Correlate: Pinpoint Probable Cause

Dramatically Reduce the Time to Detect and Resolve IT Issues to Minimize Business Impact

The Aternity FPI Platform preemptively detects performance problems, dynamically isolates impacted applications, processes, and users, and automatically identifies business impact and probable cause.

In order to determine the cause of detected problems, the Aternity FPI Platform uses all of the performance metrics captured, such as the Computer Model, OS Patch Level, Memory, CPU Speed, Installed Applications, Running Processes,  Location, Subnet, DNS, Presentation Server, Web Server, Target IP Server, etc.,  for each frontline user monitored, and indentifies the unique and common attributes of the impacted group – deriving the probable cause of the problem.

Probable Cause Analysis

The Aternity FPI Platform’s Analytic Engine automatically identifies the unique commonalities of the impacted group, via two levels of correlation:

1. Positive Correlation: Identifies the unique attributes the impacted group have in common
2. Negative Correlation: Identifies the attributes common to the impacted group that are also common to the non-impacted groups

The intersection of these two correlations, i.e. the Query Group and the Impact Group is shown as the “Match Group” above. The attributes that produce the strongest Match Group are “surfaced” as a Probable Cause.

For example, suppose that 200 users, who share a common applications server but are distributed across North American and Europe, are experiencing slower than normal response times for Siebel. All 200 users may have the same OS, as do all other Siebel users in the enterprise or 100 of the affected users may have antivirus installed and running as do 300 non-affected users. Therefore, OS and antivirus would not be surfaced as probable causes.

However, each impacted user is connected to the same application server while non-affected users are connected to a different applications server. This common application server would be surfaced as a probable cause. A quick check of system and diagnostic monitors for this server would specifically identify the problem.

When the Aternity FPI Platform performs classification on a detected problem, it scores each of the unique commonalities according to the statistical probability of it explaining the problem. This score is called the grade, and is included in the problem report. In addition, each of the problem correlations is further classified into silos that map to the organizational teams dealing with detected problems, i.e., Desktop, Application, Network, etc. This silo is shown to the left of the classification grade.

As the problem nears resolution, the Aternity FPI Platform continues to monitor the affected End Points, eventually closing the problem when there are no more suffering end users. By having quantitative, empirical evidence about the impact of problems, tickets are closed only when a problem is actually resolved, reducing the guesswork currently involved in understanding resolution impact.

To learn more about Aternity’s Analytic Engine preemptively detects problems and performs probably cause analysis, download the new technical whitepaper on Analytics today.