Locating the root cause of problems

The Performance Analytics Engine provides additional analysis and performance diagnostic capabilities to BMC Real End User Experience Monitoring. It helps you locate the root cause of performance problems by using queries, summary views, and investigation of deeper levels of hierarchical performance data.

The Performance Analytics Engine component stores traffic data from multiple sources in a central, off-board storage location. To locate the root cause of performance problems, you can run Performance Analytics queries, which sort and filter data in an intuitive and efficient way (by "zooming in" on categories of most relevance).

• • To locate the root cause of a problem
  • Query information
  • Query details
  • Aggregated results
  • To filter the original query
  • To add expressions to the filter
  • To refine your query with a different set of additional expressions

For information about the role of the Performance Analytics Engine in the overall system, see Architecture.

To locate the root cause of a problem

1. On the menu of the Home page of the Console, click Performance Analytics Query.
2. On the Action menu, click Query.
3. As shown in the following figure, select Page or Object as the Record Type.
4. Select a time range.
5. (Optional) Select an identifier and an operator, edit the value in the Expressions Builder and click Add Expression.

6. (Optional)Add additional filters.

   Note

   The quotation marks (") enclosing the parameters that are strings are mandatory.

7. Click Submit.

Composing a query

The Traffic Browser Query page displays the query in progress (the percentage of data that has been searched) and the aggregated results. The results display the records that match the query criteria and are aggregated according to various metrics.

The Performance Analysis Query page displays the following information:

• Details and progress of the query
• Results of the query, broken down into the Application, User, and Infrastructure categories
• Filtered results in the current query, which helps you to pinpoint the cause of problems experienced by the end user of your application

Query information

As shown in the following figure, the query panel displays details about the query:

Query in progress

It displays the following information:

• Query Summary — A description of the query. It displays the record type, the monitored traffic, and the time range.
   For example, [Page] query on [all traffic] over [Last 60 minutes]. When you hover your mouse over the context, the query ID (an ID that the server assigns the query) is displayed in a tooltip.
• In Progress — The percentage of data that has been searched, based on the time range selected (corresponds to the In progress bar). For example, if the time range selected is 60 minutes, the in progress percentage displays 50% when the first 30 minutes of data has been searched.
• Pages Scanned — The number of pages or objects that has already been scanned for the query
• Pages Found — The number of pages or objects found that meet the query criteria
• Progress bar — The amount of data that has been searched, based on the time range selected (corresponds to the In progress percentage)

Query details

As shown in the following figure, you can view additional information about the query by clicking Details. Click Details again to hide the information.

• Issued on — Date and time the query is created
• Time Range — The interval that the query is searching
• Watchpoint — The name of the Watchpoint that the owner selected as the initial group of data for the query to begin its search (if applicable)
• Filter — The initial expression used to search through the data (if applicable)

Query summary with details

Aggregated results

As shown in the following figure, while the query is in progress, the aggregated results are shown in the following sections:

• Application (section a in the figure)
• User (section b)
• Infrastructure (section c)

Aggregated results sections

Excluding the latency metrics, attributes display only the top nine most frequent values. All the remaining values are aggregated into others.

To filter the original query

While your query is in progress and statistics begin to appear, you can refine your query by filtering the results. Further filtering enables you to localize your issue.

To filter your initial query, click the results that you want the next query to focus on. As shown in the following figure, when you click a result, it briefly displays to the right of the Edit Filter button and a star appears on the Edit Filter button to indicate that the filter has been updated.

Query with filter added

Click Edit Filter and the filter box displays the results that you selected. You can add as many predicates to the filter as needed to focus your next query. After you have all the predicates selected, you can edit them within the filter text box. You might want to edit an operation or the value of the selected result to further refine your query. After the filter is refined, click Restart to abandon the previous query (if the previous query has not completed executing) and start another query with the new filter.

To add expressions to the filter

To narrow the scope of the search by using additional criteria, you must add additional expressions to your current query, as follows:

1. Click all the results that you want the next query to focus on.
2. Click Edit Filter.
   The filter text box displays the results that you selected.
3. (Optional) Edit the filters within the text box.
4. Click Restart.

The previous query is canceled and the new query with the additional expressions begins.

To refine your query with a different set of additional expressions

To refine your query after adding additional expressions, you must remove expressions from the filter.

1. Click Edit Filter.
2. Delete the expressions in the text box that you want to remove from the next query.
3. (Optional) Click a result in the aggregated results to add it to the filter and refine your query.
4. Click Restart.

The previous query is canceled and the new query with the updated additional expressions begins.