Implementation:Apache Druid Supervisor Lifecycle Api

Knowledge Sources	Apache Druid Druid Supervisor API
Domains	Streaming_Ingestion, Operations
Last Updated	2026-02-10 00:00 GMT

Overview

Concrete API integration for managing streaming supervisor lifecycle operations through the Druid Supervisor REST API.

Description

The supervisor lifecycle management spans two main components:

SupervisorsView (L590-L757): Renders action dialogs for individual supervisor operations using AsyncActionDialog for confirmation and execution. Actions include resume, suspend, hard reset, terminate, and set offsets.
SupervisorResetOffsetsDialog (L97-L246): Specialized dialog for manually setting consumer offsets/sequence numbers per partition, with type-specific labels (offsets for Kafka, sequence numbers for Kinesis).

Bulk operations (L1194-L1270) provide fleet-wide suspend, resume, and terminate for all supervisors.

Usage

These operations are triggered from the supervisor row context menu in the Supervisors view. Each operation shows a confirmation dialog via AsyncActionDialog before executing.

Code Reference

Source Location

Repository: Apache Druid
File: web-console/src/views/supervisors-view/supervisors-view.tsx (L590-L757, L1194-L1270)
File: web-console/src/dialogs/supervisor-reset-offsets-dialog/supervisor-reset-offsets-dialog.tsx (L97-L246)

Signature

// Individual supervisor actions:
Api.instance.post(`/druid/indexer/v1/supervisor/${id}/resume`)
Api.instance.post(`/druid/indexer/v1/supervisor/${id}/suspend`)
Api.instance.post(`/druid/indexer/v1/supervisor/${id}/reset`)
Api.instance.post(`/druid/indexer/v1/supervisor/${id}/terminate`)
Api.instance.post(`/druid/indexer/v1/supervisor/${id}/resetOffsets`, {
  type: supervisorType,  // 'kafka' or 'kinesis'
  partitions: {
    type: 'end',
    stream: streamName,
    partitionOffsetMap: { '0': 12345, '1': 12340 },
  },
})

// Bulk actions:
Api.instance.post('/druid/indexer/v1/supervisor/resumeAll')
Api.instance.post('/druid/indexer/v1/supervisor/suspendAll')
Api.instance.post('/druid/indexer/v1/supervisor/terminateAll')

Import

import { Api } from '../../singletons';

I/O Contract

Inputs

Name	Type	Required	Description
supervisorId	string	Yes	Target supervisor ID
supervisorType	string	No	'kafka' or 'kinesis' (for resetOffsets)
partitionOffsetMap	Record<string, number>	No	Partition-to-offset mapping (for resetOffsets)

Outputs

Name	Type	Description
Success/Error	toast notification	Action result shown via AppToaster

Usage Examples

Supervisor Operations

import { Api } from '../../singletons';

// Suspend a supervisor:
await Api.instance.post('/druid/indexer/v1/supervisor/my-kafka-supervisor/suspend');

// Resume:
await Api.instance.post('/druid/indexer/v1/supervisor/my-kafka-supervisor/resume');

// Reset offsets to specific positions:
await Api.instance.post('/druid/indexer/v1/supervisor/my-kafka-supervisor/resetOffsets', {
  type: 'kafka',
  partitions: {
    type: 'end',
    stream: 'events-topic',
    partitionOffsetMap: { '0': 1000, '1': 2000, '2': 1500 },
  },
});

// Terminate permanently:
await Api.instance.post('/druid/indexer/v1/supervisor/my-kafka-supervisor/terminate');

Related Pages

Implements Principle

Principle:Apache_Druid_Supervisor_Operational_Management

Requires Environment

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment