Skip to main content

Hi everyone.

 

I'm working on a project that we have the following scenario: Extractor Server is running on a Virtual Machine, that has access to an SAP system, and we have an uplink connection between Celonis Cloud and Extractor Server. This is working fine during working hours, let's say 9-to-5.

 

However, due to resources saving, the Virtual Machine is running just 08 hours a day. And some tables take up to 13 hours to extract (full load).

 

Suppose that we start extracting data at 8 am: RFC Call => Job Running on SAP => Generating files from 0...N => Extracting to Celonis => Repeat. Then at 5 pm the Virtual Machine shuts down. What will happen to the files generated after 5 pm on SAP? When the Virtual Machine is up again, Celonis will recover the work from where it stopped or we need to start it over again?

 

It happened on our project during a big extraction, we had to cleanup the files manually (Report /CELONIS/CLEANUP) and start extraction again, since Celonis didn't recover its previous jobs on SAP.

 

Is there any configuration that we are missing here? Or any other way to recover previous extraction after Extractor Server went down?

 

Thank you

Hi Guilherme,

 

As far as I know, the data batches will be staged in Celonis, and only if the extraction is succesful, it will be loaded in the Data Pool.

My suggestion would be to make your loads smaller, by using filter statements.

 

Kind regards,

Jan-peter


Hi Guilherme,

besides Jan's suggestion, it is worth checking whether all these tables have indexes created. This should accelerate your jobs.

Alternatively, if keeping the VM running is not an option, see if you could instead of running daily extractions, schedule it to run only certain days of the week, but ensuring the VM is online.

 

Best,

Gabriel

 


Reply