Are you sure the 2 applications have different end product databases assigned? This sounds like there is either a hardcoded database declaration somewhere in the code. Or the applications might be sharing the same end product database. If you could share some more information, what the system flow looks like, part of the code or something else to clarify the problem I might be able to assist. 

Hi Erwin,

I checked the runtime configuration and yes, the 2 applications have different Databases on the same Server.

The system flow consits of couple of subflows

The “Multiple running instances” is set to off

Each subflow does some clearance of some DTO tables, in order for them to be populated in the next steps by doing some GET requests on Microsoft endpoints. At the end of each subflow, the data in checked with the data from the server (normal inserts,updates, deletes with joined DTO table) in a transaction. Long storry short is a pulling system from Microsoft since some features does not support changed notification.

The first decision and the execute system tasks from the main process flow only changes the data of a flag value from the database in order to allow for the process to start or not (i will remove it since the checkbox for multiple instances is off anyway)

Both of the branches have a different application created in IAM on wich the database is also different (same as it is in SF). On both branches I ran a full deploy and sync in IAM.

I checked the process and is running on just 1 branch for about 20 hours now, the data is stable, no unexpected delition of the data and so.

I’m not familiar with Microsoft endpoints but based on this information I would search for the problem there, and not in the software factory. Perhaps someone else on the community has experience with Microsoft endpoints to indicate if the problem might lie there or if there is an option to add logging for the Microsoft endpoint.