TOCPREVNEXTINDEX

Lund Performance Solutions


Template Menu and Commands
Chapter 7 describes the options found in the Template menu. See "Menus and Commands." and "Chart Menu and Commands." for information relating to other Performance Gallery Gold menus. This chapter contains the following sections:

Overview

In addition to providing a variety of pre-configured graphs and tables, the Performance Gallery Gold graphics program enables you to customize charts. The Template menu provides the tools to create specialized charts, either by modifying an existing chart template, or by building an entirely new chart template.

Properties

To create custom chart templates, click Properties in the Template menu, or click the Properties toolbar button .

The Template Properties dialog box displays six tabs: General, Graph, Table, Thresholds, Data, and Links; and eight buttons: New, Rename, Apply, Save, Save As..., Delete, Open Chart, Help, and Close (see Figure 7.1).

Clicking in the Validate check box narrows the options displayed in the Template Name drop down menu to only those templates that can be used with the type of data file currently open. For instance, if the current data file is from a Unix system, only Unix-compatible templates will be displayed. Leaving this box unchecked, however, allows you to modify templates for any system without having to open a data file first. This setting is saved between Performance Gallery Gold sessions.
Parameters for new or customized chart templates can be specified in the Properties dialog box, then saved to the tmplates.usr file using the Save and Save As... buttons (see "Save Templates"). Tools found in the Properties dialog box are:

Tabs
Buttons
 


Figure 7.1 Template Properties dialog box

Buttons

New
Clicking this button clears the Template Name drop down window so that a different template may be chosen. If you want to create your own template with a new name, click New, then Save As... to give the new template a name. In order to create a new template, any existing template data should be cleared. If modifying an existing template, select the template to modify in the Template Name drop-down list box.

NOTE If you select New, then begin altering the template without clicking Save As..., then you may lose all of your changes if you accidently choose another template from the drop down menu before saving. We recommend giving your new template a name before making any changes to it.

Rename
This button opens the Set Template Name dialog box. From this dialog box, user-defined templates can be renamed. Performance Gallery Gold templates cannot be renamed, so the Rename button is only active if the chart selected in the Template Name window is a user-defined template.
Apply
Click Apply to apply your changes to any template without closing the Template Properties dialog box. These changes are lost once Performance Gallery Gold is closed unless you opt to save your template changes upon closing Performance Gallery Gold.
Save
Close temporarily saves changes to the template and closes the Template Properties dialog box. If changes were made to a user-defined template, you will be prompted to save these changes again before closing Performance Gallery Gold.
Save As...
If you want to make changes to an existing Performance Gallery Gold template and save it to be used during another session, use the Save As... button to give the template a new name.

NOTE When saving a new or modified template, the customized template isn’t saved in place of an existing template. The template to be saved should be renamed. Modified templates should be saved using the Save As... command.

Delete
Use the Delete button to remove user-defined templates from the drop down menu. Again, because Performance Gallery Gold templates cannot be modified, this button is only active if a user-defined template is selected in the Template Name window. Use this button with caution, as it will not ask you to verify your delete decision.
Open Chart
This button opens the Open Chart dialog box (see "Open Chart"), and allows you to open new charts without closing out of the Template Properties dialog box
Help
Clicking on the Help button from the Template Properties box opens the online context-sensitive help that describes the Template Properties dialog box.

Tabs

General Tab
Under the General tab are three sections: Chart Type, Timeline Type, and Template Warnings.
  • To set the type of chart to be graphical or tabular, select the appropriate chart type in the Chart Type area (see "General Properties").
  • To condense the chart data into a single data point - an average of the eligible data, select Single Point in the Timeline Type area. Continuous is selected by default. For more information on Single Point and Continuous timelines, see "Timeline Type".
  • Any warnings regarding template properties are displayed in this section (see "Template Warnings").
  • Graph Tab
    Under the Graph tab are five options: Labeling, Type, Scale, Stacking, and Marker.
  • The Labeling text box displays the Y axis label of the currently-selected graph (percent, ms, kb, pages, etc.).
  • Select one of the eight graph types found in the Type area (see "Graph Properties").
  • Select one of the three y axis scale options in the Scale area (see "Scale").
  • Check the Enable Data Stacking check box if chart data should be displayed without overlap (applicable to line, area, bar, radar, and area radar charts only) as described in "Stacking".
  • To add a horizontal marker as a reference line, check the corresponding check box, and enter a value in the Marker field (see "Marker").
  • Table Tab
    Under the Table tab are three options: Type, Order, and Options.
  • Select one of the two table formats available in the Table area: Time Indexed and Snap Shot (see "Type").
  • Select the order in which to display the chart data in the Order area (see "Order").
  • The Scroll Snap Shot and Color Coded options control how the data is displayed in the table (see "Options").
  • Thresholds Tab
    Under the Thresholds tab there are two options: Maximum, and Minimum.
  • Set the Maximum threshold and select the Enable Threshold option in the Maximum options section (see "Maximum and Minimum Thresholds").
  • Set the Minimum threshold and select the Enable Threshold option in the Minimum options section (see "Maximum and Minimum Thresholds").
  • Data Tab
    Under the Data tab are six options: Group, Element, "All" group type, Sign, Add Options, and Delete Options. The Data tab also contains a large computation text box.
  • Select a group from the Group drop-down list box. This represents a group for which a particular element exists (see "Group").
  • Select an element of the group from the Element drop-down list box. This represents a particular item in the group (see "Element").
  • To select an entire set of groups (for example, all drives, all processors), select all and the group of to be monitored (such as, "All discs" or "All workloads") in the Group drop-down list box. To set how the data is calculated and displayed, click the appropriate option in the All group type section. Refer to "All group type" for more information.
  • To set the value of the divisor and numerator (see "Add Options" and "Delete Options") as being either negative or positive, select the appropriate sign in the Sign area (see "Sign").
  • To add an entry to the currently selected template, click Entry, located under the Add Options area (see"Add Options").
  • To delete an entry from the currently selected template, click Entry, located under the Delete Options area (see"Delete Options").
  • Links Tab
    Under the Links tab there are two option sets: Link (1-5), and Separators.
  • To make other relevant charts readily available, select the appropriate charts in the Link 1 through Link 5 drop-down list boxes (see "Links 1-5").
  • To place a separator between links in the right-click shortcut menu, check the corresponding Separator check box (see "Separators").
  • General Properties

    The General tab includes three sections: Chart Type, Timeline Type, and Template Warnings. The Chart Type and Timeline Type options control the essential format of the data.

    Figure 7.2 Template Properties dialog box: General tab

    Chart Type

    The Chart Type menu area offers two options for presentation: Graph, and Table.
    Graph
    Graph allows for a graphical presentation of data in several different formats. These formats include Area, Line, Bar, Pie, 3D Bar, 3D Surface, Radar, and Area Radar. These graph types are described in "Graph Properties".
    Table
    Table allows for a tabular presentation format. For more information on tables, refer to "Table Properties".

    Timeline Type

    The Timeline Type area allows data to be displayed as a continuous graph or a single, combined data point.
    Single Point
    A single point graph is displayed as an average. All of the data points that are logged are then averaged into a single point.
    A graph using the Single Point option may look like the example in Figure 7.3.

    Figure 7.3 A single point bar graph
    Continuous
    A continuous graph displays the data points from each collection interval. This kind of graph shows trends and variations over the collection period, rather than the average of all data points of each data set for the entire collection period.

    Figure 7.4 A continuous bar graph

    Template Warnings

    The third section of the General tab displays template warnings. If you have a Performance Gallery Gold template open, you will be alerted that you will not be able to save changes to the template (see Figure 7.1). If you click New to create a new template, all of the following warnings will be displayed until you give your new template a name, links, and data:
  • No template name specified
  • No template links defined
  • This template contains no data
  • If a chosen template lacks defined links (see "Links Properties") or data (see "Data Properties"), then the appropriate message will display in the Template Warnings dialog box.

    Graph Properties

    The Graph tab offers several different options for Labeling, Graph Type, Scale, Stacking, and Marker.

    Figure 7.5 Template Properties dialog box: Graph tab

    Labeling

    The Y Axis text box displays the default Y axis label for the currently selected graph.
    To change the Y axis label:
  • In the Graph tab of the Template Properties dialog box, click in the Labeling Y Axis text box.
  • Type the new label (percent, kb, queue length, pages, etc.).
  • Click Apply. After all modifications have been made, click Close.
  • Graph Type

    Among the possible graph types that can be displayed are the following: Area, Line, Bar, Pie, 3D Bar, 3D Surface, Radar, and Area Radar.
    Area Graphs
    Area graphs are primarily used to plot changes in quantity or position and are useful for illustrating trends. A few examples of how this type of graph my be useful follow:
  • Checking CPU workloads.
  • Comparing between multiple data sets.
  • Comparing storage device performance.
  • Area graphs can be displayed in two separate fashions: Overlapped or Stacked. For more information on stacked graphs, see "Stacking".
    An example of a stacked area graph may look like this:

    Figure 7.6 A stacked area graph
    The CPU Pause data is stacked on top of the CPU Busy data, and all data points are visible.
    An overlapped area graph is usually not used to view graph data due to possible obscuring of data. Line graphs are generally used instead. This is an example of an overlapped area graph:

    Figure 7.7 An overlapped area graph
    Line Graphs
    Line graphs present much the same data as area graphs, however they are not filled. Line graphs may be more practical than other types of graphs when:
  • The data points from different data sets overlap to the extent that individual data points are hidden from view.
  • Viewing only a few data sets.
  • An example of a line graph that is not stacked is shown in Figure 7.8.

    Figure 7.8 An unstacked line graph
    Bar Graphs
    Bar graphs are primarily used to:
  • Compare items at one particular time or over a period of time.
  • View changes in one item over a period of time.
  • Compare portions of a single item.
  • The horizontal and vertical axes represent the two elements being illustrated, such as time and quantity, and can be viewed either stacked or unstacked (see "Stacking"). A bar graph that is stacked may look like the example shown in Figure 7.9.
    .
    Figure 7.9 A stacked continuous bar graph
    Occasionally, there may be a need to view two data sets side-by-side as opposed to stacked. When the Stacking option is disabled, the graph may be viewed in this manner (Figure 7.10).

    Figure 7.10 An overlapped (unstacked) bar graph
    Pie Graphs
    Pie graphs use a circle to represent a whole unit. A "whole unit" would normally represent percentages or a quantitative analysis. Examples of this could be:
  • CPU Utilization by process or workload
  • Comparing Memory Usage to Cache Usage
  • Comparing Disk Pause and Read Hit Percentages
  • Pie charts can only be displayed when using the Single Point Timeline type (see "Timeline Type"). An example of a pie graph is shown in Figure 7.11.

    Figure 7.11 A pie chart
    Radar Graphs
    Radar graphs are line graphs wrapped around a center point. They are most useful when looking at either 12-hour or 24-hour data, as it presents data in such a fashion as to be in a "clock" format. This makes seeing intervals of spiking easier. An example of a radar graph with data stacking enabled is shown in Figure 7.12.

    Figure 7.12 A stacked radar graph
    The figure on the left represents a conceptual model of a radar graph. Each degree of the circle represents a percentage of time (in this case, hours), the outermost circle represents the maximum (in this case, 100), and the center circle represents a halfway point, presented here as 50.0 for clarity.
    A radar graph is essentially a line graph (see "Line Graphs") "wrapped" around a center point.
    Figure 7.13 Figure 7.14 shows an example of a radar graph with stacking disabled.

    Figure 7.14 An unstacked radar graph
    Area Radar Graphs
    Area radar graphs are used in the same fashion as radar graphs. In some circumstances area radar graphs may be easier to read due to the fact that they are filled. Figure 7.15 shows an example of an area radar graph with stacking enabled.

    Figure 7.15 An area radar graph
    Figure 7.16 shows an example of an area radar graph with data stacking disabled.

    Figure 7.16 An unstacked radar graph

    3D Graphs

    Performance Gallery Gold offers three-dimensional (3D) bar and 3D surface graphing capabilities. Three-dimensional graphs have three axes: x, y, and z:
  • The x axis represents the range of the graph. In most cases the range is the time or length of the reporting period divided into logical and equidistant segments (minutes, hours, days).
  • The y axis represents the scale (the quantity of the data over the given reporting period).
  • The z axis represents the categories or groups of data (processors, drives, applications).
  • Figure 7.17 identifies each of the three axis.

    Figure 7.17 X, Y, and Z axes of a three-dimensional graph
    Pre-configured, three-dimensional graphs can be opened from the Open Chart dialog box (see "Open Chart").
    3D Bar Graphs
    3D bar graphs are much the same as a normal bar graph, with one significant exception: it utilizes a third axis. This allows for data to be displayed using three labeled chart axes. Here are some ways that 3D bar graphs may be beneficial:
  • Viewing multiple processor loads over multiple days
  • Viewing Processor Utilization with multiple jobs
  • Monitoring multiple drives over a span of time
  • A 3D bar graph can be configured to display data as a continuous timeline display (see Figure 7.18) or in a single point (see Figure 7.19).

    Figure 7.18 Continuous data displayed in a 3D bar graph

    Figure 7.19 Single point data displayed in a 3D bar graph
    3D Surface Graphs
    3D surface graphs with built-in color coding are used to compare varying levels of activity, for example:
  • CPU Utilization
  • Disk I/O
  • Disk Queue Length by Drive
  • Figure 7.20 shows an example of a 3D surface graph of CPU Utilization.

    Figure 7.20 A 3D surface graph

    Scale

    A graph’s scale (along the y axis) can be tailored to better "fit" its data. There are three scaling options:
  • Percent scaling allows for a graph to be displayed from 0 to 100 of the original scaling. For example: If the graph being viewed has fairly tight increments (for instance, if one point cannot be discerned from the next) scaling could assist in viewing the data.
  • Automatic scaling allows for a graph to be displayed in a scale that fits all data in the graph appropriately (for example, the highest point on the graph is the highest point viewable, while the lowest point on the graph is the lowest point viewable).
  • Thresholds scaling allows for a graph to be displayed in a scale defined by the threshold values located under the Thresholds tab (see "Thresholds Properties").
  • Stacking

    Data elements in an area, line, or bar graph can be "stacked" vertically, which means that the Performance Gallery Gold program will display all of each data element without overlap. Rather than every data set starting over at zero, data sets are added together, one on top of the other, to equal the total. For instance, In the next figure, Disk Utilization by Drive is displayed unstacked. Each of the five drives is represented as its own data set starting at zero and reaching to its individual disk utilization value, in this case, ranging from 1 to 2.5%. The unstacked graph focuses on the individual data sets (drives) and their values.

    .
    Figure 7.21 A bar graph displaying unstacked data sets
    Enabling the stacking option using this same data shifts the focus of the graph from each data set value to how they combine to form the "big picture." In the next example (Figure 7.22), each data set is still separated (by color) and an accurate representation of how each drive is performing is still presented. However, this graph shows that these five data sets combined to utilize almost 10% of available disk space.