application maps. Both configuration management and ADM solutions simplify the process of diagnosing run-time problems. Such products will gain momentum over the next few years.

EMA is in the process of publishing a paper titled, A New Paradigm for End-to-End Application Management: Products, Processes, and Where to Start, detailing these products as well as products from vendors including Compuware, A10 Networks, Citrix, Coradiant, iRise, and NetScout. Virtually all of these solutions feature capabilities that translate to “doing more with less”. Product profiles will also be published online in early Q1 of 2009, and can be viewed free of charge at EMA’s site (www.emausa.com).

The challenges of the new year will likely separate the wheat from the chaff, as far as strong and weak companies are concerned. A key element of a company’s strength lies in its ability to effectively strategize and execute through lean times. Hard times shouldn’t mean that companies tread water in anticipation of better times to come. Instead, smart investments can drive significant efficiencies, and investments in today’s “smart” management solutions will do just that.

Julie Craig is a senior analyst with Boulder, Colo.-based Enterprise Management Associates (www.enterprisemanagement.com), an industry research firm focused on IT management. Julie can reached at jcraig@enterprisemanagement.com