如何隨機生成一個大表的資料

create table test_table(a number,b varchar2(100),c varchar2(10),d date);
insert into test_table
select rownum as a,
       dbms_random.string('a', 50),
       dbms_random.string('u', 3) as c,
       to_date('2016-01-01', 'yyyy-mm-dd') + dbms_random.value * 360 as d
  from dual
connect by level <= 4000000;  
commit;


建立索引：
create index idx_node_c_a on test_table(c,a);
create index idx_node_c on test_table(c);
create index uk_node on test_table(a);