10.4.4 Event and Performance chart tabs area (Performance chart tab selected)

Organization of this subsection

(1) Window configuration
(2) Window description
(3) Supplemental notes

(1) Window configuration

[Figure]

(2) Window description

This window displays performance charts for a monitored target within a monitored service.

The following table lists the items that are displayed:

No.	Item	Description
1	`X` - `Y` - `Z`	This area displays the names of the service group, monitored service, and monitored target within the monitored service whose performance charts you want to view. `X` is the name of the service group, `Y` is the name of the monitored service, and `Z` is the name of the monitored target within the monitored service.
2	Node state display pull-down menu	Use this pull-down menu to select one of the following items as the base for determining the node state display for monitoring items (the default is Event): Event Monitor item state The display status of Configuration information and Graph depends on your selection here.
3	Display interval pull-down menu	Use this pull-down menu to select the interval to be used for aggregating the data to be displayed on the performance charts. The following values can be selected (the default is 1 hour): 1 minute 10 minutes 30 minutes 1 hour 6 hours 1 day Regardless of the value you select, the number of values that will be plotted on each performance chart will be 61. The values calculated from data obtained over the time period you select will be plotted onto a chart divided into 61 equally-sized display intervals.
4	Specify date and time button	Select the data and start time that will be subject to being displayed in the performance chart. Specify the data and time with the Specify date and time dialog box that appears when the Specify data and time button is clicked. You can specify a date and time up to 60 days in the past.
5	Access log button	Click this button to display the Access log area in the Troubleshoot window, where you can check the past status of the selected monitored target within the monitored service. For details about how to check the past status, see 4.4.2 Checking past data.
6	`message`	This area is displayed when you display charts from a list of events. `message` in this area consists of an icon indicating the type of event, the type of event in text, the event details, the monitoring item, and the date and time the event occurred. The following format is used: `icon` `event-type` `:` `detail` `:` `monitoring-item` `YYYY/MM/DD` `hh:mm:ss` (`year/month/date` `hour:minute:second`) If you click or at the left or right end of this area, the message for the preceding or subsequent event, respectively, is displayed. If there is no preceding or subsequent event, the corresponding icon cannot be clicked.
7	Configuration information	This area displays the relationships between monitored services and hosts. When you display performance charts from a list of events, this area displays configuration information for the monitored service in which the selected event occurred. When you select a monitored target from the Services area, this area displays configuration information for the service group or monitored service selected from the Services area. The configuration information that is displayed depends on what you select on the Node state display pull-down menu: Event is selected The information that is displayed is based on events that occurred within the period of time displayed in Graph. Monitor item state is selected The information that is displayed is based on the status of the monitoring item. Priority for the monitoring status is set in the order of Error > Warning > Normal. The monitoring status will be propagated according to the following rules: For the service monitoring status, the monitoring status of the highest level of priority for a monitored target (All Web Access, Web transaction) of service performance in a parent-and-child relationship with the service will be displayed. However, the monitoring status of each host of the system performance will not be propagated to the monitoring status of the service. To configure propagation in such manner, edit the dashboardPropagateSystemStatus property of the system definition file. For details on the system definition file, see 5.6 Editing the system definition files to change settings. For the host monitoring status, the monitoring status of the highest level of priority for a monitored target in a parent-and-child relationship with the host will be displayed. For the monitoring status of a monitored target, the monitoring status of the highest level of priority for a monitoring item in a parent-and-child relationship with the monitored target will be displayed. For the monitoring status of a monitoring item, the monitoring status of events that have occurred for that monitoring item during the currently displayed period will be displayed. If multiple events have occurred, the monitoring status of the highest level of priority will be displayed. If no events have occurred, the normal icon will be displayed. Even when the parentnode is normal, if an error or warning status has been generated with the child node, the status of the child node will be propagated to the parent node. When multiple child nodes belonging to a parent node are in different statuses, the status will be propagated to the parent node in order of those with a higher level of priority. However, a different status will not be propagated to a node in the status where monitoring is being stopped. In cases when an error or warning has been displayed for a monitoring item regarding system performance, the status will be propagated to the monitored service if true has been configured for the dashboardPropagateSystemStatus property of the system definition file (`jp1itslm.properties`) of SLM - Manager. The status will not be propagated to the monitored service if false has been configured. Click for a monitoring item to see a performance chart for that monitoring item. Up to 10 charts can be displayed in the Graph area. To hide a displayed performance chart, click for the monitoring item or click to the right of the performance chart itself. Click for a monitored target associated with a host to launch the PFM - Web Console window in a separate browser.
8	Configuration information legend	You can use the check box next to each legend icon to show or hide the corresponding monitor items in the Configuration information area.
9	Graph	This area displays performance charts for the selected events. The date and time (`YYYY/MM/DD` `hh:mm:ss` (`year/month/date` `hour:minute:second`)) are displayed with the following performance charts: Avg. response (in ms) Throughput (per sec) Error rate (%) If you are linked to Performance Management, you can also display performance charts for system performance. In each performance chart, the following items can be displayed as a line graph: Measured (avg) The average measured values for the period divided into 61 equally-sized display intervals. Measured (max) The maximum measured values for the period divided into 61 equally-sized display intervals. Measured (min) The minimum measured values for the period divided into 61 equally-sized display intervals. Threshold The value set under SLO monitor settings in the Settings window. Baseline The average baseline value for the period divided into 61 equally-sized display intervals. This graph is not displayed unless Error Predict. settings is set in the Settings window. Error Predict. (upper limit) The maximum baseline value for the period divided into 61 equally-sized display intervals. This graph is not displayed unless Error Predict. settings is set in the Settings window. Error Predict. (lower limit) The minimum baseline value for the period divided into 61 equally-sized display intervals. This graph is not displayed unless Error Predict. settings is set in the Settings window. The upper part of the performance chart depends on what you select on the Node state display pull-down menu: Event is selected Events that occurred within the period of time displayed in the performance chart are displayed as icons. Monitor item state is selected Events that occurred within the period of time displayed in the performance chart are displayed as icons, plus the status of the monitoring item is displayed as a band. (error): The band is displayed in red. (warning): The band is displayed in yellow. If your Display interval specification is six hours or one day, events are displayed as one icon per minute. For finer-grained checking of the details of event occurrence, specify one hour or a smaller interval in Display interval. A legend showing the meaning of each graph line is generated and displayed on the right side of a performance chart. Use the Select items to be displayed dialog box to select the values to be plotted and displayed. Click the legend to display the Select items to be displayed dialog box. Select the check boxes for the items whose values you want to have plotted and displayed, and then click the Settings button. The values for Measured (avg), Measured (max), Measured (min), Threshold, and Baseline are displayed by default. Each graph line is displayed based on values aggregated over the time period set in the Display interval pull-down menu. However, the width of the bands indicating timeframes of variation is the same regardless of the display interval.
10	Performance chart details	Details are displayed when you hover the cursor over a graph line on a performance chart. The following items are displayed: Type of values plotted for the graph line Date and time of the value where the cursor is hovered, in the format `YYYY/MM/DD` `hh:mm:ss` (`year/month/date` `hour:minute:second`) Value at the point where the cursor is hovered

Table 10‒3: Items displayed in the Configuration information area
No.	Header	Items displayed
No.	Header	Service performance-related node	System performance-related node
1	Service	Displays the monitored service.
2	Host	Nothing	Displays the host belonging to the monitored service in the parent node.
3	Monitored target	Displays one of the following belonging to the monitored service in the parent node: All Web Access `web-transaction-name`	Displays the monitoring agents belonging to the host in the parent node.
4	Monitor item	Displays the service performance monitoring items belonging to the monitored target in the parent node. The following monitoring items can be displayed: Ave. Response Throughput Error rate	Displays the system performance monitoring items belonging to the monitored target in the parent node.
5	Average^#	Displays for each monitoring item the average value calculated from the data points (maximum of 61) used in the Measured (avg) graph.
6	Maximum^#	Displays for each monitoring item the maximum value calculated from the data points (maximum of 61) used in the Measured (max) graph.
7	Minimum^#	Displays for each monitoring item the minimum value calculated from the data points (maximum of 61) used in the Measured (min) graph.
8	Unit	Displays the unit of measurement used for each monitoring item.

#: If there are no data points for a calculation in the entire range from the left end to the right end of the performance chart, a hyphen (-) is displayed.

Cases where items are not displayed on the performance chart

The performance chart baseline will not be displayed until the number of days since the start of monitoring has reached the Days till start value that was set under Error Predict. settings in the Monitor settings area of the Settings window. For details about the Days till start setting in the Settings window, see 3.2.9 Setting up the monitoring items for service performance.
If there is no information for a specific time period (for example, if there is a period during which no data was stored because monitoring of the relevant monitored service had stopped), the applicable lines on the performance chart will not be displayed. In addition, if monitoring of a monitored service stops before the calculation of a point of variation on the graph, the band indicating the variation point will not appear immediately prior to the stop.

Handling of missing performance data on the performance chart

If there is a period for which no performance data was stored in the database, the immediately preceding status is displayed continuously up to the next recorded event.

Interval displayed in the performance charts

When an event is selected from the event list while the Performance chart tab is displayed in the Troubleshoot window, the performance charts are displayed so that the display interval selected in Display interval is centered at the time the event occurred.
When a monitored target is selected from the Services area or the reload button is clicked while the Performance chart tab is displayed in the Troubleshoot window, the performance charts are displayed so that the current time is positioned at the right end of the chart using the display interval selected in Display interval.
When you specify a date and time in the Specify date and time dialog box, the performance charts are displayed so the time that you specified is centered using the display interval selected in Display interval.

Manipulating performance charts

By dragging a performance chart left and right, you can check the status of the monitored service in time periods before and after the occurrence of the event. With a single drag, you can see an interval that is the same duration as the current display interval. For example, if the display interval is 10 minutes, one drag lets you check a total of 20 minutes.

[Figure]

To Page Top

(3) Supplemental notes

The Performance chart tab cannot be selected unless you have selected a monitored target within a monitored service in the Services area.
Depending on how you display the Troubleshoot window, the Event and Performance chart tabs area will display different events that occurred at different times.
- If you display the Troubleshoot window by clicking the Troubleshoot button at the top of a window, the displayed Troubleshoot window displays a list of the events that had occurred at the time of your login.
- If you display the Troubleshoot window by clicking the Details column for an event in the Events in the last 7 days area of the Home window or in the Event and Performance chart tabs area of the Real-time Monitor window, the displayed Troubleshoot window displays performance charts representing the service performance at the time you clicked. If you then click the Event tab, the Troubleshoot window displays a list of the events that had occurred at the time you clicked the Details column. However, if you click the Details column in the Troubleshoot window, it will not change the time displayed in the Event and Performance chart tabs area.
- If you display the Troubleshoot window by clicking the Troubleshoot button in the Event and Performance chart tabs area of the Real-time Monitor window, the displayed Troubleshoot window displays performance charts representing the service performance at the time you clicked. If you then click the Event tab, the Troubleshoot window displays a list of the events that had occurred at the time you clicked the Troubleshoot button.
- If you redisplay the Troubleshoot window (while the Troubleshoot window is already being displayed) by clicking the (reload) button, which is one of the Troubleshoot buttons at the top of the window, the display will be of a list of events that had occurred at the time you clicked.
If you select another service group, monitored service, or monitored target in the Services area while the Performance chart tab is being displayed, the display will switch automatically to the Event tab.
Performance charts are displayed in the Troubleshoot window based on the results of aggregating past data. However, when the Display interval pull-down menu is set to 1 minute, the performance charts will be displayed based on the most recent data.
If the data update interval for a system performance monitoring item is too long with respect to the value set in the Display interval pull-down menu, the performance charts might be displayed as points. In such a case, set a longer display interval.
The performance chart might not be displayed correctly when data from a version earlier than 10-10 is stored in the database and Monitor item state was selected from Node state display. The following are examples:
- The status of the monitoring item is treated as normal only when no event has been issued in the past.
- Even when the monitoring item has recovered from an overage of a threshold or baseline, its band is not displayed correctly because the monitoring item's previous status (error or warning) is displayed until the end of the display range.
- If the status of the monitoring item changed to a warning error within the display range, that change is displayed correctly, but if the status changed from error to warning, error continues to be displayed.

To Page Top