Registration

Dear SAP Community Member,
In order to fully benefit from what the SAP Community has to offer, please register at:
http://scn.sap.com
Thank you,
The SAP Community team.
Skip to end of metadata
Go to start of metadata

Issue: There are no alerts at all in the BPO Alert Inbox.

Description: You are using the BPO Alert Inbox in 7.2. There are no alerts displayed at all.
This issue can have several root causes:

Root Cause 1: There are no current alerts

The BPO Alert Inbox only displays open alerts (alerts not marked as "solved"). The following scenarios would result in no alerts in the alert inbox even though the monitoring is working as it should:

  1. There is a problem with the context selection of the BPO Alert Inbox
  2. All alerts have been solved (confirmed).
  3. The data collection only returned green ratings, meaning no alert records would be created.
  4. All monitoring objects are configured without the flag "Alerting active" set

Use the solutions listed below to find out whether there are alerts and if it's only the BPO alert inbox that has problems displaying them.

Solution 1.1.: Check for BPMon Alerts in the SysMon Alert Inbox

Start the Solution Manager launchpad. Go to area "System and Application Monitoring". In there, find the tile "Alert Inbox" and select this tile.

In the Alert Inbox, create a new query for BPMon alerts according to the attached document. Check whether BPMon alerts are displayed in this new query.

If there are alerts displayed, most likely you have an issue with your context selection. In that case:

  • Delete the persisted solution context for your user using report E2E_BPM_VARIANT_CLEANUP.
  • Clear your browser cache and afterwards call the alert inbox again.
  • Ensure that the embedded search is working correctly.

If there are no alerts displayed, there are no current alerts. If you feel that there should have been alerts, proceed with the next solution to search for confirmed alerts. 

Solution 1.2: Check for confirmed alerts 

Start the Solution Manager launchpad. Go to area "System and Application Monitoring". In there, find the tile "Alert Search" and select this tile.

In the Alert Search, look for alerts of use case BPMon.

If there are confirmed alerts, they will be displayed here:

If the alerts should not be confirmed, you need to check the configuration for the BPMon objects to determine whether the automatic alert confirmation was switched on. This is best done via the tile "Object Administration" in area "Business Process Monitoring" in the Solution Manager launchpad (see solution 1.4.)

If there are no confirmed alerts and no current alerts, you need to check in the Object Administration whether the alert creation was switched off for the BPMon objects (see solution 1.4).

Solution 1.3: Check the last data collection

Before checking the configuration of monitoring objects, let's check whether the data collection for the monitoring objects is running. This you can do by calling the BPMon application. You can do so via the tile "Business Process Monitoring" in the Solution Manager launchpad.

In the metric view for your monitoring objects you will be able to see the time of the last data collection (observe that this screen might vary between the different Solution Manager SPs).

Check whether the time stamp is in the recent past (keep in mind the data collection period for each individual monitoring object). Observe also whether a rating for the metric is available. If the data collection is running and a rating (yellow or red) + measured value are returned, the most likely cause is that the creation of alerts has been switched of (see solution 1.4.). If the flag for alert creation is set for the monitoring objects,  it is also possible that there are problems with the alert creation. Most likely, this is due to outdated MAI content.

How to check and update MAI content

To check this, call transaction SOLMAN_SETUP and go to area Business Process Operations → Basic BPO Confiuration. There, go to step 1.4.Update Content. Check whether new content for MAI is available.

If new content is available, apply the new content. Afterwards, generate and activate your monitoring objects again.

How to check the Runtime Overview

Another way to check whether the data collection for the monitoring objects is running is by checking the runtime overview in the BPO Object Administration. Choose the respective tile in the launchpad:

In the Object Administration, choose your solution/branch and select the relevant system role. Keep in mind that the navigation in the Object Administration might vary between the different Solution Manager SPs.  

In the Runtime Overview, filter for the active monitoring objects (custom filter "A" for column Status). Check whether or not the data collection got triggered in the recent past for your monitoring object (column DPC Last Trigger) and check that the data collection took place on the managed system (column Last BPMon DC). You can additionally check when the next data collection is due (column Next Run). As of SP5, the runtime overview will tell you in column Health whether or not the infrastructure for data collection works correctly. Column Rating tells you the last rating returned for the monitoring object. Observe that also Analytics objects are displayed in this list, which by definition have no rating.

If the entries for DPC Last Trigger or BPMon Last DC are missing in this overview, most likely there is an issue with the infrastructure. In this case proceed with Root Cause 3.

If the rating is missing even though the time stamps are there, most likely this is an issue on with the data collection on that specific managed system, either due to RFC problems or authorization issues. If the data collection is supposed to happen in background, it might also be an issue with the background data collection.

Solution 1.4. Check the configuration in the Object Administration  

In the Object Administration, check whether the monitoring objects have the flag for "Alerting Active" set. To do so, use tile "Object Administration and navigate to the correct solution-branch, system role and site in the Object Administration. 

In tab "Further Settings", make sure you include the columns for "Alerting Active" and "Automatic Alert Confirmation". Note that these settings cannot be maintained for Analytics Objects (because Analytics objects do not raise alerts). Additionally, only monitoring objects of use case "Business Process Monitoring" have these flags. So for Interface Channel Monitors and Job Montiors you cannot maintain the flags in the Object Administration.

Make sure that you set the flag for "Alerting Active" for those monitoring objects that should raise alerts. Also review if the flag for "Automatic Alert Confirmation" is correctly set for these alerts where a green rating should automatically confirm all open alerts. Save, generate and activatee your changed monitoring objects.

Root Cause 2: Wrong solution context selected

You are accessing the BPO Alert Inbox for a specific solution context. In case you are using the wrong solution context, there might be no monitoring objects assigned to the business process context, consequently there are no alerts displayed. Use the solutions below to rectify this root cause

Solution 2.1: Switch the solution context

 Make sure you select the correct solution context. Ensure that you use the correct solution, branch, site and system role.

Keep in mind that the last selected solution context for your user is persisted. When you access the alert inbox again, your last solution context is used.

Clean up persisted Solution Context

In case of issues, you can delete the persisted solution context via report E2E_BPM_VARIANT_CLEANUP.

If you are sure you have selected the correct solution context and the alerts are visible in the System Monitoring alert inbox, you will have to open an SAP incident.

Root Cause 3: Issues in the Infrastructure

There can also be issues in the MAI infrastructure that prevents the MAI to create alerts. The check whether this is the case, use the solutions listed below:

Solution 3.1: Check the DPC PULL extractors

If the data collection for the monitoring objects does not get triggered, check whether the DPC PULL Extractors are scheduled and regularly executed. To do so, go to area "Solution Manager Administration" int he Solution Manager launchpad and execute tile "Extractors- Alerting Administration"

Extractors are scheduled per managed system. Filter for the Name "DPC PULL CORE" and your manaed system in column "System".

BPMon data collection mostly happens via DPC PULL extractors with extended context ST_PI_CLNT<client>_*. Typically, there are 3 entries available:

  • ST_PI_CLNT<client> which runs every 5 minutes and triggers the data collection of monitoring objects running every 5-10 minutes.
  • ST_PI_CLNT<client>_15 which runs every 15 minutes and triggers the data collection for montioring objects running every 15 minutes - up to once a day.
  • ST_PI_CLNT<client>_60 which runs every 60 min and triggers the data collection for monitoring objects running less often than once a day.

The extractors are scheduled latest with the activation of the monitoring objects. Depending on your data collection periods, it might be that not all of these extractors are scheduled (e.g. if you only have monitoring objects scheduled to run every 5 minutes, then you might only have the extratcor ST_PI_CLNT<client> ). Please also note that system monitoring uses some of these extractors, so even if you have no BPMon active, these extractors might already be scheduled.

When checking the etxractors, you need to check the following:

  1. The extractors are available for the managed system. If they are not available, generate and activate monitoring again. This should make the etxractor entries reappear.
  2. The extractors are scheduled for the managed system (Last update is in the recent past). If they are not scheduled, check whether the extractor is active (column Active Status). If the extractor is not active, switch to edit mode and use button "Activate" to activate the extractor. If the extractor is already active, but not running, check whether background job
    EFWK RESOURCE MANAGER is running. If not, schedule this background job via SOLMAN_SETUP

Solution 3.2: Check the configuration data in the Alerting Directory Broswer

If the extractors are correctly running, check whether the configuration data in the MAI is correct. To do so, access the Alerting Directory Browser.

How to access the Alerting Directory Browser

Call transaction MAI_TOOLS. In there, select the tool "Alerting Directory Browser"

This opens a browser window. There, use monitoring use case "BP_MON" and select "Apply".

 

This opens a tree with the alphabetically sorted list of monitoring objects. Go to one of the monitoring objects that is not working.

Important Information in the Alerting Directory Browser

There should be a metric entry that is active. Mark the respective line. This opens the confifguration for the metric below. There are many important fields. Here, we are only going to highlight some of the most important configuration fields.

There should be a data provider listed for the metric. This data provider is the function module that is used to collect the data for this metric. Ther should also be some data collection information (e.g. the data collector frequency and the days of the week the collection is supposed to take place). An last but not least there needs to be a collection context ID with a system on whcih the data collection takes place.

If some of these entries are missing, make sure you have the latest MAI content available in your Solution Manager system. To check this, call transaction SOLMAN_SETUP and go to area Business Process Operations → Basic BPO Confiuration. There, go to step 1.4.Update Content. Check whether new content for MAI is available and if yes, download the latest MAI content to your Solution Manager system.

Afterwards, try to save, generate and activate your monitoring object again. Check afterwards in the alerting directory browser whether the configuration data is now available. If not, open an SAP incident.

Solution 3.3: Check for banned providers

One of the reasons why data collection might not be executed could be a banned provider. To check for banned providers, use tile "Extractors - Alerting Administration".

In the next screen, use tab "Banned Providers". Look for providers starting with /SDF/E2E_BPMon* (this is the same data provider here as listed in the Alerting Directory Browser). In case you use ICMon and Job Mon, remember that these data providers use a different name space.

The providers that are banned will not be triggered for that managed system. Usually, providers get banned because the data collection for these providers took too long. This can be caused by a long runtime of an individual montioring object or by a large number of monitoring objects using the same provider being collected at the same time.

You can release the provider from this screen again to "un-ban" it, but if you leave the configuration as it is, most likely the provider will get banned again pretty soon. So in case you find banned providers, check the configuration of your active monitoring objects for this provider first. Schedule data collection in background if possible and distribute the data collection across the day.

If there are no banned providers, proceed with the next solution.

Solution 3.4: Check the result of the ad-hoc data collection

You can manually trigger the data collection for a metric from your SAP Solution Manager. This will only work for monitoring objects where data collection happens in dialog. To do so, first you need to get the managed object ID and metric ID fromt he Alerting Directory Browser. See Solution 3.3: Check the configuration data in the Alerting Directory Broswer for how to access the Alerting Directory Browser.

In there, go to tab "Others" and copy the metric ID and the managed object ID.

With these, go to SE38 and execute report E2E_DPC_GET_DATA_CHECK. In there, enter the managed object ID and metric ID. Afterwards, execute.

The data collection will bring back a result screen. If the data collection was successful, it should look something like this:

The rating should be green, there should be an entry in column "Rating" and a metric text.

If there is a red rating, it means that something went wrong with the data collection on the managed system.

  • No labels