| id | datasource |
|---|---|
| title | Datasources |
A data source is the Apache Druid equivalent of a database table. However, a query can also masquerade as a data source, providing subquery-like functionality. Query data sources are currently supported only by GroupBy queries.
The table data source is the most common type. It's represented by a string, or by the full structure:
{
"type": "table",
"name": "<string_value>"
}This data source unions two or more table data sources.
{
"type": "union",
"dataSources": ["<string_value1>", "<string_value2>", "<string_value3>", ... ]
}Note that the data sources being unioned should have the same schema. Union Queries should be always sent to a Broker/Router process and are NOT supported directly by the Historical processes.
This is used for nested groupBys and is only currently supported for groupBys.
{
"type": "query",
"query": {
"type": "groupBy",
...
}
}