Druid Connector

Overview

The Druid Connector allows access to Druid data from Presto. This document describes how to setup the Druid Connector to run SQL queries against Druid.

Note

It is highly recommended to use Druid 0.17.0 or later.

Configuration

To configure the Druid connector, create a catalog properties file etc/catalog/druid.properties with the following contents, replacing the properties as appropriate:

  1. connector.name=druid
  2. druid.coordinator-url=hostname:port
  3. druid.broker-url=hostname:port
  4. druid.schema-name=schema
  5. druid.compute-pushdown-enabled=true

Configuration Properties

The following configuration properties are available:

Property Name

Description

druid.coordinator-url

Druid coordinator url.

druid.broker-url

Druid broker url.

druid.schema-name

Druid schema name.

druid.compute-pushdown-enabled

Whether to pushdown all query processing to Druid.

druid.case-insensitive-name-matching

Match dataset and table names case-insensitively.

druid.case-insensitive-name-matching.cache-ttl

Duration for which remote dataset and table names will be cached. Set to 0ms to disable the cache

druid.coordinator-url

Druid coordinator url, e.g. localhost:8081.

druid.broker-url

Druid broker url, e.g. localhost:8082.

druid.schema-name

Druid schema name.

This property is optional; the default is druid.

druid.compute-pushdown-enabled

Whether to pushdown all query processing to Druid.

the default is false.

druid.case-insensitive-name-matching

Match dataset and table names case-insensitively.

The default is false.

druid.case-insensitive-name-matching.cache-ttl

Duration for which remote dataset and table names will be cached. Set to 0ms to disable the cache.

The default is 1m.

Data Types

The data type mappings are as follows:

Druid

Presto

VARCHAR

VARCHAR

BIGINT

BIGINT

DOUBLE

DOUBLE

FLOAT

REAL

TIMESTAMP

TIMESTAMP

(others)

(unsupported)