Use the values from the current plot chart to identify
daily trends in threshold conditions. You can then better anticipate
problems by comparing current values with expected future values or
with past values, and overlaying the chart with situation comparison
criteria, one plotted line for each day specified.
Before you begin
The
Add Historical Baseline tool does not appear in the plot chart
toolbar if any of these conditions exists:
- Historical data collection has not been configured for the queried attribute group and started
on the managed system (or managed systems) assigned to the Navigator item.
- A time span was specified. Click
Specify Time Span for Query and select the
Real time to remove the time span.
- Any of the chart values are the result of a calculation.
-
Attribute(s) across multiple rows is selected. To see if it is, click Edit
Properties in the view's toolbar, Style tab, then the plot area of the thumbnail graphic.
For
historical data beyond the past 24 hours, you must have a
data warehouse configured and the
warehouse proxy agent that transfers data to it.
Procedure
- Click Add Historical Baseline in a plot
chart.
The tool does not display if historical data collection has not be configured for the
attribute group (indicated by the presence of the
Time Span view tool).
- To plot data samples taken yesterday, the same day last
week, or a particular day or past days, select the check box for that
time range.
- Show
yesterday's data
- Draws a plot line of data samples from the previous 24 hours. For example, if it is May 30 at
8:45 a.m., the samplings will be from May 29, starting at 8:45 a.m.. The legend indicates the line
color for each plotted attribute value from Yesterday. As new samples are
added, the oldest are removed from the chart.
- Show last
week's data
- Draws a plot line of data samples from the same day last week. The legend shows the line color
for each plotted attribute value from Last Week. For example, if this is
Wednesday, the samples will be from last Wednesday.
- Show this
day's data
- Draws a plot line of data samples that have been taken on the day shown. If you would like to
see a different day, click the date button to open a calendar. Select a month and day prior to
today. For example, you had selected the check box to show yesterday's data, you could select
the day before yesterday in the calendar to show the past two days.
-
Show data for last __
- Draws a plot line for each of the last number of weekdays or
calendar days you specified. For example, if today is Saturday and you select
the last 7 calendar days, today and all days through last Sunday are plotted. The last 7 weekdays
would be Monday through Friday of this past week plus the Thursday and Friday of the week
before.
- Time Column
- The time stamp type displayed will be shown along the X-axis. If the attribute group has other
time stamp attributes, you can select one from the list. Not all time stamps will be displayed along
the axis because of space limitations.
- Click Apply to render the chart
with a baseline for each chosen time range.
The
time ranges chosen are listed in the legend along with the color of
the plot line, one line for each day.
- Click OK to close the dialog
box, and perform any of these actions:
- Move the mouse pointer over a baseline to highlight
it; move the mouse pointer over a plot point to see the value, name,
and time stamp.
- Click and drag a smaller time range to zoom in, then press Esc to zoom
out, once for each time you zoomed.
- Right-click a data value that falls outside normal statistical
baseline and click Model Situation to open the Model Situation chart and populate the
attributes table with the values from the data point. After performing additional analysis with the
modeling tool, you can create a new situation with the calculated values and edit them as
needed.
What to do next
After you have adjusted the baseline
calculations and the time span to best illustrate the metric you are
watching, you can save the workspace, and then open it whenever you
want to observe the same baselines with current and recent data samples.