I have read few articles about table partioning but still I am bit confused on its uses.
My case is as follows.
I have a big Table TA containing 10 millions record approx and is daily loaded with 30-40K records.
Table TA contains many column including date field and one more critical column is project# which is varchar.
Now I have an option of partioning table TA on date field.
But if I see my query mainly I am going to fetch data for one date alone and since date field is indexed so data fetching is not a big prob.
Similary qry contains project# also in “WHERE” condion with “IN” clause i.e I have to give N number of project# as input through “IN” clause.
Now suggest me how should I proceed.
If you will always read the full daily data, do this:
Step 0) Drop the global index on date, you won’t need it with partitions.
Step 1) Create daily partitions
Step 2) Create a secondary partition local index on project
If I misundertood you, and you won’t read the full daily partition, just parts of it, then leave your schema as it is.