I'm trying to get a feel here from some other database professionals... what criteria would you use to decide whether or not a new system of tables should have its own database or be included in an existing database?
---
Our company database has about 40 tables which load daily from another extract environment. New requirements dictate that I need to load and retain a subset of 13 of these tables at a few different time intervals.
I could create the new schema in 1 of 2 ways...
1) add the new tables to the existing company database with time suffixes... as in :
tbl_Customer
tbl_Customer_6AM
tbl_Customer_8AM
tbl_Employee
tbl_Employee_6AM
tbl_Employee_8AM
tbl_Plan
tbl_Plan_6AM
tbl_Plan_8AM
...etc
OR
2) create two new databases named EARLY_REPORT_DB_6AM, EARLY_REPORT_DB_8AM, as in...
COMPANYDB
tbl_Customer
tbl_Employee
tbl_Plan
...etc
EARLY_REPORT_DB_6AM
tbl_Customer
tbl_Employee
tbl_Plan
...etc
EARLY_REPORT_DB_8AM
tbl_Customer
tbl_Employee
tbl_Plan
...etc
Other Notes :
------
- Regular users need access to the original tables only.
- A special report process will run using the 6AM DB & another special report process will run using the 8AM DB.
- By definition, these tables will never intersect (that is, we will never need to access tbl_Customer_6AM & tbl_Employee_8AM in the same query).
Would you include the new tables in the current database, or would you create two new databases. What would be the pros & cons of each?
Thank you all,
Leon