Scheduling data refresh
As we learned before, cubes are a big cache in the Intelligence Server memory. Their data is static. We need to update them when new data is available, typically after an ETL job is run.
Refreshing a cube's data involves connecting to the data warehouse and re-executing the underlying SQL statements. Normally, the best time to do it is during off-peak hours, and since BI developers like to sleep at night, there must be a way to automate that.
MicroStrategy offers a scheduling mechanism to automate cube updates very similar to cron on *nix or the Task Scheduler on Windows.
In order to automate a cube refresh, we need two objects: a schedule and a subscription.
A schedule is just a recurring timer, like every day at 3 A.M., or every Monday; it does not perform any action. A subscription is an object that binds a report, document, or cube to a schedule and a delivery type. There are several delivery types in MicroStrategy, such as e-mail and printer; in our case we will...
