The ECE aggregate statistics program is an expert tool for finding errors within the end-to-end Monitoring and Alerting Infrastructure of SAP Solution Manager. The program displays the aggregates for events and metrics that are stored in relevant table (MES_DB_AGGREGATE
) of the Event Calculation Engine
(ECE).
For example, an aggregate of an event includes all instances of the event that have the same event type, the same managed object, and, for successive runs of the Event Calculation Engine, the same valuation (that is, the same color: red, yellow, green, or gray). An aggregate is therefore characterized by the following data (its key):
Managed object, such as an instance of a managed system
Metric type or event type, such as the dialog response time
Metric path; in metric groups, this distinguishes between the different metrics of a managed object.
Example
An example of a metric group is the file system, since for a managed object (in this case, the host), there are usually several file systems (File System C:
or File System D:
) that are monitored. The variable part of the metric name (that is, C:
or D:
) is therefore the metric path; this is also technically uniquely identified by an ID.
Use this program primarily to check the following points:
Metrics or events with a large number of aggregates; a large number of aggregates is primarily created if the color of the corresponding metric or event frequently changes, that is, if it is flickering between colors. This is disadvantageous, since the aggregates form the basis of the data that is displayed in the Alert Inbox, and this therefore makes working with the Alert Inbox increasingly complex.
Metrics or events for which values are no longer delivered.
Number and distribution of aggregates to the different monitoring use cases.
Caution
Depending on the number of aggregates in the database, executing this program can take several minutes.
To call the aggregate statistics tools, call the transaction MAI_TOOLS. Go to expert mode, and, in the Analysis
tab, choose Aggregate Statistics
.
When you call the program, the aggregate statistics are displayed in an ABAP list. The first two tables display the number of aggregates, broken down by data type or monitoring use case. The table below this displays information about the aggregates in accordance with the keys described above:
Column Title | Description |
---|---|
| ID of the managed object (MO) |
| Type of the managed object. There are different MO types for each monitoring use case. For example, in system monitoring, there are the MO types Note This column, like the columns End of the note. |
| Name of the managed object; these follow a particular naming convention, depending on the type of the managed object. For example, the naming convention for ABAP instances is <System ID>~ABAP~<instance name> |
| ID of the metric or of the event |
| Name of the metric or event, for example dialog response time |
| ID of the metric path of the metric or event |
| Name of the metric path of the metric or event |
| Number of aggregates for the relevant key, which consists of MO-ID, type ID, and ID of the metric path. A high number indicates that the event or metric is frequently changing color. |
| The lifetime of the aggregates for the relevant key; this begins at the beginning of the of the oldest aggregate and ends at the end of the newest aggregate. Together with the value |
| Frequency with which the relevant key changed color. A high value here means frequent changes of color; this can have the following causes, among others:
|
| The age of the newest aggregate of the relevant key in hours. A large age indicates that the associated event or metric is no longer collected, or even that is has been deleted. |
You can use the following function keys to hide and show columns, or to change the columns by which the output in the list is sorted:
Key | Meaning |
---|---|
F5 | Switch the sort column from left to right; the sort column is highlighted by being displayed in a different color. |
F6 | Switch between showing and hiding the technical IDs. |
F7 | Switch between showing and hiding the displayed names. |
F8 | Refresh |
If a high age was determined for a metric type or event type, check whether the metric still exists and whether data provision is working correctly. You can do this, for example, in the Metric Error Handling program or in the monitoring use case System Monitoring
at metric level by choosing the context menu entry with the same name.
If a frequent change of colors was determined for a metric type or an event type, first check, as in the previous case, whether there is a problem in the data provision. If the problem is occurring due to the measured values fluctuating around the threshold values, check whether it would be useful to make appropriate changes to the threshold values. Note that you can set the threshold values between two colors differently, depending on whether the measured value is rising or falling, which is also a way of preventing the colors constantly changing.