Consider the following normalized relations from a database in a large retail chain:

STORE (StoreID, Region, ManagerID, SquareFeet) EMPLOYEE (EmployeeID, WhereWork, EmployeeName, EmployeeAddress) DEPARTMENT (DepartmentID, ManagerID, SalesGoal) SCHEDULE (DepartmentID, EmployeeID, Date)

What opportunities might exist for de-normalizing these relations when defining the physical records for this ­database? Under what circumstances would you consider creating such de-normalized records?

Assume that a data communications network links a computer at corporate headquarters with a computer in each retail outlet. The chain includes 75 stores with an average of 150 employees per store. There are 10 departments in each store. A daily schedule is maintained for five months (the previous two months, the current month, and next two months). Further assume that the following: 

• Each store manager updates the employee work schedule for her or his store roughly 10 times per hour. 

• An external payroll provider generates all payroll checks, employee notices, and other mailings for all employees for all stores. 

• The corporation establishes a new sales goal each month for each department, in collaboration with the respective store managers.

• The corporation hires and fires store managers and controls all information about store managers; store managers hire and fire all store employees and control all information about employees in that store.

a). Would you recommend a distributed database, a centralized database, or a set of decentralized databases for this retail store chain?

b). Assuming that some form of distributed database is justified, what would you recommend as a data distribution strategy for this retail store chain?