Performance Intelligence

Performance Intelligence is technology that uses historical trend data to resolve current database performance problems. Ignite uses Performance Intelligence to analyze millions of granular data points, captured in the Performance Data Warehouse along multiple dimensions, and identify which issues have most impact on the IT end user. Then Ignite presents the results in a way that leads the user directly to the most critical problems and the correlated evidence showing the root cause.

Trend Analysis

The first thing Ignite Performance Intelligence shows is the historical trend analysis, covering data captured months, weeks, days or just seconds ago. While many tools focus only on the current situation, Ignite is unique in looking for the repetitive clues that show why database response is slow. The Performance Intelligence engine runs through months of data to identify both the most urgent immediate issue, as well as the most important long term issues impacting database output. And the graphical output makes it easy for anyone, DBAs, managers, or developers, to see the trends and patterns that tell the true performance story.

Correlation

Step 2 of the Performance Intelligence picture is the correlation across the different types of data gathered – Response Time, query statistics, real time sessions, and server health resources. Only Ignite gives the user a way to see how the different indicators correlate at different time periods. Since Ignite Performance Intelligence is unique in its capture of detailed Response Time info, it can illustrate how server memory and I/O bottlenecks, blocked sessions, or specific queries are directly responsible for applications waiting for database response.

Top Problems Analysis

Confio Ignite DBAs have analyzed thousands of databases for our customers using Response Time clues to determine root problems. Performance Intelligence now captures the best analysis experience from our DBAs and automatically runs the same analysis for the user. Ignite Top Problems recommendations are the result of expert DBA rules applied systematically to the full data warehouse of collected statistics. As a result, Ignite tells the user, whether experienced or novice, which problems should be the first priority to improve user response, using the same evaluation rules of our best DBAs.

Alarms

Tying it all together and leading the user through Ignite are the Performance Intelligence alarms. From the main screen on down into the detailed charts, alarms highlight when problems have been identified and their severity. By simply clicking on the alarm at each level, Ignite leads the user to the underlying detected problem. Unlike tools where a user must hunt for a problem in a sea of data, Ignite alarms create a clear path to follow.

Multi-dimensional Cube

The heart of Performance Intelligence analysis is the 25+ dimensional cube structure of the Performance Data Warehouse, which drives all of the analysis techniques described. Ignite captures information from the monitored systems including response time dimensions such as Wait Events/Types, programs, user, blocking session, and objects; plus server resource dimensions such as CPU, memory, and disk I/O. Ignite can then isolate the specific parameters that impact a performance problem. Using the terminology of Business Intelligence, Ignite “slices and dices” the along the multiple dimensions to expose the conditions truly responsible for query bottlenecks. No other database performance solution captures so granular data along so many different dimensions to give this depth of analysis.