ClickHouse
due to: Not able to find suitable driver for datasource
select * from jdbc('jdbc:mysql://rm-xxx.mysql.rds.aliyuncs.com:3306/davinci?useUnicode=true&user=xx&password=xxx','xxx','select * from cron_job');
Such SQL will report an error
select * from jdbc('mysql5', 'xxx', 'select * from cron_job')
Such SQL runs fine
Our mysql is dynamic, there is no way to configure it in datasources in advance
select * from jdbc('jdbc:mysql://rm-xxx.mysql.rds.aliyuncs.com:3306/xxx?useUnicode=true&user=xx&password=xxx','xxx','select * from cron_job');
Why does this SQL run fail ？
I follow this documentation using
select * from jdbc('jdbc:mysql://rm-xxx.mysql.rds.aliyuncs.com:3306/xxx?useUnicode=true&user=xx&password=xxx','xxx','select * from cron_job');
Why does this SQL run fail ？
Have you tried user:password@host_or_host_sublist as mentioned at here?
Update: and make sure user name and password are URL encoded.
select * from jdbc('mysql://davinci:[email protected]:3306/davinci','davinci','select * from cron_job');
I make sure user name and password are URL encoded.
select * from jdbc('jdbc:mysql://davinci:[email protected]:3306/davinci?useUnicode=true','davinci','select * from cron_job');
Do I have to define such a json file?
We have a lot of mysql, and they are added dynamically, there is no way to define the json file in advance
select * from jdbc('mysql5', 'davinci', 'select * from cron_job')
This SQL statement works fine，mysql5 is configured in advance in the mysql5.json file
but：We have a lot of mysql, and they are added dynamically, there is no way to define the json file in advance
Do I have to define such a json file?
No, you don't. But named datasource in general provides better security and performance. Also, you don't have to restart JDBC bridge each time adding a new JSON file - it will reload automatically.
If the json file is not defined, how to put the driver in the bridge?
You can put the driver in lib or any other directory and make sure it's included in CLASSPATH. Or, you can simply use docker-compose to start the environment and make sure you put JDBC drivers under $DRIVERS_DIR.
jdbc-bridge:
image: clickhouse/jdbc-bridge:2.0
hostname: jdbc-bridge
# In general you don't need to define any environment variable
# Below are all default settings just for demonstration
environment:
CONFIG_DIR: config # configuration directory
SERIAL_MODE: "false" # whether run query in serial mode or not
HTTPD_CONFIG_FILE: httpd.json # httpd configuration file
SERVER_CONFIG_FILE: server.json # server configuration file
VERTX_CONFIG_FILE: vertx.json # vertx configuration file
DATASOURCE_CONFIG_DIR: datasources # named datasource directory
DRIVER_DIR: drivers # driver directory
EXTENSION_DIR: extensions # extension directory
QUERY_CONFIG_DIR: queries # named query directory
CUSTOM_DRIVER_LOADER: "true" # whether use custom driver loader or not
JDBC_BRIDGE_JVM_OPTS: # use CPU and memory allocated by container
# You may want to keep datasources, queries, SQL scripts, and maybe drivers in a git repo
volumes:
- ./jdbc-bridge/config:/app/config
- ./jdbc-bridge/drivers:/app/drivers
- ./jdbc-bridge/scripts:/app/scripts
mem_limit: 512m
restart: always
@zhicwu Does bridge support k8s deployment? Is there a complete chart file？
Yes, but it should be able to come up with one quickly based on docker-compose file - please feel free to submit a pull request.