Hi Guys,
Please comment your valuable feedback on my queries.
Is it good to have single table with 90 partitions where in each partition would have at least 160 million records OR should I create separate tables (90 different tables) and each table should have at least 160 millions records?
Which approach is good for performance/management and maintenance point of view?
Another point came to my mind in case of first scenario is to create 90 different views(on single table with 90 partitions and each partition should have 160 million records) where in each view will fetch data from respective partitions. So basically creating views for each partitions.
The requirement is : There are 90 questions and each question has at least 160 millions records with different answers for a particular business date. We are also storing data for different business date with at least 160 millions records.
Please comment your thoughts and let me know which way should i proceed?
Thanks.
This is what I am planning.
Will partition this table on BUSINESS_DATE
Will have sub partition on QUESTION
Will also have index on ANSWER column. This column will be used in queries where clause.
-- Table description as follows
WH_QUESTION_ANSWER_F
(
BUSINESS_DATE,
QUESTION,
--------- OTHER FIELDS ----
--------- OTHER FIELDS ----
--------- OTHER FIELDS ----
--------- OTHER FIELDS ----
--------- OTHER FIELDS ----
ANSWER
)