Skip to main content

Hi,

I want to show that a data model is successfully refreshed, in studio with the updated refresh date.

What is the best possible way to achieve this ?

I would opt to show some kind of max timestamp in a KPI component.

 

I see two possibilities:

  1. This could either be a natural timestmamp of e.g. MAX(activity-timestamp). Effectively saying something like : "Latest loaded activity from 10.01.2024".
  2. Or if you don't want to use a natural process based timestamp I would create an artificial table in the backend which writes the current timestamp during transformation. Assuming you only reload your datamodel in combination with extraction and transformations you can then access the MAX(of this load_timestamp) from the artificial table. The artificial table would need to be part of datamodel though. But not necessarily connected to any other table. This would then be "Latest data load: was 10.01.24"

 

I personally would prefer to see the latest timestamp of an activity in the frontend as its more explcit and doesn't allow interprationambiguity which data is loaded.


I would opt to show some kind of max timestamp in a KPI component.

 

I see two possibilities:

  1. This could either be a natural timestmamp of e.g. MAX(activity-timestamp). Effectively saying something like : "Latest loaded activity from 10.01.2024".
  2. Or if you don't want to use a natural process based timestamp I would create an artificial table in the backend which writes the current timestamp during transformation. Assuming you only reload your datamodel in combination with extraction and transformations you can then access the MAX(of this load_timestamp) from the artificial table. The artificial table would need to be part of datamodel though. But not necessarily connected to any other table. This would then be "Latest data load: was 10.01.24"

 

I personally would prefer to see the latest timestamp of an activity in the frontend as its more explcit and doesn't allow interprationambiguity which data is loaded.

Complementing Manuel's answer, you could use the column "_CELONIS_CHANGE_DATE", this column is automatically generated in each of the tables extracted in your data pool.

This column stores the timestamp in which the last extraction was done, so you can avoid making calculations to get this timestamp.

 

Hope it helps.


Hi @Raman Silotia,

 

Most of the times I use the solution as proposed by @andré.tafur above, which is easy to implement and quite accurate.

If you need a more specific timestamp of the exact data model load, you could enable Data Monitoring in your Celonis environment. In a new data pool, the information about the Data Jobs and Data Model is then stored. By using data sharing between pools, you can add this information to your Data Model of choice and display the information in your analysis.

 

See Custom Monitoring (celonis.com) for more information on how to enable the Data Monitoring Pool.

 

I hope this helps.

 

Best regards,

Jan-peter

 


Hi @janpeter.van.d , @andré.tafur , @manuel.wetze 

 

Thank you all

 

These responses are quite helpful and Custom monitoring looks quite a game changer for monitoring backend

 

Looking forward to implement these solutions

 

Thanks again !!


Hi @janpeter.van.d , @andré.tafur , @manuel.wetze 

 

Thank you all

 

These responses are quite helpful and Custom monitoring looks quite a game changer for monitoring backend

 

Looking forward to implement these solutions

 

Thanks again !!

Hi @Raman Silotia,

 

Custom Monitoring just got an upgrade, which might be interesting to check out: Update from an earlier version of Monitoring (celonis.com)!


Reply