Assign Roles and Access Privileges
Mart configuration is usually considered an administrator function, and is not a business user function. Data Mart’s privileges in User Manager are granular enough to support the differentiation among configuration roles that most organizations need. Where there is more than one Data Mart role required, the roles usually fall under the following categories:
Administration, with all privileges.
Model management, with Maintain, Configuration, Migrate and View privileges.
Production, with Submit and View privileges.
Data quality assurance, with Audit, View and Validate privileges.
Adopt an Archive Policy
Archiving Data Mart tables can improve index rebuilding and query performance simply by reducing the number of rows to be navigated. A logical policy is one that preserves each data frequency (daily and monthly) for the time span that reporting may need it. Many clients preserve monthend data indefinitely, but keep daily data for a few months or a year.
Remember that all as-of reporting against Data Mart requires that data be kept for all possible prior dates you might want to report. If you need data for time periods already purged, the lost rows can be rebuilt, but there will likely be restatements of numbers previously reported.
You should purge the Data Mart execution tables and leave them with no more than 1 months’ worth of data. Your specific needs may be different. Not purging can result in system performance issues.
0 Comments