Replies: 1 comment 8 replies
-
@ndr-brt is already working on a general eventing system that auditing can be built on top of as opposed to introducing a specific interface into the runtime to deal with auditing. I would sync with him. |
Beta Was this translation helpful? Give feedback.
8 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
A topic that has come up with us is the special tracking of a data history and the use of this data in particular, so separate from a logging which deals with the state of the data ecosystem / the connector and its component state.
For this I/we would develop an extension which offers an interface internally, like the Monitor already offers, that within the Connector as soon as data is manipulated or moved, this must be logged actively, or an observation on it can take place (passively). So that a data transfer from Connector A to Connector B arrive also logged.
Besides the internal logging there should also be an external interface where external certified software can write logs how they use the data or where they move the data to for example.
For this the first idea would be to write a simple and small extension which can store in-memory logs. A simple API interface which could log logs from external with a POST method. In the first step an active logging mechanism is easier to implement, but if possible in a next step it should become an observing logging, so that the error-proneness within the system would be reduced.
Accordingly, the following information should first be logged over the data:
• DatenID
• SourceId
• Timestamp
• Action (USAGE, DATATRANSFER)
• LogMessage (More detailed description but free text)
What should be logged first?
• Access/use to data
• Data manipulation
• Move data
Critical is the concrete implementation of the identification of data (data ID), so that the logs can be assigned to them.
Beta Was this translation helpful? Give feedback.
All reactions