Using log outlier analytics

Log outlier analysis helps resolve production issues quickly by summarizing large volumes of individual log events into a smaller set of broad event types. This enables you to focus on the most important logs for troubleshooting. Log outlier analysis is available on any log widget on Tanzu Platform.

Verbose logging environments generate an overwhelming amount of data, which makes it difficult to pinpoint the root cause of an issue. Log outlier analysis simplifies the process of identifying the root cause by automatically grouping similar log events into categories. Thus, allowing you to focus on critical logs rather than sifting through endless entries.

Summarizing log events provides the following benefits:

• Highlights the most frequent and the most rare events by automatically grouping logs by their signature
• Identifies key patterns in log data, thus helping you identify the root cause of issues faster.

View log outliers

You can view log outliers in the Logs tab for applications and infrastructure. You can also view log outliers on any runbook.

To view log outliers on the Logs tab:

1. On Tanzu Platform hub, select the Logs tab for a Cloud Foundry foundation or an application.

   • Go to to Infrastructure > Foundations > click on a Cloud Foundry foundation.
   • Or, go to Applications > Apps and Microservices > click on an application.
2. Go to Logs > Analytics.

3. On the top right, select a time window from the drop-down menu.

   The analytics data is presented in the form of a table. You can sort the table by Count or by Trend. When you sort by Count, the rows are sorted by highest Count to the lowest count. When you sort by Trend, the rows are sorted by the pattern that has a higher rate occurrences to the lowest rate of occurrences.

   Item	Description
   Count	Number of occurrences of the pattern for a specified time.
   Trend	Graphical representation of the count across the period of time.
   Pattern	Sample message that represents a group of similar log messages. To filter out a pattern, click on the three vertical dots adjacent to a row and select Events not like this.