Merge pull request 'migration to ckickhouse' (#26) from OS-220 into main

Reviewed-on: https://gitea.de-infra.servehttp.com/de-infrastructure/DWH/pulls/26
Reviewed-by: aleksey bogatyrev <abogatyrev@example.com>
This commit is contained in:
2026-01-07 21:13:50 +04:00
2 changed files with 279 additions and 1 deletions
+278
View File
@@ -0,0 +1,278 @@
CREATE TABLE ybirk.core_user ON CLUSTER clickhouse_cluster
(
user_id Int32,
name String,
age Int32,
email String,
registration_date Date
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/ybirk/core_user',
'{replica}'
)
ORDER BY user_id;
CREATE TABLE ybirk.core_course ON CLUSTER clickhouse_cluster
(
course_id Int32,
title String,
category String
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/ybirk/core_course',
'{replica}'
)
ORDER BY course_id;
CREATE TABLE ybirk.core_lesson ON CLUSTER clickhouse_cluster
(
lesson_id Int32,
title String,
duration_min Int32,
course_id Int32
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/ybirk/core_lesson',
'{replica}'
)
ORDER BY lesson_id;
CREATE TABLE ybirk.core_fact_lesson_views ON CLUSTER clickhouse_cluster
(
user_id Int32,
lesson_id Int32,
course_id Int32,
viewed_at DateTime64(3)
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/ybirk/core_fact_lesson_views',
'{replica}'
)
ORDER BY user_id;
CREATE TABLE ybirk.core_enrollments ON CLUSTER clickhouse_cluster
(
id Int32,
user_id Int32,
course_id Int32,
enrolled_at Date
)
ENGINE = ReplicatedMergeTree(
'/clickhouse/tables/{shard}/ybirk/core_enrollments',
'{replica}'
)
ORDER BY user_id;
--------------------------------
CREATE TABLE ybirk.inactive_users_summary ON CLUSTER clickhouse_cluster
(
user_id UInt32,
name String,
email String,
age UInt8,
registered_courses_count UInt32,
registration_date DateTime64(3, 'UTC')
)
ENGINE = ReplicatedAggregatingMergeTree(
'/clickhouse/tables/{shard}/ybirk/inactive_users_summary',
'{replica}'
)
ORDER BY user_id;
CREATE TABLE ybirk.lesson_popularity_summary ON CLUSTER clickhouse_cluster
(
lesson_id UInt32,
lesson_title String,
course_id UInt32,
course_title String,
total_views UInt32,
unique_users UInt32,
first_view DateTime64(3, 'UTC'),
last_view DateTime64(3, 'UTC')
)
ENGINE = ReplicatedAggregatingMergeTree(
'/clickhouse/tables/{shard}/ybirk/lesson_popularity_summary',
'{replica}'
)
ORDER BY (lesson_id, course_id);
CREATE TABLE ybirk.course_completion_rate ON CLUSTER clickhouse_cluster
(
user_id UInt32,
user_name String,
course_id UInt32,
course_title String,
lessons_in_course UInt32,
lessons_viewed UInt32,
completion_rate Float32
)
ENGINE = ReplicatedAggregatingMergeTree(
'/clickhouse/tables/{shard}/ybirk/course_completion_rate',
'{replica}'
)
ORDER BY (user_id, course_id);
--------------------------------------
CREATE MATERIALIZED VIEW ybirk.mv_lesson_popularity_summary ON CLUSTER clickhouse_cluster
TO ybirk.lesson_popularity_summary
AS
SELECT
flv.lesson_id AS lesson_id,
l.title AS lesson_title,
flv.course_id AS course_id,
c.title AS course_title,
count() AS total_views,
uniqExact(flv.user_id) AS unique_users,
min(flv.viewed_at) AS first_view,
max(flv.viewed_at) AS last_view
FROM ybirk.core_fact_lesson_views AS flv
JOIN ybirk.core_lesson AS l ON flv.lesson_id = l.lesson_id
JOIN ybirk.core_course AS c ON flv.course_id = c.course_id
GROUP BY
flv.lesson_id,
l.title,
flv.course_id,
c.title;
DROP VIEW IF EXISTS ybirk.mv_inactive_users_summary ON CLUSTER clickhouse_cluster;
CREATE MATERIALIZED VIEW ybirk.mv_inactive_users_summary ON CLUSTER clickhouse_cluster
TO ybirk.inactive_users_summary
AS
SELECT
u.user_id AS user_id,
u.name AS name,
u.email AS email,
u.age AS age,
countDistinct(e.course_id) AS registered_courses_count,
u.registration_date AS registration_date
FROM ybirk.core_user AS u
LEFT JOIN ybirk.core_enrollments AS e ON u.user_id = e.user_id
LEFT ANTI JOIN ybirk.core_fact_lesson_views AS flv ON u.user_id = flv.user_id
GROUP BY
u.user_id,
u.name,
u.email,
u.age,
u.registration_date;
CREATE MATERIALIZED VIEW ybirk.mv_course_completion_rate ON CLUSTER clickhouse_cluster
TO ybirk.course_completion_rate
AS
SELECT
u.user_id AS user_id,
u.name AS user_name,
c.course_id AS course_id,
c.title AS course_title,
countDistinct(l.lesson_id) AS lessons_in_course,
countDistinct(flv.lesson_id) AS lessons_viewed,
if(countDistinct(l.lesson_id) = 0, 0,
round(countDistinct(flv.lesson_id) * 1.0 / countDistinct(l.lesson_id), 2)
) AS completion_rate
FROM ybirk.core_enrollments AS e
JOIN ybirk.core_user AS u ON e.user_id = u.user_id
JOIN ybirk.core_course AS c ON e.course_id = c.course_id
JOIN ybirk.core_lesson AS l ON c.course_id = l.course_id
LEFT JOIN ybirk.core_fact_lesson_views AS flv
ON flv.user_id = e.user_id
AND flv.course_id = e.course_id
GROUP BY
u.user_id,
u.name,
c.course_id,
c.title;
----------------------------------------------
INSERT INTO ybirk.core_user (user_id, name, age, email, registration_date) VALUES
(1, 'Alice', 25, 'alice@mail.com', '2023-01-10'),
(2, 'Bob', 30, 'bob@gmail.com', '2023-02-05'),
(3, 'Charlie', 22, 'charlie@mail.com', '2023-02-20'),
(4, 'Diana', 28, 'diana@mail.com', '2023-03-01'),
(5, 'Ethan', 35, 'ethan@gmail.com', '2023-03-10');
INSERT INTO ybirk.core_course (course_id, title, category) VALUES
(1, 'SQL для начинающих', 'data'),
(2, 'Python для анализа данных', 'programming'),
(3, 'BI с нуля', 'business');
INSERT INTO ybirk.core_lesson (lesson_id, course_id, title, duration_min) VALUES
(1, 1, 'SELECT и FROM', 10),
(2, 1, 'JOIN', 15),
(3, 2, 'Pandas', 20),
(4, 3, 'Основы BI', 12),
(5, 3, 'Метрики и дашборды', 18);
INSERT INTO ybirk.core_fact_lesson_views (user_id, lesson_id, course_id, viewed_at) VALUES
(1, 1, 1, '2023-01-16 10:00:00'),
(1, 2, 1, '2023-01-16 10:15:00'),
(2, 1, 1, '2023-01-21 09:00:00'),
(3, 3, 2, '2023-03-06 12:00:00'),
(4, 4, 3, '2023-04-02 10:00:00'),
(4, 5, 3, '2023-04-03 10:00:00');
INSERT INTO ybirk.core_enrollments (id, user_id, course_id, enrolled_at) VALUES
(1, 1, 1, '2023-01-15'),
(2, 1, 2, '2023-02-01'),
(3, 2, 1, '2023-01-20'),
(4, 3, 2, '2023-03-05'),
(5, 4, 3, '2023-04-01');
--------------------------------------------------
--Примеры
SELECT
lesson_title,
course_title,
total_views
FROM ybirk.mv_lesson_popularity_summary
ORDER BY total_views DESC
LIMIT 5;
SELECT
name,
email,
registered_courses_count
FROM ybirk.inactive_users_summary
WHERE registered_courses_count > 0
ORDER BY registration_date DESC;
SELECT
course_title,
AVG(completion_rate) AS avg_completion
FROM ybirk.course_completion_rate
GROUP BY course_title
ORDER BY avg_completion DESC
LIMIT 5;
-- ЗАпросы
-- самый популярный курс
SELECT
course_title,
SUM(total_views) as sum_views
FROM ybirk.mv_lesson_popularity_summary
GROUP BY course_title
ORDER BY sum_views DESC;
-- Кто застрял
SELECT
user_name,
course_title,
completion_rate
FROM ybirk.mv_course_completion_rate
WHERE completion_rate > 0 AND completion_rate < 1.0;
-- тест3
SELECT
t1.user_name,
t1.course_title,
t2.email
FROM ybirk.mv_course_completion_rate AS t1
JOIN ybirk.mv_inactive_users_summary AS t2
ON t1.user_id = t2.user_id
WHERE t1.completion_rate = 1.0;