I have table in which data is like this
Mapping Relationship
1 task_id is mapped to multiple task_detail_id (1 to Many)
x
Task_detail_id Task_id Creation_date
1 20 2020-05-02 20:28:23.354
2 21 2020-05-02 20:28:23.354
3 22 2020-05-02 19:28:23.354
4 22 2020-05-02 18:28:23.354
5 22 2020-05-02 17:28:23.354
6 22 2020-05-02 16:28:23.354
7 22 2020-05-02 15:28:23.354
8 23 2020-05-02 10:28:23.354
9 24 2020-05-02 09:28:23.354
10 24 2020-05-02 08:28:23.354
11 24 2020-05-02 07:28:23.354
What I want is to traverse the table and fetch record as if same task_id exist more than 2 times then fetch top 2 (latest) records for that task_id
Sample Output
Task_detail_id Task_id Creation_date
1 20 2020-05-02 20:28:23.354
2 21 2020-05-02 20:28:23.354
3 22 2020-05-02 19:28:23.354
4 22 2020-05-02 18:28:23.354
8 23 2020-05-02 10:28:23.354
9 24 2020-05-02 09:28:23.354
10 24 2020-05-02 08:28:23.354
Advertisement
Answer
This is typically done with a window function:
select task_detail_id, task_id, creation_date
from (
select task_detail_id, task_id, creation_date,
row_number() over (partition by task_id order by creation_date desc) as rn
from data
) t
where rn <= 2
order by task_detail_id, task_id;