Monitoring results

After selecting an application or a group in the Analysis menu, the results of monitoring are displayed. The chapters below describe the displayed charts, tables, and controls of the displayed components.

The time frame for which you want to see the monitoring results is selected in the Analysis panel.

Expand the panel by clicking on it or the arrow in the middle:

You can select the required time frame on the right.

You can select a specific time frame or view the results for the last hour, 6 hours, 12 hours, 1 day, 3 days, and so on.

Confirm the selection of the time interval by clicking Search.

The meaning of the displayed APM Index number is described under heading APM index.

You can also use the refresh button (see below) to update the time frame (this makes sense in particular when we are monitoring data for the last hour,for example). You can also find a miniature form of this button in all charts and tables; it has the same functionality.

Charts

A time frame can be selected by mouse in any chart:

The selected frame is displayed in gray. After clicking the **magnifying glass icon with the + sign**, the time frame in the upper panel is automatically set to the selected interval. The new time frame is automatically applied to all other charts and tables.

After hovering over the bottom right corner, you can activate the Logarithmic axis or activate a Stacked chart, where the data series in the chart are stacked on each other. The stacked chart mode is only supported in charts where such stacking is meaningful (for example, in a transaction count chart).

A description of the charts follows.

Performance and transactions

The blue area of the chart corresponds to the current APM index value, which ranges between 0 and 100. The red line shows the transaction count per minute at a given time.

APM index is a number 0–100:

• If the application response time for all transactions is less than the SLA, then the APM index = 100.

• If the application response time for all transactions was longer than four times the specified SLA, then the APM index = 0.

It is therefore a simple universal metric for expressing the overall performance of the application with a number between 0 and 100.

The specific formula to calculate the APM index is:

• L1 = count of transactions with application response time lower than the SLA

• L2 = count of transactions with an application response time between the SLA and 2nd multiple of the SLA

• L3 = count of transactions with application response time between 2nd and 3rd multiple of the SLA

• L4 = count of transactions with application response time between 3rd and 4th multiple of the SLA

• CNT = total transaction count

Counts of transactions at levels L1–L3 can be also seen in chart under heading Transaction count.

Application response time

This chart displays statistical values of the application response time:

• Average – average application response time.

• Median – median application response time.

• 99% percentile – the maximum application response time if we ignore 1% of the slowest transactions.

• 95% percentile – the maximum application response time if we ignore 5% of the slowest transactions.

Number of parallel users

The chart shows the total number of users working in parallel (concurrently).

(The method of calculation depends on the settings of the identification of user session (session id), which is described under heading Session identification of chapter Application).

SLA fulfilment

The chart displays a color-coded transaction count, structured by application response time.

Error count/error codes

These charts display the counts of error return HTTP codes.

The table of transaction occurrences can be filtered by error codes (see heading Transaction instances table). When you click a slice in the pie chart, the context menu opens. When you click Filter this error code in the table, only occurrences with the respective error code will display in the table of transaction occurrences.

Transport size and network transport time

• Network transport time – the average time for which transactions were transferred over the network.

• Transport size sum – the sum of sizes of all transactions.

Tables

Transactions table

The table displays a list of transactions. For example, if transaction = URL, the table displays a list of all URLs that were monitored in the selected interval and the selected application/group. Each transaction is listed in this table only once. Statistical data is displayed in each column.

Additional columns can be displayed using the Show/hide columns button.

If you click Save composition, the table will save currently visible columns and sort is as its default composition.

Click the transaction name to display a pop-up window with transaction details.

Also in this field, you can click a particular field to display its value in a separate window (convenient for viewing and copying long values).

If you click the chart icon in the first column, a new tab opens with results only for this single transaction. If the application has assigned subordinated applications, paired transactions are also displayed (see heading Paired transactions).

If the application has configured subordinate applications, paired transactions from the child applications are also displayed.

You can sort the table data by clicking the column headers. You can also search the table using the Set filter button.

First, select the field by which you want to filter (search) the table and confirm it with the Add filter button.

The = or ~ operators can be selected for text fields.

• = – a test for a full match (all items that fully match the searched expression are found).

• ~ – a test for a partial match (all items that include the searched expression in any place are found).

Confirm or cancel the filter using the Search or the Reset filter buttons.

Another possibility to filter data in a table by a certain value is the context menu. The context menu is opened by right-clicking a cell in the table. There are several options in the menu: Filter same values (=), Filter different values (!=), Filter lower values (<), Filter greater values (>), and Filter similar values (~). You can find a detailed description of these operators earlier in this chapter. These options are displayed based on the type of column. If an option is selected, the respective column filters the selected value.

Show transaction’s instances shows instances of selected transactions in the lower Instances of transactions table.

Transaction instances table

The table displays the history of all user transactions for a selected application or group for the selected time frame.

Additional columns can be displayed using the Show/hide columns button.

If you click Save composition, the table will save currently visible columns and sorting as its default composition.

You can sort the table data by clicking the column headers. You can also search the table using the Set filter button. A format with a mask can also be used in the filter for IP addresses (for example, 192.168.100.0/24 for IPv4 or fc00::0/48 for IPv6).

After clicking on the graph icon, a new tab opens with the details for this type of transaction.

After clicking the transaction row in the table, a new tab opens with the detail of the selected occurrence of the transaction:

You can click on a specific field to show the whole value in a separated window (useful for viewing and copying long values). If the application has assigned subordinated applications, paired transactions are also displayed, see heading Paired transactions.

Sort by columns

You can sort data by one or more columns. By clicking the arrow in the column header you sort the data in one order, by clicking again in the opposite order. If you want to sort data by another column, click the arrow in the column header while holding the SHIFT key.

Save filters

You can save the filtering and sorting configuration for a table can for later. If filtering is configured using the window described under heading Transactions table, such filtering can be subsequently saved using the Save filter button. A filter can be saved under the default generated name or a custom name. If the name of an existing filter is used, this filter will be overwritten. Filters are saved separately for the transactions table and the transaction instances table.

A saved filter can be applied by selecting it from the list which appears after hovering over the Select filter button. Saved filters can be removed by clicking the removal icon in the displayed list.

CSV export

Both tables can be exported to CSV format (for Microsoft Excel) by clicking Export a CSV.

The following dialog appears:

The Delimiter field contains the character used to separate CSV columns. A semicolon is suitable in most cases. If the Add heading toggle is set to YES, a special first row will be added to the CSV containing column titles. If you set Add only visible columns to YES, hidden columns will not be added to the CSV (hidden columns can be set on the right above the table using the Show/hide columns button). The All records toggle allows you to export all records from the table for the selected period; however, if you select this option, the export can take up to several minutes. Otherwise, a limited number of records will be exported (the Number of records). Click Export and wait until the CSV file downloads. Alternatively, you can cancel the export by clicking Cancel.

Paired transactions table

The table displays child transactions paired with selected parent transactions. The Frequency column indicates the percentage of occurrence of the child transaction during the parent transaction. It is the average number computed from the last few days’ statistics. Hit rate 100 means: whenever the parent transaction occurs, there is 1x the child transaction.