It is a prototype of Clickhouse cluster deployment to test storing different tables on different hosts but with the ability to use them jointly.
2 clusters with one shard and 2 replicas:
cluster_first
1.1.ch1
1.2.ch2
cluster_second
2.1.ch3
2.2.ch4
Run the cluster
docker-compose up
Populate data
./populate_databases.sh
Execute select request joining data from both clusters
./invoke_sql.sh ch1 cluster_first_join_request.sql
The solution is based on the configuration defining 2 different clusters and Distributed
engine.
Distributed engine allows us to create a virtual table in which data is actually stored on another cluster and processed too.