Whats the best way to quickly insert hundreds of millions of records to an indexed and partitioned table?

Hi!
I have a table 'table1' containing about 3.8 billion records. The table has range partions(weeky dates) and hash partitions( id). The thing is, i have realised that i would need to add about a 150 m records to each partition(30 in total). Every time i insert data for a single partition, it seems to take about 20 hours before it finishes. i know its looking up the entire table before it inserts a record and inserting to a specific partition will improve performance;
How do i insert to a specific partition? How do i disable the indexes for that partition and rebuild them after inserting?
Will it be faster if i disabled all the partitions for the entire table, do the insertions and rebuild them later rather than partition by partition?
If i drop all Indexes, what's the estimated time will it take to rebuild them (the table will have grown to a 5 billion records table)? Let me know the best way of going about all these please.

I know Oracle 10g gathers statistics from the box every night, but how can i update the statistics of the table as i wish - Like after insertion to each partition to make sure it worked? I tried

EXEC dbms_stats.gather_table_stats( 'schema_name','Table1',cascade=>TRUE);

but got an error in SQL Developer(04098.00000  trigger %s.%s is invalid and failed re-validation).
Is this right way of generating the statistics? When i ran the same command on SQL *Plus (and logged on as DBA), it shows a busy Cursor. It has been running for sometime now, does it mean the procedure is correct? The error came immediately in SQL developer but i have seen nothing in SQL *Plus -other than the busy cursor, and am not sure whether it's executing or hunged. How quickly does an error show up in SQL *Plus compared to SQL Developer?

Any other ideas unknown to guy relatively knew to oracle that would help improve access speeds to this table will be greatly appreciated.

Thanks