menshikh-iv / clickhouse-replication-example

Create an example replicated + distributed dataset using Clickhouse

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Play with cluster

Cleanup old data

If you're Tim, you need to do these first, because you tried to run previous version

sudo rm -rf data/*

Up cluster

docker-compose up

Check that clickhouse are up

docker exec -ti clickhouse-replication-example_ch-client_1 /bin/bash
clickhouse-client --host ch1
select 1;

Create dabase on cluster

CREATE DATABASE IF NOT EXISTS testdb ON CLUSTER '{cluster}';

Create distributed table on cluster

CREATE TABLE IF NOT EXISTS testdb.test ON CLUSTER '{cluster}'
(
    userid UInt32,
    timestamp DateTime DEFAULT now()
) ENGINE = ReplicatedMergeTree('/clickhouse/tables/{cluster}/{shard}/default/test', '{replica}')
PARTITION BY toYYYYMM(timestamp)
ORDER BY (timestamp, userid)

Insert smth into every replica

docker exec -ti clickhouse-replication-example_ch-client_1 /bin/bash -c 'clickhouse-client --host ch1 --query "INSERT INTO testdb.test(userid) VALUES (1)(2)(3)(4)(5)"'
docker exec -ti clickhouse-replication-example_ch-client_1 /bin/bash -c 'clickhouse-client --host ch2 --query "INSERT INTO testdb.test(userid) VALUES (1)(2)(3)(4)(5)"'
docker exec -ti clickhouse-replication-example_ch-client_1 /bin/bash -c 'clickhouse-client --host ch3 --query "INSERT INTO testdb.test(userid) VALUES (1)(2)(3)(4)(5)"'

Check that we have same # of rows on each replica

docker exec -ti clickhouse-replication-example_ch-client_1 /bin/bash -c 'clickhouse-client --host ch1 --query "SELECT COUNT(*) FROM testdb.test"'
docker exec -ti clickhouse-replication-example_ch-client_1 /bin/bash -c 'clickhouse-client --host ch1 --query "SELECT COUNT(*) FROM testdb.test"'
docker exec -ti clickhouse-replication-example_ch-client_1 /bin/bash -c 'clickhouse-client --host ch1 --query "SELECT COUNT(*) FROM testdb.test"'

About

Create an example replicated + distributed dataset using Clickhouse