By Dennis Drogseth
As an industry analyst, one of the pleasures of doing research is that you get paid to learn. If you can pick a topic that’s changing rapidly while transforming the industry in the process, then the pleasure is all the greater—both for you and for your clients and sponsors.
The research Enterprise Management Associates (EMA) recently conducted earlier this year on advanced IT analytics (AIA)—often referred to as IT operations analytics—seemed to hit a homerun in terms of being both a changing environment, as well as one that’s industry transformative.
I’d like to share just a few of the more salient highlights, but first let’s examine the premise for our research a little more closely.
The requirements-driven definition for AIA in our research included that it be cross-domain and not restricted to just network, systems, application, database, or even business outcomes; it should require the assimilation of many different sources, whether third-party monitoring tools or different data sources such as log files, events, time series data, or spreadsheets; and it requires the application of advanced heuristics such as machine learning, advanced correlation, anomaly detection, and predictive trending.
Beyond this, we let the data speak for itself.
We used the term AIA because we saw a growing interest in having it provide a layer of commonality for decision-making far beyond operations. In fact when we looked at the data on role support, we saw that our average respondent targeted four domain-specific roles, four cross-domain roles, and three business stakeholder roles for a single AIA initiative. We also saw strong leadership from IT service management teams in providing governance, workflow, and leadership in working with operations in support of AIA.
Targeted Use Cases
For this research, we targeted two key use case areas—requiring active deployments in at least one of two options, performance management or change management and capacity optimization.
We found that 100 percent of our respondents were doing performance management and only 60 percent were doing change and/or capacity optimization. In other words, and perhaps not surprisingly, ‘performance management’ turned out to be the pervasive place to start.
I should mention that we also saw strong requirements for integrated security management, including top-of-the-list priorities for security in triage and diagnostics; data sources—security event and information management or SIEM was a number one priority; domain role support; and AIA enablement in support of the move to cloud. All this was especially striking since only six percent of our respondents were dedicated security professionals.
AIA Technology Design
In terms of AIA technology design, the average respondent wanted five different types of data sources—from SIEM to log files to spreadsheets to events, nearly four different types of advanced heuristics—process analytics, anomaly detection, predictive trending, and data mining led; and the ability to assimilate on average 15 different third-party monitoring or other sources. Perhaps not surprisingly, 90 percent of respondents were aggregating at least some AIA-related data into a central data store.
We also saw the service modeling interdependencies and AIA go hand in hand—a priority for 96 percent of our respondents. Application discovery and dependency mapping for performance led in prioritized sources, followed by a service-modeling dashboard, a topology built into the analytic tool, and a federated configuration management systems. All these options were close—separated by only seven percentage points.
Benefits and Obstacles
The average respondent had achieved four unique benefits. The top five were more efficient use of cloud resources, more efficient use of storage, faster time to remediate problems, faster time to deliver new IT services, and improved operational efficiencies.
Interestingly, all but five percent of the respondents were involved with cloud and 69 percent had linkages between AIA and DevOps or agile initiatives.
As for obstacles, the number one concern was technology—the products are not fully baked yet. Data quality issues were tied for second place with internal politics/leadership and the availability of relevant skill sets.
But perhaps the most significant finding was conformation that AIA is a strategic rather than purely tactical investment, most often led by the executive suite. This is because optimizing AIA requires a fresh approach to how IT organizations work together. In other words, the move to AIA is not purely about technology, but also about process and organization.
These are just a few of the many highlights from data EMA collected in quarter one of this year from 150 respondents in North America and 100 respondents in Europe—including France, Germany, and the U.K. To learn or hear more, access a Webinar on the topic through research.enterprisemanagement.com. SW
Jul2016, Software Magazine