Skip to Content

The idea is to build a real time dashboard that displays status of process chains in SAP BW.

 

This dashboard made our on-call monitoring activity very easy, saved us lot of time, we are able to proactively identify long running/delayed jobs and take necessary action in a timely manner.  

 

This real time Dashboard should fulfill following requirements.


1. Automatic refresh, easy to launch and monitor.

2. Show currently running or failed process chains that started in the past 24 hours.

3. Show upcoming meta process chains that have not yet started for the day.

4. A progress bar to show the progress of current run of a PC based on historic trends.

5. Notification of long running process chains (Visual and Audio Alerts).

6. Notification of delayed process chains (Visual and Audio Alerts).

7. Show only latest failed instance in case of multiple failures of same process chain.

8. Do not show a failure if the more recent run of the same process chain is successful.

 

This dashboard is intended to address following pain points.

 

1. The native process chain monitoring transaction (RSPCM) requires manual refresh and sometimes is slow.

2. No native t code/report to only show running or failed process chains.

3. Process chains sometimes get stuck in yellow status for long time causing delays in data loads, we fail to notice these as there is no failure/error.

4. Sometimes process chains may not start at specified time and again we fail to notice this as there is no failure/error.

5. Need to manually add new process chains to RSPCM.

Below is the screenshot of the dashboard that we built using XCELSIUS on top of BW statistics cubes.

 

image

 

This is an XCELSIUS dashboard built on top of BW statistics cubes. We use Business Objects WIDGETS to launch and monitor this dashboard from our desktop. BI WIDGETS are great for real time monitoring because we can setup WIDGETS to automatically launch during startup, it will essentially become a desktop widget which one can always keep on top of other windows or take quick peek using keyboard shortcuts.

As one can see from the screenshot the top part of the dashboard shows currently running/failed process chains, there is also an audio alert which will go off when a process chain fails. There is a progress for each process chain which tells the progress of process chain (This is based on current running time vs average running time in the past 30days). The color of the progress bar changes from yellow to red if a PC is running longer than 130% of average time, there is also an audio alert for this.

 

This dashboard automatically refreshes every five minutes.

 

The bottom part of the dashboard shows upcoming process chains and their start times. A red alert icon with the message “Delayed” will appear under comments section if for any reason a PC does not start at the specified time. 

 

I will follow up with more technical details and step by step instruction in the next blog post.

To report this post you need to login first.

9 Comments

You must be Logged on to comment or reply to a post.

  1. Ponnurangam Rajesh
    Hi There,

    It’s a very nice idea of dashboarding BW Statistics data and use it for real-time monitoring. I have never seen a good use case for BI widgets being used…So this blog should be quite interesting..

    (0) 
  2. Tobias Haas
    Hi,
    do you know the transaction ST13 -> BW Tools -> Process Chain Analysis.
    Here you can monitor every process chain and fix dataload problems directly.

    Kind regards,
    Tobias

    (0) 
  3. Peggy Bishop
    Hi,

    How often do the BW statistics cubes get refreshed in BW in your environment?  If we took your approach, which we would love to do, we would see day-old data.  If your statistics cubes get updated more often, can you say how that is happening?

    Also, when will part 2 of the article be released?  We are very interested in doing a dashboard like this.

    Thanks,
    Peggy

    (0) 
    1. Mahidhar B Post author
      We are using a virtual cube for this dashbord. So it is real time data. I will explain it in more detail in next par very soon.
      (0) 
  4. James Elliott

     

    Hello Mahidar,

     

    I understand this is an older post, but do you have the second one for the instructions? I think we can use this dashboard….

     

    Thanks..

    Jay

    (0) 

Leave a Reply