Q: I assume that the indexing in Exadata is not as important due to cell storage. Generally speaking, “full table scans” are not to be avoided as with non-Exadata databases. Exadata has numerous mechanisms which make full table scans preferable under many circumstances. What is the best way to optimize SQL for Exadata? Should I use the Exadata edition of Toad for instead of Toad for Oracle?
A: Yes, Toad DBA Suite for Oracle – Exadata Edition knows about Smart Scans, offloading, storage indexes and so on. Also, keep in mind that explain plan does not always report the truth; it indicates only what qualifies as possible, not what Exadata actually did. Because the Exadata edition knows how to test whether these actually occurred, it is far smarter for optimizing in Exadata.