Transactional RFC Activity situation alerts on SAP Agent

Body

ITCAM for Application SAP Agent provides an attribute group, called
"Transactional RFC Activity", that allows to monitor documents that are exchanged between the SAP system itself and external system.
It shows the involved function module, the involved nodes, the transaction code and so on.

While working with this attribute group, we noticed that a situation was closed almost immediately after it fired, despite the condition was still true,

at least looking at the information in the workspace view "Transactional RFC Status Information".

After having investigated about the method used to collect data for this attribute group, we understood the root cause for this scenario.

First of all, let's consider the situation interval, that was really small, only 30 seconds.

In an SAP environment, some attributes are available as snapshot data while others are collected using a given time span, basically from a begin date to an end date (date + time).
You can get this information from this technote, by finding the attribute group of interest and looking at the field type:

/support/pages/node/93653

All attribute groups that contain the sample_interval_end attribute are capable of returning data over a given time span.
This is the case of our R/3_Transactional_RFC.
By default, the situations provided by the agent request data for the time span that matches the situation sampling interval.
So, if the situation interval is 30 seconds, it will get all the data generated from "current time" to "current time - 30 seconds".

By default, the workspaces provided by the Monitoring Agent for mySAP request data for the default time span.
The default time span is defined per attribute group, and it can be checked and changed using /IBMMON/ITM_PERIOD function module.
Usually the default time span for workspaces is 30 minutes or one hour, depending on the attribute group workload.

So basically, when you refresh a TEP view, the ABAP code running on SAP returns data collected basing on a time span that is different than the one use by the situations.

Let's suppose that you get a new entry in the Transactional RFC Status Information view, if you have a 30 seconds interval situation with a formula that matches one of the value showed in the rows,
I expect it  to fire accordingly because the time span for both includes the data when alert has been generated.
Anyway, after 30 seconds the situation will run data collection again using a different time span.
Data previously collected will not be collected and as a consequence the situation becomes false and alarm is closed.

This is actually working as designed.

We have two possible solutions for it:

1) Increase the situation interval. If you set it to 5 minutes, it means that the situation alert will remain open at least for 5 minutes, if you use 10 minutes it will stay open for 10 minutes and so on.
2) Change the behavior within /IBMMON/ITM_PERIOD to make situation behaving like workspaces.

Step 1: Logon to SAP GUI and go to transaction code /n/ibmmon/itm_config
Step 2: Click on button /IBMMON/ITM_PERIOD
Step 3: Check/select checkbox under "Use in situations" column for "R/3 Transactional RFC" and save the changes.
Step 4: If you want you can also change the default sample period to increase it accordingly.

Step 5: restart the agent
In this way the data collected by the situation will be consistent with data collected by workspace and you will see the alert until the information is also visible in the TEP view.

Hope it helps

Best Regards

Subscribe and follow us for all the latest information directly on your social feeds: