Audit logs are chronologically logged processes in a system. In SAP Data Warehouse Cloud (DWC), it is also possible to trace which user did what and when. First and foremost, this allows access to SAP Data Warehouse Cloud to be logged.
How to connect SAP Data Warehouse Cloud with Power BI and Tableau
Follow the traces: Audit logging in SAP Data Warehouse Cloud
Using the logs, developers can see how the system got to its current state. The available data can also be beneficial for a compliance audit if the auditor requires proof of compliance with the applicable regulations.
However, audit logging also comes with a disadvantage: depending on the time period chosen for storing the operations, the logs can take up a lot of storage space at the SAP Data Warehouse Cloud tenant. Retention periods between 7 and 10,000 days are available for selection. The default value set is 30 days.
Every space and database user in SAP Data Warehouse Cloud has the option of activating audit logging and specifying a corresponding time period. It is distinguished whether the user wants to log only the changes or also the read accesses.
The logs are not displayed in an overview by SAP Data Warehouse Cloud, but must be prepared in a view. To do this, the user navigates to System/Configuration/Audit. There he specifies a space in which the audit tables are made available. The recommendation is to use a space exclusively for audit logging. This way the accesses to the audit data can be limited.
The user then creates a new view in the selected space. Under “Sources” there is “DWC_AUDIT_READER” with three tables.
The view “AUDIT_LOG_OVERVIEW” contains the number of log entries for the different audit policies.
The view “AUDIT_LOG” contains the data of the logs.
View “ANALYSIS_AUDIT_LOG” contains all audit logs that can be assigned to a database user.
These views can now be used for creating a view in space or also for modeling. In this way, the created view can also be connected to SAP Analytics Cloud to display the audit data graphically accordingly.