Data Ingestion & Infrastructure

Hi all, I'm currently enrolling Celonis Summer School and I have no experience on PQL neither SQL languages, so I please ask your help. I am trying to write something like a "SUM IF" expression. Basically, when column "_APX_WDCRP_CASES_csv"."ORDER_DATE_MET" = 0, I want to sum "_APX_WDCRP_CASES_csv"."ORDERED_QUANTITY" values.  I already tried to mix CASE WHEN function with SUM but I was not succeeded. I also searched for other functions related with aggregation but could not find one that applied. Many thanks!Guilherme Luz

Hi together, We are facing an issue with the data load since 9th April. We cannot load data from SAP HANA databases anymore. Everything worked till 8.4. We have these issues in different teams (all on eu-1) and for different SAP HANA databases from different hosters. Therefore, it seems that something changed with the driver within Celonis EMS. Error: Unexpected error when creating database connection: Cannot load JDBC driver class 'com.sap.db.jdbc.Driver'(see attachment) Any ideas? Thanks in advance,Clemens

I established the connection with service now.When i tried to install the service request management but i got error sys_audit table is not present.so I check in Extraction part to extract the sys_audit table but i dont have the table.

Hi All,I am looking for filter condition which can filter out records from activity table once specific activity is first time occurred.Tried with REGEX, PU_FIRST but some how this doesn't work.Is there any one applied same kind of filter condition? Below is what I am expecting:Anything after first event 'Approved' is what I wanted to exclude. 🙂

Quick Tip!  For faster run times, move filters from the where clause into the appropriate inner join. The top example is an optimized version of the bottom code. Let me know if you have any other tips or questions!  

Dear Team,  Please give me an advice about how I can get the timestamp which represents when was the last execution of a specific data job Thanks in advence!Regards, Roman

Example: CASE WHEN YEAR("X"."y") >= 2019 THEN FILTER ROUND_DAY("X"."y") != ROUND_DAY("A"."b") OR ROUND_DAY("X"."y") != ROUND_DAY("M"."n")OR ROUND_DAY("A"."b") != ROUND_DAY("M"."n") ELSE FILTER ROUND_DAY("A"."b") != 2019-01-01 OR ROUND_DAY("M"."n") != 2019-01-01 END Neither CASE WHEN nor IF worked for me. I'm trying in a Component Filter, though a solution on a Sheet level could also solve my problem.

Hi everyone!I have a doubt about a calculation related to the duration between activities. Normally the CALC THROUGHPUT formula is used to calculate durations between 2 activities. However, I need to calculate the duration between more than 2 activities in a single KPI (e.g. the duration between activities A and B to activity C, or from A and B to C and D), but I can't get the formula.  I already have tried the following formula to calculare the duration between activities 'A' and 'C' to 'B' AVG (CASEWHEN PROCESS EQUALS 'A' TO ANY TO 'B' THEN CALC_THROUGHPUT(FIRST_OCCURRENCE['A'] TO LAST_OCCURRENCE['B'],  REMAP_TIMESTAMPS("activity_table"."timestamp"))  WHEN PROCESS EQUALS 'C' TO ANY TO 'B' THEN CALC_THROUGHPUT(FIRST_OCCURRENCE['C'] TO LAST_O

Hi, We have installed Extractor for JDBC connection for Oracle inside Oracle Data base Server, oracle DB server installed in a Physical server, during connection test with uplink connector we are getting below error. 2021-07-12 14:31:32.680 INFO [,,,] 30305 --- [uplink-thread-0] c.celonis.uplink.manager.UplinkManager  : Uplink could not connect: 'I/O error on POST request for "https://tcscboprod.us-2.celonis.cloud/uplink/api/public/uplink": Read timed out; nested exception is java.net.SocketTimeoutException: Read timed out', waiting 3000ms before retrying --EOM-- We have enabled the port & IP for source and IBC Please suggest for possible solution to resolve the above error

Hi!I have a doubt about a calculation related to the duration between activities. Normally the CALC THROUGHPUT formula is used to calculate durations between 2 activities.However, I need to calculate the duration between more than 2 activities in a single KPI (e.g. the duration between activities A and B to activity C, or from A and B to C and D), but I can't get the formula. I already have tried the following formula to calculare the duration between activities 'A' and 'C' to 'B' AVG (CASEWHEN PROCESS EQUALS 'A' TO ANY TO 'B' THEN CALC_THROUGHPUT(FIRST_OCCURRENCE['A'] TO LAST_OCCURRENCE['B'],  REMAP_TIMESTAMPS("activity_table"."timestamp"))  WHEN PROCESS EQUALS 'C' TO ANY TO 'B' THEN CALC_THROUGHPUT(FIRST_OCCURRENCE['C'] TO LAST_OCCURRENCE[&#

Hi! I am using the same simple formula to calculate the average throughput time in the OLAP table and in the Gauge Component. OLAP table has dimensions (e.g. Job Family). The case is a Vacancy ID, and there are 2 basic tables in the data model: dimensions table (containing Job Family attribute) and facts table (activity table for the Vacancy status changes) where Vacancy ID is a key. So the data model is pretty simple as well. Could you please help me to figure out why I am constantly getting different results when I am comparing throughput time calculated based on OLAP table selection with the throughput time calculated by a single component (e.g. Gauge or Throughput Time Search)?  Throughout Time formula: TRIMMED_MEAN(CALC_THROUGHPUT(ALL_OCCURRENCE['Process Start'] T

I would like to compare an attribute (e.g. price) of one activity from e.g. caseID 123 with another activity from the following caseID e.g. 124 - is there any way to solve this problem in Celonis?   

Hi! I have a use case when a specific field within 1 case can be changed multiple times during the case lifecycle (e.g. fields containing metadata around the Vacancy case). The aim is to figure out what is the influence of specific field changes on the Vacancy throughput time. For this, I am using range buckets calculated based on the number of specific field changes per Vacancy.  I've created 2 tables with buckets: the first table allocates a Vacancy to a range bucket (OLAP table dimension) using PU_COUNT function and the second table does the same using CALC_REWORK function (all formulas are provided below). Then I calculated the # of vacancies in each bucket and the average # of field changes in a bucket (both tables show the same result). Then I am adding throughput time calculation to the tables, and the result is the same at a first glance. But when I hide the column with average # of field changes in a bucket from the second table, throughput t

Hi, is there a way to add weekdays to a date?E.g. Goods Issue date is on a Fridayand we need to add 2 weekdays (Monday through Friday are weekdays)to the Goods Issue Date, so that we end at a Delivery Day of Tuesday (Friday + 2 weekdays)ADD_DAYS would give us Sunday as output, and ADD_WORKDAYS only works with a workday calendar (needs TFACS table which we do not have)REMAP_TIMESTAMP even with the ADD_DAYS function would bring us to Monday as the next weekday.

We have multiple cases when we want to make sure there is an explicit correlation between two data series apart from visually representing those series on a graph. Is there any simple KPI/PQL formula we can use, or do we need to go with a workaround deriving corr from calculating standard deviations etc?

Hi There, I am trying to Filter on component level. The filter has two parts and looks like this: 1.Filter : Filter all cases, which contain only activities, which does NOT match user_type = A FILTER PU_COUNT ("AP_VIM_CASES" , "_CEL_AP_ACTIVITIES"."_CASE_

Hello, I am looking for something similar to a lookup function in Celonis. Is it something possible? I have a child table "X" linked to Parent table "A" another child table "Y" linked to same parent table "A". Both links are based on different keys. I want to lookup a value from X to Y is it something possible as in is there PQL formula which could help me do that? Thanks.

We are trying to connect to SAP ECC using uplink connector, post Test Connection got successfull, tried to create Data Job and trying to load tables but not moving ahead from log "Job not yet started in SAP. Current status: SCHEDULED", so it could be issue in SAP RFC module, unable to understand the issue please suggest.

Do you have a tips how I can create the sorting numbers? Do I have to create the sorting number while i running the procedures to create the event log entries afterwards when the event log is created? But I don't know how to define the sorting number with SQL.

Best 20% and worst 20 % Cases based on their duration Hi everyone, we are looking for a solution to the following problem. We would like to get the best 20% and worst 20% of the Cases based on their duration, including all cases that fall between the scope, e.g. in a pie chart.So we could filter on the specific cases that are in the defined scope and as an output we would like so see them individually (with Case Key and their specific duration) in an OLAP table. Does anyone know what formula to use to get the relationship to each other?

Does anyone know how can you remove disk space (hence reduce APC consumption) removing specific rows from a table? I would need to remove permanently rows from a table based on a specific sub-query. I have limited space available so i cannot use operations that require, even for a small period of time, twice increase in space occupation Thanks in advance  

UTC TO LOCAL TIMEHello, I'm curious if anybody else has already faced the following issue: For some tables, timestamps are stored in UTC in SAP, although in display mode of transaction timestamp is shown in local time. This leads to UTC timestamps for activities in Celonis and can cause confusion of end users.  I already tried SQL best practise to transfer UTC in local time but standard operations/ local time zones are disabled in Event Collection. SELECT GETDATE() only provides UTC (Vertica base)getutcdate() is not supported (Vertica base)SELECT * FROM sys.time_zone_info is not allowed/missing rights (SQL Server base) It would be better to avoid hard coded transformations due to summer and winter time in Europe which will lead to maintaining effort. Does anybody else has an idea how to transfor UTC in local time?  Looking forward to hear your experience

API Key vs. Bearer Token for Data Push API Hi all,  We are trying to push data towards the EMS using the Data Push API. In our tests with Postman, we were able to do this, but we found out that the API key in the EMS should be used as a Bearer token.In the application we use, a Bearer token can only be obtained by requesting it from a authorization endpoint.Moreover, we cannot use the Celonis API key as a REST API key authorization, since you need a key name for that (and this is not the key name in the EMS). Has someone of you experienced the same issue, and is there a workaround for it?  Kind regards,Jan-peter

Good morning everyone, Is there a way for me to extract more than 100 rows of data from the Transformations / Event Collection workspaces? For some context, we don't have access to the SAP environment from where we are extracting the data; we do, however, need to provide a considerable sample of the data we're working with to validate with the team. If I do a regular SELECT * FROM CDPOS for example, I will only be able to get 100 rows of data at a time Thank you in advance! P.S: Apologies if this is not posted in the appropriate place!

Hi Celopeers :) We have a specific use case, based on JIRA data. For a case a limited set of activities (~7, JIRA stories) need to be performed. For each activity (/story) we know when it was ready for pickup, in progress, and done (or rejected). If we would like to calculate the total time the activity/story spent in the status 'in progress' I was thinking of something like this: SUM( CASE WHEN "TABLE_STATUS"."STATUS" = 'In progress'   THEN HOURS_BETWEEN("TABLE_STATUS"."STATUSDATETIME",ACTIVITY_LEAD("TABLE_STATUS"."STATUSDATETIME")) / 24 --to get fractional days   ELSE 0.0 END) Where the dimension could be the different activity/story names for instance. However, the activity_lead often is a very different activity (/JIRA story), hence the calculation fails.I also cannot use a process pattern matching using fir