
Appendix A. Data placement tests 111
Results
As you can see in Figure A-2, the elapsed times for Loads were equivalent for all tests (less
than 3% variance), regardless of placement on the arrays or the method used to ‘carve up’
the arrays.
However, enabling four paths compared to a single path per ESS logical disk improved the
elapsed time of the loads approximately 15%.
Figure A-2 Large versus small logical disks elapsed LOAD times
A tablescan was then run for each test setup using a query similar to the following
(Example A-1):
Example: A-1 Tablescan query
select * from test.rb2a where L_EXTENDEDPRICE < 0;
Again, as shown in Figure A-3, the position ...