Navigation Menu

Skip to content

zeromicro/cds

Repository files navigation

ClickHouse Data Synchromesh

Data syncing in golang for ClickHouse.

based on go-zero

ARCH

Data workflow of a typical data warehouse architecture

avatar

Design of Data Sync

Automatically synchronizing data from MySQL/MongoDB data source to ClickHouse cluster in real time(< 2min).

sync

start up

git clone https://github.com/zeromicro/cds.git
cd cds
make up

After the build , pay attention to check if any container exits abnormally.

click http://localhost:3414/cds.html to accessing the web interface.

using username and password below to login

user: admin@email.com
password: 123456

create table in ClickHouse for syncing MySQL(or MongoDB) data

Chose "create table" tab

1. Click "Target ClickHouse Database Info",
2. Click “connect”
3. Select the schema ("default") synchronized to Clickhouse
4. switch to "Data Source" 
5. MySQL input connection string root:root@tcp(mysql:3306)/test_mysql
or
MongoDB input connection string mongodb://mongo1:30001/test_mongo
6. Click “connect”
7. Select the table, such as the default example_mysql (MySQL) or example (MongoDB)
8. click "Generate create Table SQL" 
 Note: select the partition field as needed, here'PARTITION BY toYYYYMM()' can be deleted, or replaced with'PARTITION BY toYYYYMM(dt)'
9. . Click "send SQL to ClickHouse", and the successful execution will pop up below

One-time full data synchronization:

Chose "full sync" tab

1. Click "+" in the upper right corner
2. MySQL input connection string root:root@tcp(mysql:3306)/test_mysql
or
MongoDB input connection string mongodb://mongo1:30001/test_mongo
3. Click “connect”
4. Select the table, such as the default example_mysql (MySQL) or example (MongoDB)
5. Click "Target ClickHouse Database Info"
6. Click “connect”
7. Select the schema ("default") synchronized in Clickhouse
10. Click "Add", a pop-up below shows successful execution

check task status

Refresh the page

Confirm data in Clickhouse

image-20201118135156133

Turn on real-time incremental synchronization

Take mysql as an example

chose "Connector" tab

1. Click "+" in the upper right corner
2. MySQL input connection string root:root@tcp(mysql:3306)/test_mysql
3. select table 
4. Click "Add"

chose "Incremental Sync" tab

1. Click "+" in the upper right corner
2. MySQL input connection string root:root@tcp(mysql:3306)/test_mysql
3. select table 
4. Click "Target ClickHouse Database Info"
5. Click “connect”
6. Select the schema ("default") synchronized in Clickhouse
7. Click "Add", a pop-up below shows successful execution

refresh page

Verify incremental update

Execute the initialization database script again, you can insert 100000 rows of data again.

cd sit/docker/
sh ./init.sh

Verify the incremental data of mysql in clickhouse:

image-20201118135503830

clean up

To clean up all the docker containers started above and restore the initial state, you can :

cd cds
make down

only clean

cd cds
make docker_clean

data model in clickhouse

CDS中ClickHouse使用的建表方案

help

提问的智慧

如何有效的报告bug


if you like this project and want to support it,please star 🤝