如何遍历 SQL 中的分区，收集特定周的行并使用收集的行创建新表

Question

My data is organized in partitions. Data is partitioned by the year, month, and day when the records were received by the servers. The dataset contains a column with the timestamp that records when an event happened and another one with the timestamp of when the data corresponding was received in the servers.

I need to go to each partition from 06/2021 to 06/2022, collect all rows that correspond to events that happened during the week of Jan. 18, 2021 to Jan. 24, 2021, and create a new table with the rows collected.

This is an example of how my datase looks like:

year	month	day	event_timestamp	server_timestamp
2021	07	01	2021-01-19 01:48:20.000	2021-07-01 01:48:20.000
2022	04	09	2022-04-08 01:48:20.000	2022-04-09 01:48:20.000
2023	01	19	2023-01-08 01:48:20.000	2023-01-19 01:48:20.000
2022	02	21	2022-01-09 01:48:20.000	2022-02-21 01:48:20.000
2021	08	05	2021-01-23 01:48:20.000	2021-08-05 01:48:20.000

What is the best way to solve this using SQL?

Answer 1

It seems like you do not need the columns year, month and day, since you have got the server_timestamp and you do not need a for loop?!

If i understood your question correctly, the answer could look something like that:

create table new_table(
year int,
month nvarchr(2),
day nvarchar (2),
event_timestamp timestamp,
server_timestamp timestamp
)


select year, month, day, event_timestamp, server_timestamp
into new_table
from dataset
where server_timestamp >= 2021-06-01 00:00:00.000 
  and server_timestamp < 2022-07-01 00:00:00.000
  and event_timestamp >= 2021-01-18 00:00:00.000
  and event_timestamp < 2021-01-25 00:00:00.000