Sizing Exadata - Oracle Exadata Recipes: A Problem-Solution Approach - page 126

Finally, you can add additional logic into the script to determine which Exadata configuration you need to target
based on your growth rate, number of similarly sized databases (think in terms of production clones), percentage of
primary database size that each clone consumes, whether you're targeting High Performance or High Capacity disk
drives, the percentage of space reserved for “data” vs. your Fast Recovery Area, and whether you're electing ASM
normal redundancy or ASM high redundancy.

Not ■ please see recipes 5-4 and 6-6 to learn more about aSM redundancy and recipe 5-9 to understand
techniques for measuring your Fast recovery area.

In the script in Listing 5-20 below:

&num_dbs = Number of additional databases that are copies or clones of the database you're
running the query against.

•

&pct_db_size = Percentage of size for each additional database with respect to the database
you're running the query against. For example, “1” means that each additional database will
be the same size, “.25” means each additional database is one-fourth the size, and so forth.

•

&hp_or_hc = Whether you're using High Performance or High Capacity Disk drives. For this
parameter, enter either “HC” or “HP.” If the string is unmatched, the query assumes High
Performance disks.

•

&pct_for_data = Percentage of disk space reserved for database files, excluding FRA
requirements. For example, if you wish to size your FRA to 50% available storage, enter .5.
If you need a 25% reservation for the FRA, enter .75.

•

&asm_redundancy = Normal or high ASM redundancy. If string is unmatched, the query will
assume NORMAL redundancy.

•

For the execution of this script in our environment, we will assume five additional databases, each equally sized
to the current database, and use of ASM normal redundancy with High Performance drives. We will further assume
we will reserve half of our Exadata storage capacity for our FRA.

Not ■ as executed, the script below assumes a five-year capacity plan. You can adjust the script for a one-year,
two-year, or three-year forecast by changing the “year5” references in the top section of the code.

Listing 5-20. lst05-20-db-growth-forexa.sql

SYS @ PROD> select curralloc,grate,year5,
decode(upper('&&hp_or_hc'),'HP','High Performance','HC','High Capacity','High Performance')
drivetype,
decode(upper('&&asm_redundancy'),'NORMAL','Normal','HIGH','High','Normal') asmredund,
case
when upper('&&hp_or_hc') = 'HP' then
case
when upper('&&asm_redundancy') = 'NORMAL' then
case
when year5 < (10547*&&pct_for_data) then 'Quarter Rack'