Merge Session/Event/Pageview CH

This commit is contained in:
Brian Cao 2022-09-12 09:55:34 -07:00
parent d377ef86e7
commit 52e036964b
18 changed files with 237 additions and 294 deletions

View file

@ -1,118 +1,50 @@
SET allow_experimental_object_type = 1;
-- Create Pageview
CREATE TABLE pageview
(
website_id UInt32,
session_uuid UUID,
created_at DateTime('UTC'),
url String,
referrer String
)
engine = MergeTree PRIMARY KEY (session_uuid, created_at)
ORDER BY (session_uuid, created_at)
SETTINGS index_granularity = 8192;
CREATE TABLE pageview_queue (
website_id UInt32,
session_uuid UUID,
created_at DateTime('UTC'),
url String,
referrer String
)
ENGINE = Kafka
SETTINGS kafka_broker_list = 'kafka1:19092,kafka2:19093,kafka3:19094', -- input broker list
kafka_topic_list = 'pageview',
kafka_group_name = 'pageview_consumer_group',
kafka_format = 'JSONEachRow',
kafka_max_block_size = 1048576,
kafka_skip_broken_messages = 1;
CREATE MATERIALIZED VIEW pageview_queue_mv TO pageview AS
SELECT website_id,
session_uuid,
created_at,
url,
referrer
FROM pageview_queue;
-- Create Session
CREATE TABLE session
(
session_uuid UUID,
website_id UInt32,
created_at DateTime('UTC'),
hostname LowCardinality(String),
browser LowCardinality(String),
os LowCardinality(String),
device LowCardinality(String),
screen LowCardinality(String),
language LowCardinality(String),
country LowCardinality(String)
)
engine = MergeTree PRIMARY KEY (session_uuid, created_at)
ORDER BY (session_uuid, created_at)
SETTINGS index_granularity = 8192;
CREATE TABLE session_queue (
session_uuid UUID,
website_id UInt32,
created_at DateTime('UTC'),
hostname LowCardinality(String),
browser LowCardinality(String),
os LowCardinality(String),
device LowCardinality(String),
screen LowCardinality(String),
language LowCardinality(String),
country LowCardinality(String)
)
ENGINE = Kafka
SETTINGS kafka_broker_list = 'kafka1:19092,kafka2:19093,kafka3:19094', -- input broker list
kafka_topic_list = 'session',
kafka_group_name = 'session_consumer_group',
kafka_format = 'JSONEachRow',
kafka_max_block_size = 1048576,
kafka_skip_broken_messages = 1;
CREATE MATERIALIZED VIEW session_queue_mv TO session AS
SELECT session_uuid,
website_id,
created_at,
hostname,
browser,
os,
device,
screen,
language,
country
FROM session_queue;
-- Create event
-- Create Event
CREATE TABLE event
(
event_uuid UUID,
website_id UInt32,
session_uuid UUID,
created_at DateTime('UTC'),
event_uuid Nullable(UUID),
--session
hostname LowCardinality(String),
browser LowCardinality(String),
os LowCardinality(String),
device LowCardinality(String),
screen LowCardinality(String),
language LowCardinality(String),
country LowCardinality(String),
--pageview
url String,
referrer String,
--event
event_name String,
event_data JSON
event_data JSON,
created_at DateTime('UTC')
)
engine = MergeTree PRIMARY KEY (event_uuid, created_at)
ORDER BY (event_uuid, created_at)
engine = MergeTree
ORDER BY (website_id, session_uuid, created_at)
SETTINGS index_granularity = 8192;
CREATE TABLE event_queue (
event_uuid UUID,
website_id UInt32,
session_uuid UUID,
created_at DateTime('UTC'),
event_uuid Nullable(UUID),
url String,
referrer String,
hostname LowCardinality(String),
browser LowCardinality(String),
os LowCardinality(String),
device LowCardinality(String),
screen LowCardinality(String),
language LowCardinality(String),
country LowCardinality(String),
event_name String,
event_data String
event_data String,
created_at DateTime('UTC')
)
ENGINE = Kafka
SETTINGS kafka_broker_list = 'kafka1:19092,kafka2:19093,kafka3:19094', -- input broker list
SETTINGS kafka_broker_list = 'domain:9092,domain:9093,domain:9094', -- input broker list
kafka_topic_list = 'event',
kafka_group_name = 'event_consumer_group',
kafka_format = 'JSONEachRow',
@ -120,11 +52,19 @@ SETTINGS kafka_broker_list = 'kafka1:19092,kafka2:19093,kafka3:19094', -- input
kafka_skip_broken_messages = 1;
CREATE MATERIALIZED VIEW event_queue_mv TO event AS
SELECT event_uuid,
website_id,
SELECT website_id,
session_uuid,
created_at,
event_uuid,
url,
referrer,
hostname,
browser,
os,
device,
screen,
language,
country,
event_name,
event_data
FROM event_queue;
event_data,
created_at
FROM event_queue;