Creating a Druid Database

This page covers the configuration specific to Druid — its Database Mode and any engine-specific settings shown below. The rest of the creation flow — opening the wizard, namespace and name, version, machine profile, storage, and optional features — is the same for every engine and is documented in Common Steps.

Database Mode

Druid is always deployed as a Topology of role-separated process tiers. Configure each tier’s node count and resources independently.

Druid topology showing Coordinators, Overlords/MiddleManagers, Historicals, and Brokers panels

NodeDescription
CoordinatorsManage data availability, segment balancing, and cluster coordination.
MiddleManagersRun ingestion tasks and index new data.
HistoricalsStore and serve queryable historical segments.
BrokersReceive queries and route them to the appropriate Historicals/MiddleManagers.

Each tier has its own Number of Replicas, Storage size, Machine, CPU, and Memory fields. Druid also requires dependent metadata storage and deep storage, configured by the platform.

Druid Dependencies

Druid relies on external dependencies for metadata, deep storage, and coordination. Each can be provisioned by the platform or pointed at an existing, externally-managed instance.

Druid deep storage, metadata storage, and ZooKeeper configuration

Deep Storage — Durable storage for Druid segments.

FieldDescription
TypeDeep storage backend: s3, google, azure, or hdfs.
Config SecretSecret holding the credentials/configuration for the chosen backend.

Metadata Storage — Relational database holding Druid metadata.

FieldDescription
TypeMySQL or Postgres.
Externally ManagedToggle on to reference an existing database instead of provisioning one.
Namespace / NameReference to the metadata database when externally managed.

ZooKeeper — Coordination service for the Druid cluster.

FieldDescription
Externally ManagedToggle on to reference an existing ZooKeeper instead of provisioning one.
Namespace / NameReference to the ZooKeeper instance when externally managed.

Create a Druid Database

  1. Open the wizard and select Druid — see Getting Started and Select a Database Type.
  2. Set the namespace and name.
  3. Pick the database version and the Database Mode described above, then set the machine profile and storage — see Configure the Database.
  4. Optionally configure Advanced Configuration (labels, deletion policy, credentials, point-in-time recovery) and Additional Options (monitoring, backup, TLS, gateway).
  5. Click Deploy.