Error message "No Common Parent"


in our internal prototyping we sometimes are getting the message “no common parent” in our analysis.
We have found some work arounds by

  • keeping the case table on the left of the join
  • adding a common “_case_key” to tables

However we would like to understand the cause of this issue


Hi Geoff,

I run into the same issue almost every day when creating new apps. I don’t have any official confirmation, but I can share what I’ve deduced from my observations. I believe, the error is rooted in the cardinality of the involved tables. If you try to combine two tables whose relationship is of type n:m, this error occurs. Celonis typically does not allow for direct n:m relationships, but if you combine three tables with the following relationships, you end up with the same Situation:

lets say you have 3 tables A,B and C. A -> C (1:n), A -> B (1:n). Trying to inlcude table B and C in one Analysis results in said error for me, especially when table C and B use a different set of keys. This situation artificially creates a n:m relationship.

This error also occurs when your case table is on the “wrong” side of a 1:n relationship, such that one row can be related to multiple rows of the case table. This actually makes sense if you think about it from a process mining perspective and it would also explain why keeping your case table on the left side of the join solves the problem.



Hi Geoff,

the reason for this error message is that Celonis only supports direct joins between two tables meaning you cannot join tables which are not directly connected in the datamodel.
For instance, you have three tables A, B and C in your datamodel. A and B are connected, B and C are connected, however A and C are only indirectly connected via B.
If you try to display columns from A and C in an OLAP table you will receive exactly this error message. A solution to this problem is to “pull up” either the columns from table A or the columns from table C to table B using a PU-function (e.g. PU_MAX, PU_SUM…) in Celonis.

A function PU_XXX( B, C.column1) aggregates the values from column1 in table C into the table B.

Best regards,


Thanks very much for the reply. The pull up function looks useful


thanks for the reply. Have a look at the reply from Celonis , this gives an explanation of why.
The other approach is to use SQL in the database to add the required fields to the case table.


Hi Kerstin,
I tested the described scenario since we have plenty constructs that should be equivalent. I connected the SAP tables EBAN (Purchase Requsition items), EKPO (Purchase Order Items) and EKBE (Goods Receipts for Purchase Order Items) , selected a column from each one and combined them in an OLAP Table without any problems. Only using columns of EBAN and EKBE doesnt yield any errors either. I connected them in the following way EBAN -> EKPO -> EKBE. Are you sure that this causes the error?



Hi Ruben,

Celonis only supports joins between tables which have a 1:N or M:1 relationship.
If tables A and B from my previous example have a M:1 relationship and tables B and C have a 1:N relationship, the tables A and C have a M:N relationship and you need to “Pull up” either the columns from table A or the columns from table C to table B.
I assume that in your example (EBAN -> EKPO -> EKBE), there is a 1:N relationship between EBAN and EKPO and there is a 1:N relationship between EKPO and EKBE meaning that there is a 1:N relationship between EBAN and EKBE. In this case, displaying columns in the same OLAP table from EBAN and EKBE is possible without any further restrictions.

Best regards,