How I can group the data for each hour of each day in a table?

Hi everybody!
Please tell me how I can group the data for each hour of each day in a table?
I have a table with the name sales. This table contains the data:
sale_id - id of the sale.
sale_start - when the sale process was started (timestamp).
sale_canceled - when the sale was canceled(timestamp).
sale_completed - when the sale was made (timestamp).
table

...
I need to calculate how many sales have been made and are available at every hour of time during the week.

In general, now I have a postgresql-query and it works when I use "LIMIT". Otherwise, the request crashes.
Postgresql-query:

WITH s AS (SELECT date_trunc('hour', '2022-04-22'::date) - generate_series(0, 24 * 7) * interval '1 hour' AS h_slot
FROM sales),
dt AS (SELECT DATE_TRUNC('hour', "sale_start") AS sale_start,
       DATE_TRUNC('hour', "sale_canceled") AS sale_canceled,
       DATE_TRUNC('hour', "sale_completed") AS sale_completed,
       FROM sales
       WHERE "sale_start" >= '2022-08-10'::date - INTERVAL '1 week'
       AND "sale_start" <= '2023-08-10'),
SELECT s.h_slot,
       COUNT(*) FILTER (WHERE dt.sale_completed = s.h_slot) AS completed_count,
       COUNT(*) FILTER (WHERE s.h_slot >= dt.sale_start 
       AND s.h_slot <= dt.sale_canceled) AS available_count
FROM s CROSS JOIN dt
GROUP BY s.h_slot
ORDER BY s.h_slot

Pay attention to the fragment:

WITH s AS (
  SELECT date_trunc('hour', '2022-04-22'::date) 
            - generate_series(0, 24 * 7) * interval '1 hour' AS h_slot
  FROM sales
)

There you generate series of size= 7 * 24 *(count of records in table sales).

Then cross join with dt (sales).
Total number of records is
7 * 24 *(count of records in table sales) * (count of records in interval (week) of sales)

First, remove FROM Sales.
This may solve your problem.

WITH s AS (
  SELECT date_trunc('hour', '2022-04-22'::date) 
             - generate_series(0, 24 * 7) * interval '1 hour' AS h_slot
)

For example, try this.

WITH s AS (SELECT date_trunc('hour', '2022-8-10'::date) - generate_series(0, 24 * 7) * interval '1 hour' AS h_slot
)
,avlbl AS (
  SELECT DATE_TRUNC('hour', "sale_start") intrvl
      ,count(*) as cnt_availible 
       FROM sales
       WHERE "sale_start" >= '2022-08-10'::date - INTERVAL '1 week'
       AND "sale_start" <= '2023-08-10'
  group by DATE_TRUNC('hour', "sale_start")
)
,dtgr AS (
  SELECT coalesce(DATE_TRUNC('hour', "sale_cancelled")
                   ,DATE_TRUNC('hour', "sale_completed")) dthour
         ,count(sale_cancelled) cnt_cancelled
         ,count(sale_completed) cnt_completed
       FROM sales
       WHERE "sale_start" >= '2022-08-10'::date - INTERVAL '1 week'
       AND "sale_start" <= '2023-08-10'
  group by coalesce(DATE_TRUNC('hour', "sale_cancelled")
                   ,DATE_TRUNC('hour', "sale_completed"))
)
SELECT * 
FROM s 
left JOIN dtgr d on d.dthour=s.h_slot
left join avlbl a on a.intrvl=s.h_slot
ORDER BY s.h_slot