This is a hypothetical question - the problem listed below is entirely fictional, but I believe if anyone has an answer it could prove useful for future reference.
We have a situation wherein multiple systems all populate the same data table on our SQL Server. One of these systems seems to be populating the table incorrectly, albeit in a consistent pattern (leading me to believe it is only a bug in a single system, not multiple) These are majoritively third-party systems and we do not have access to modify or view their source code, nor alter their functionality. We want to file a bug report with the culprit system's developer, but we don't know which one it is as the systems leave no identifiable trace on the table - those in charge before me, when the database was new and only occasionally used by a single system, believed that a single timestamp field was an adequate audit, and this has never been reconsidered.
Our solution has to be entirely SQL-based. Our thought was to write a trigger on the table, and somehow pull through the source of the query - ie, where it came from - but we don't know how, or even if that's possible.
There are some clear solutions to this - for eg contact all the developers to update their software to populate a new software_ID
field, and then use the new information to identify the faulty system later (and save my fictional self similar headaches later) - but I'm particularly interested to know if there's anything that could be done purely in-house on SQL Server (or another clever solution) with the restrictions noted.