The E-MapReduce Druid cluster integrates the Superset tool, which is integrated with E-MapReduce Druid and supports a variety of relational databases. Because E-MapReduce Druid supports SQL, you can access E-MapReduce Druid through Superset in two ways: Apache Druid's native query language or SQL.
Superset is installed in emr-header-1 by default and does not support high availability at present. Before you use this tool, make sure that your host can access emr-header-1. You can connect to the host by establishing the Connect to the master node of an EMR cluster in SSH mode.
- Log on to the Superset
Enter http://emr-header-1:18088 in your browser to go to the Superset logon page. The default username is admin and the default password is admin. When you log on for the first time, we strongly recommend changing your password.
- Add an E-MapReduce Druid cluster
The English interface is displayed by default. You can select the appropriate language by clicking the flag icon in the upper-right corner. In the menu bar along the top, select to add an E-MapReduce Druid cluster.
Configure the addresses of the coordinator and broker. The default port number in E-MapReduce is the corresponding open source port number with "1" added in front. For example, if the open-source broker port number is 8082, the port number in E-MapReduce is 18082.
- Refresh or add a new data source
After adding the Druid cluster, you can click to add new data sources. The data sources on the E-MapReduce Druid cluster loaded automatically.
You can also customize a new data source by clicking on the interface. (This operation is equivalent to writing a JSON file for data source ingestion.)
Enter the necessary information for custom data sources, and save it.
Click the second of the three small icons on the left side to edit the data source. Enter the appropriate information, such as dimensions and metrics.
- Query E-MapReduce Druid
After the data source has been added successfully, click it to go to the details page.
- (Optional) Use E-MapReduce Druid as a database
Superset provides SQLAlchemy to support a wide variety of databases with various dialects, as shown in the following figure.
Superset also supports accessing E-MapReduce Druid in this way. The corresponding SQLAlchemy URI of E-MapReduce Druid is druid://emr-header-1:18082/druid/v2/sql. When you add E-MapReduce Druid as a database, check the "Expose in SQL Lab" check box.
You cannow use SQL to query in the SQL toolkit.
