/itmo-clickhouse-bigdata

Docker-compose based ClickHouse sharded, replicated installation

Primary LanguageShell

ITMO BigData course 2020

Getting started

This repo contains a preset for starting 4 clickhouse instances, 1 zookeeper and a client to connect to them. 4 instances are orginized into 2 shards with 2 replicas in each of them.

  1. install docker and docker-compose on your machine
  2. configure docker to have reasonable resorces
  3. launch server:
    1. go to /server directory
    2. run $ docker-compose up --build
  4. launch client:
    1. go to /client
    2. run $ docker-compose run client

There is a convenince script ./clean.sh, that is usefull if you need a fresh start

Uploading data

As everything has been set up, you should have a container with clickhouse-client set up and running. The whole /client directory is mounted to the container, so you can put what's nececcary in there. As part of the assignment there is predefined set of tables and queries that are task specific.