15 Using the Monitoring Dashboard

This chapter describes the Monitoring Dashboard, which provides views and tools for graphically presenting diagnostic data about servers and applications running on them. The underlying functionality for generating, retrieving, and persisting diagnostic data is provided by the WebLogic Diagnostics Framework (WLDF). The Monitoring Dashboard provides additional tools for presenting that data in charts and graphs.

This chapter includes the following sections:

Running the Monitoring Dashboard

You can launch the Monitoring Dashboard from the WebLogic Server Administration Console, or you can run it separately in a web browser. The Monitoring Dashboard is always displayed in its own tab, or window, depending on the preferences you have set for your browser. You do not need to be logged in to the Administration Console to use the Monitoring Dashboard; but if you are not logged in, you are prompted for your username and password credentials.

For more information, see "Launch the Monitoring Dashboard" in Oracle WebLogic Server Administration Console Online Help.

Scope of the Diagnostic Information Displayed

The diagnostic data displayed by the Monitoring Dashboard consists of run-time MBean attributes with numeric or Boolean values that are useful to measure, either as their current values or as their changes over time. These values, referred to in the Monitoring Dashboard as metrics, originate from one or more run-time MBean instances from one or more servers in the domain.

The Monitoring Dashboard obtains metrics from two sources:

  • Directly from active run-time MBean instances — these metrics are sometimes called polled metrics in this chapter.

  • From the Archive that have been collected by the Harvester — these metrics are also known as collected metrics to distinguish them from metrics whose values are obtained directly from active run-time MBean instances and returned to the Monitoring Dashboard.

About the Monitoring Dashboard Interface

The Monitoring Dashboard has two main panels: the explorer panel and the view display panel, as shown in the following figure.

Figure 15-1 Monitoring Dashboard Panels

Description of Figure 15-1 follows
Description of "Figure 15-1 Monitoring Dashboard Panels"

The explorer panel provides access to the following:

  • View List — Set of existing built-in and custom views. It also contains controls for creating, copying, renaming, and deleting views. For details, see View List.

  • Metric Browser — Provides a means to navigate to and select the specific MBean instance attributes whose metric values you want to display in a chart in a view. For details, see Metric Browser.

View List

To display a view, select it from the View List, shown in Figure 15-2.

Figure 15-2 Built-in and Custom Views Displayed in the View List

Description of Figure 15-2 follows
Description of "Figure 15-2 Built-in and Custom Views Displayed in the View List"

Views are presented in two primary categories:

  • Built-in views

    The Built-in views are a set of predefined views of available run-time metrics for all running WebLogic Server instances in the domain. These views surface some of the more critical run-time WebLogic Server performance metrics and serve as examples of the Monitoring Dashboard's view and charting capabilities.

    Note the following about built-in views:

    • Built-in views are dynamic. For example, if four servers are running, the set of available built-in views and its charts are related to those four servers. If five servers are running, then the set of built-in views and its charts expands for each additional server. In addition, if the number of running server instances changes while you are using dashboard (for example, a server is started or stopped), and you want to see the new built-in views for the current set of running server instances, refresh the view list by selecting Refresh from the View List menu.

    • Built-in views are automatically available with every WebLogic Server installation and can be used by every user logged into Administration Console or Monitoring Dashboard.

    • You cannot modify a built-in view, but you can copy it. Once copied, the view can be modified, renamed, saved, and deleted.

  • Custom views

    A custom view is any view created by a user. Custom views are available only to the user who created them. Custom views are automatically persisted for the user and are in effect only for that user account and only in the current domain. (However, note that polled metric values that are displayed in custom views are not persisted if you close the Monitoring Dashboard window, just as they are not persisted for built-in views either.)

    No custom views are available by default.

For more information, see the following topics in Oracle WebLogic Server Administration Console Online Help:

Metric Browser

Charts display metrics, which are attributes of MBean instances. Metrics can be either of the following:

  • Metrics whose values are obtained from active MBean instances in a running WebLogic Server instance.

    The running server instances are polled at regular intervals, and the charts that display the metric values that are returned are continually updated (see Current Time Range Charts).

  • Collected metrics whose values are obtained from the Archive.

    Collected metrics have been previously captured by the WLDF Harvester and placed in the Archive, and they provide a record of past state. Charts that display only collected metrics are not updated (see Custom Time Range Charts).

You use the Metric Browser to select the metrics that you want to add to a chart. The Metric Browser, shown in Figure 15-3, displays:

  • Currently registered WebLogic MBean types

  • Currently registered instances of MBean types

  • Attributes of the listed registered instances

As a convenience for selecting metrics that have been collected by the Harvester, the Metric Browser includes the Collected Metrics Only button. When you select this button, the Metric Browser displays only collected metrics.

To see metrics for all run-time MBean types regardless of whether instances of them are currently active, select Include All Types. To determine whether a metric was collected by the harvester, select the metric, or leave the mouse positioned over it. A note window is displayed that provides information about the metric, including whether or not it is a collected metric (that is, collected by the harvester).

Figure 15-3 Metric Browser

Description of Figure 15-3 follows
Description of "Figure 15-3 Metric Browser"

To use the Metric Browser, select the server instance containing the metric values you want to display. The Metric Browser can optionally constrain the list of MBean types, registered instances, and metrics that are displayed to only those for which metric data has been collected, or display all MBean types for the server even if they have no active instances.

In addition, you do not need to find a metric by first selecting its MBean type and then the instance in which it exists. You can select a metric in any order; for example, you can start by first selecting a metric, or by first selecting the MBean instance if you prefer. In addition, you can apply filters to each list to further constrain the items that are displayed.

You can select and filter in any order. Selecting an item in one list may make a selection in another and may also constrain other lists. Note the following behavior:

  • Initially the Types list box shows all MBean types (as determined by the settings of the Collected Metrics Only and Include All Types checkboxes), the Instances list box shows all MBean instances, and the Metrics list box shows all metrics.

  • Selecting a specific MBean type causes the MBean instances list to be constrained to instances of that type and the metrics list to be constrained to metrics of that type.

  • Selecting (none) in the Types list specifies that no type is selected, which causes the entries in the Instances and Metrics lists to be unconstrained.

  • Selecting a specific MBean instance, either before or after making any other selection, causes:

    • The corresponding MBean type in the Types list box to become selected.

    • The entries in the Metrics list to become constrained to only those metrics for that MBean instance.

  • Selecting a specific entry in the Metrics list box, either before or after making any other selection, causes:

    • The specific MBean type to which the metric corresponds to become selected in the Types list.

    • The Instances list to be constrained to the MBean instances to which the metric corresponds.

  • When you enter a filter string into any of the list boxes, you constrain the list contents to include only the items that match the filter. The behaviors described in the preceding items that are used in combination with the filter result in a behavior similar to a "logical and."

The effect of these behaviors is to reinforce the relationships among MBean types, MBean instances, and metrics. Each MBean instance is of a specific MBean type, and each metric corresponds to a particular MBean type. The MBean type determines both all the instances of that type as well as all the metrics that the type has.

For information about using the Metric Browser, see the following topics in Oracle WebLogic Server Administration Console Online Help:

View Display Panel

A view is a collection of one or more charts that display captured metric values, as shown in Figure 15-4. Only one view is displayed at a time in the Monitoring Dashboard; however, multiple views can be running simultaneously.

Figure 15-4 View Containing Four Charts

Description of Figure 15-4 follows
Description of "Figure 15-4 View Containing Four Charts"

Each chart in the view contains a legend, labels, and controls for identifying and displaying the data. The following chart styles can be included in a view:

  • Time-series charts, such as a line plot or bar graph that show changes in each metric's value over a period of time

  • Gauges, which show the current or most recent value of a metric along with the following statistics that have been collected for the metric's values:

    • Minimum

    • Maximum

    • Average

    • Standard deviation

Charts can show the metrics for a current time range, meaning that the chart is updated continually as the Monitoring Dashboard obtains new values for the metric at regular intervals. Or, for charts for which you specify a custom time range that has already passed, charts can display collected metrics obtained from the Archive that were captured by the Harvester.

For information about displaying and starting views, and arranging charts in them, see the following topics in Oracle WebLogic Server Administration Console Online Help:

For general details about Monitoring Dashboard charts, see The Parts of a Chart.

Understanding How Metrics Are Collected and Presented

As mentioned in Scope of the Diagnostic Information Displayed, the Monitoring Dashboard displays metrics from two sources:

  • Real-time, polled metric values that are obtained at regular intervals from running WebLogic Server instances and returned to the Monitoring Dashboard.

  • Metrics collected by the Harvester and placed into the Archive.

To view real-time, polled metrics in the Monitoring Dashboard, it is not necessary to configure the Harvester. When a view is started with charts that contain one or more real-time, polled metrics, the run-time MBean instances corresponding to those metrics are polled at each configured interval, and the requested metric values are returned to the Monitoring Dashboard. A polled metric is stored only once in the Monitoring Dashboard, even if that metric has been added to multiple charts or multiple views. The run-time MBean instance corresponding to that metric is also polled only once at each interval, regardless of the number of charts or views in which its metric values are displayed. So when an updated value for a metric arrives in the Monitoring Dashboard, all charts containing that metric are updated simultaneously. This enables the Monitoring Dashboard to minimize the performance overhead on your system and maximize its overall efficiency.

To be able to view collected metrics, you must first configure the Harvester to collect the data you want to monitor and have it available in the Archive. In a view with one or more custom time range charts containing collected metrics, the values for those metrics that correspond to the specific custom time ranges are fetched once from the Archive and displayed in those charts. Note that collected metrics data is also available for programmatic access, and it is written to a standard log, HarvestedDataArchive, which you can view using the standard Administration Console as well as the Monitoring Dashboard. For information about configuring the Harvester to collect metrics, see Chapter 7, "Configuring the Harvester for Metric Collection."

About Metrics and Chart Types

The way in which the Monitoring Dashboard presents metrics depends upon the chart in which they are displayed. After you add a chart to a view, you can use the Chart Properties dialog box to specify either of the following time ranges:

  • Current

  • Custom

The following sections provide key information about how metrics are presented in each chart type.

Current Time Range Charts

This is the default time range for charts in the Monitoring Dashboard. Use this time range for displaying real-time, polled metrics, which can be displayed only in current time range charts. These charts are updated at regular intervals, which by default is every 20 seconds. (The sample interval can be customized in the Dashboard Preferences dialog box.)

When you add a metric to a current time range chart, the Monitoring Dashboard fetches a small number of historical values for that metric from the Archive, if they are available. Note the following about metric values obtained from the Archive for current time range charts:

  • The number of values fetched is derived from the amount of time over which the stored samples can range, in which the sample interval is multiplied by the maximum samples for the chart. (The default sampling interval is 20 seconds and the default sample maximum is 100, which yields a time range of 2000 seconds, or approximately 33.3 minutes.)

  • If the sampling interval used by the Harvester is different from the one configured for the Monitoring Dashboard, some distortion may be evident in the graphing of that metric.

Custom Time Range Charts

Charts configured with a custom time range display collected metrics only. When you specify a custom time range for a chart and add a collected metric, the Monitoring Dashboard fetches the metric's values from the Archive that match the specified time range. These charts are static: once the Monitoring Dashboard displays collected metrics in a custom time range chart, the values of those metrics are never updated.

Note the following:

  • Custom time range charts never include real-time, polled metric values.

  • As a convenience for creating custom time range charts, the Metric Browser includes a button labeled Collected Metrics Only. When you select this button, the Metric Browser displays only collected metrics.

Sequence in which Metrics Data is Displayed

If the Harvester is configured to collected run-time MBean metrics, collection can begin independently of whether the Monitoring Dashboard is running. This section shows the sequence of activity that occurs when the Monitoring Dashboard collects and displays metrics in current time range and custom time range charts.

  1. If the Harvester is configured to collect data for a metric, it starts to harvest that data after the server is started. The data is persisted in the Archive.

  2. When the Monitoring Dashboard is launched, the list of available built-in and custom views is displayed. However, the real-time polling of metric values directly by JMX does not begin until one or more views are started.

  3. When a view containing a current time range chart is started:

    • The Monitoring Dashboard begins polling the run-time MBean instances corresponding to the metrics contained in the chart.

    • If the Harvester has collected data for this metric in the Archive, that data added to the chart immediately. The number of samples that the Monitoring Dashboard obtains from the Archive corresponds to the time range for the chart.

    • If the Harvester was not configured to harvest data for this metric, no historical data is retrieved from the Archive for the metric and therefore none is displayed.

  4. When a view containing a custom time range chart is created, the Monitoring Dashboard fetches from the Archive the set of values for the metric that match the custom time range specified for that chart. Once the values are displayed in the chart, the chart is never updated. The view in which a custom time range chart has been added does not need to be started in order to have the values of its collected metrics displayed.

  5. As polled data values for a metric arrive in the Monitoring Dashboard, the new values are added to the chart. The oldest values obtained from the Archive, if available, are purged.

    The chart always displays the most current data. The maximum samples for a chart determines how many samples can be saved for metrics, in both current and custom time range charts. After a chart reaches its maximum samples threshold, the oldest metric values are removed as newest arrive.

Notes about Metric Data Retention

If you exit from the Monitoring Dashboard, either by closing the Monitoring Dashboard window or by logging out, the browser prompts you to confirm your choice because all metric values captured by the Monitoring Dashboard during the session will be lost. Exiting from the Monitoring Dashboard has no effect on collected metrics persisted in the Archive. However, note that the Archive may have a data retirement policy in effect that limits how long data is retained there. For more information, see Retiring Data from the Archives.

The Parts of a Chart

A chart consists of the following:

  • Chart name

  • Chart viewport, which shows the data values of one or more metrics that are displayed according to the chart type. The type can be a time-series chart that plots individual data points over a specified time span, or a gauge that shows the current or most recent value of a metric along with statistics indicating maximum, minimum, average, and standard deviation values.

  • X- and Y-axes for plotting diagnostic data

    • For time-series charts, data point plots against a time-based X-axis. You can zoom in or out to see a larger or smaller time segment in the viewport.

    • The Y-axis has a range and, by default, the range is automatically set to include all the data points in the chart.You can optionally configure minimum and maximum values for the Y-axis.

  • A legend for each metric that includes the name of the metric and the colored marker symbol that is used for that metric in the chart viewport.

    The metric legend includes a button that, when selected, provides access to operations that can be performed with the metric, such as:

    • Changing the name that is displayed for the metric in the chart, as well as the shape and color used for the metric data points displayed in the chart viewport

    • Copying or moving the metric to another chart, moving the legend within the current chart, or deleting the metric from the chart

  • Chart series overview

    The chart series overview, which is available for time-series charts, indicates the portion of metrics data currently visible in the chart in relation to the whole set of data that has been collected for the corresponding metrics for the represented period of time. You can "drag-select" in either the viewport or the chart series overview to zoom in or out of the chart's data.

    The display of the chart series overview can optionally be suppressed, which can be useful for reducing the number of UI artifacts that are displayed simultaneously in the Monitoring Dashboard and also improving performance on slower systems or browsers.

    For information about customizing the display settings for the chart series overview, see "Set dashboard preferences" in Oracle WebLogic Server Administration Console Online Help.

  • Buttons for panning the and zooming the data displayed on the chart's X-axis. These buttons are part of the chart series overview, so the display properties set for the chart series overview also apply to these buttons.

  • Optional Y-axis units label

  • Chart menu, available by selecting the chart menu button

    You can use the chart menu to add metrics, change the chart type, pan and zoom data shown in the viewport, and set various chart properties.

  • Edit tool

    Select the edit tool to modify the chart name, Y-axis units label, and names used to identify each metric added to the chart.

Figure 15-5 shows each of these parts as they appear in a line plot chart.

Figure 15-5 Parts of a Chart

Description of Figure 15-5 follows
Description of "Figure 15-5 Parts of a Chart"

A gauge chart, shown in Figure 15-6, contains the following additional information about each metric that has been added to it:

  • Minimum and maximum values

  • Average value

  • Standard deviation

Figure 15-6 Data Values Shown in Gauge Chart Types

Description of Figure 15-6 follows
Description of "Figure 15-6 Data Values Shown in Gauge Chart Types"

To display the numeric values indicated by each of these artifacts associated with a particular metric in a gauge chart, position the mouse pointer over that metric's marker symbol, indicated in Table 15-0 by the label Current value.

For information about how to create, modify, and work with charts in the Monitoring Dashboard, see the following topics in Oracle WebLogic Server Administration Console Online Help: