Practice 1 Create a simple table without indexes, insert 1,000,000 rows into the table, and write down the query cost and execution time. Add one non-clustered index to the table, truncate the table, re-insert the 1,000,000 rows, and write down the query cost and execution time again. Do this for a few more non-clustered indexes and review the performance difference.
Practice 1 Create a simple table (with a row width of at least 50 bytes) without indexes and insert 1,000,000 rows into the table. Execute queries that retrieve 10, 100, 1,000, 10,000 and 100,000 rows (for example, use a BETWEEN filter on an identity column in the table) against the table without indexes and write down the query cost and number of page reads for each query. Create a clustered index on the column being searched, reexecute the queries, and write down the performance metrics again. Do the same thing for a covered and uncovered non-clustered index.