YARN Data Retrieval

Acceldata Pulse APIs enable you to programmatically retrieve YARN and application-level data for use in custom dashboards, reports, data pipelines, and operational workflows.

You can use these APIs to:

Authenticate with Pulse.
Retrieve Application Explorer data.
Retrieve YARN queue and capacity data.
Monitor resource allocation and cluster utilization.
Integrate Pulse data with external systems.

Pulse exposes these capabilities through GraphQL operations available at the /graphql endpoint.

Before You Begin

Before invoking the APIs:

Obtain a valid JWT token and role value.
Identify the Pulse URL for your environment.
Identify the monitor group when required by the GraphQL operation.
Ensure that the authenticated user has access to YARN data.

For authentication instructions, see Authenticate with Pulse APIs

Call the API Endpoint

All API requests go through the GraphQL endpoint:

POST https://<your-pulse-url>/graphql
Example:

https://trial.acceldata.dev/graphql

Retrieve Application Explorer Data

What You Can Get

You can fetch the same data you see in the Application Explorer UI, including:

Application ID
Type
State
Name
User
Queue
Progress
Start Time
Elapsed Time
Finished Time
Priority
Cluster Usage (%)
Queue Usage (%)
Application Tags
Allocated Memory (MB)
Allocated vCores
Memory Seconds
vCore Seconds

Use This GraphQL Operation

YarnAppDataForExplorer

Example API Request

curl 'https://<your-pulse-url>/graphql' \
-H 'content-type: application/json' \
-H 'Role: <role-name>' \
-b 'jwt=<your-jwt-token>' \
--data-raw '{
"operationName": "YarnAppDataForExplorer",
"variables": {
"tableMetadata": {
"pageNo": 1,
"pageSize": 20,
"start": "<start_timestamp>",
"end": "<end_timestamp>"
}
},
"query": "query YarnAppDataForExplorer($tableMetadata: TableMetadata!) { yarnAppDataForExplorer(tableMetadata: $tableMetadata) { id name queue user state applicationType progress startedTime finishedTime elapsedTime priority clusterUsagePercentage queueUsagePercentage applicationTags allocatedMB allocatedVCores memorySeconds vcoreSeconds } }"
}'
Replace:

<your-jwt-token> with the JWT token obtained during authentication.
<role-name> with a role assigned to the authenticated user (for example, ADMIN, SUPER_ADMIN, or another custom role).

Example Response

{
"data": {
"yarnAppDataForExplorer": [
{
"id": "application_123456",
"name": "Spark Job Example",
"queue": "default",
"user": "hadoop_user",
"state": "FINISHED",
"applicationType": "SPARK",
"progress": 100,
"startedTime": 1774809183525,
"finishedTime": 1774812783525,
"elapsedTime": 3600000,
"priority": 1,
"clusterUsagePercentage": 12.5,
"queueUsagePercentage": 45.2,
"applicationTags": ["etl", "daily"],
"allocatedMB": 4096,
"allocatedVCores": 4,
"memorySeconds": 123456789,
"vcoreSeconds": 98765432
}
]
}
}

Explore YARN Data

What You Can Access

You can also retrieve YARN-related insights such as:

Queue List

All queues and hierarchy
Capacity allocation

Example Request:

curl -sS -X POST '<Pulse_UI_URL>/graphql' \
-H 'Content-Type: application/json' \
-H 'MonitorGroup: name_here'\
-H 'role: encoded role here' \
-H 'Cookie: jwt=PASTE_YOUR_FULL_JWT_COOKIE_HERE' \
-d '{
"operationName": "YarnQueueSnapshotDistinctQueue",
"variables": {
"start": "1775544117899",
"end": "1775630517899"
},
"query": "query YarnQueueSnapshotDistinctQueue($start: String, $end: String) {\n data: yarnQueueSnapshotDistinctQueue(start: $start, end: $end)\n}\n"
}'
Example Response:

{"data":{"data":["root","root.default"]}}
Queue Details

Capacity and usage

Example Request:

BASE='http://10.100.10.18:4000'
curl -sS -X POST "$BASE/graphql" \
-H 'Content-Type: application/json' \
-H 'MonitorGroup: odp_usopp' \
-H 'role: eyJuYW1lIjoiYWRtaW4iLCJjbHVzdGVycyI6W119' \
-H 'Cookie: jwt=PASTE_YOUR_FULL_SIGNED_JWT_COOKIE_VALUE' \
-d '{
"operationName": "YarnQueueSnapshotTimeHistogram",
"query": "query YarnQueueSnapshotTimeHistogram($qNames: [String]!, $start: String!, $end: String!, $metrics: [String]!, $accumulators: [String]!, $nodeLabel: String!, $step: Int) { yarnQueueSnapshotTimeHistogram(qNames: $qNames, start: $start, end: $end, metrics: $metrics, accumulators: $accumulators, nodeLabel: $nodeLabel, step: $step) { _id vcoresUsed_avg memoryUsed_avg usedCapacity_avg absoluteUsedCapacity_avg } }",
"variables": {
"qNames": ["root.default"],
"start": "1775544117899",
"end": "1775630517899",
"metrics": ["vcoresUsed", "memoryUsed", "usedCapacity"],
"accumulators": ["avg", "max"],
"nodeLabel": "",
"step": 1800
}
}'
Example Response:

{
"data": {
"yarnQueueSnapshotTimeHistogram": [
{
"_id": "1775543400000",
"vcoresUsed_avg": "1",
"memoryUsed_avg": "1024",
"usedCapacity_avg": "3.1746032",
"absoluteUsedCapacity_avg": null
},
{
"_id": "1775545200000",
"vcoresUsed_avg": "1",
"memoryUsed_avg": "1024",
"usedCapacity_avg": "3.1746032",
"absoluteUsedCapacity_avg": null
},
{
"_id": "1775547000000",
"vcoresUsed_avg": "1",
"memoryUsed_avg": "1024",
"usedCapacity_avg": "3.1746032",
"absoluteUsedCapacity_avg": null
},
],
},
}

Running and pending applications
Resource allocation

YARN Summary

Total cluster resources
Used vs available memory and vCores
Node and application counts

How You Discover These APIs

YARN APIs are powered by GraphQL queries used internally by the UI. To access them:

Open Pulse in your browser
Go to the YARN tab
Open Developer Tools → Network tab
Filter for GraphQL requests
Click on relevant requests (queues, summary, etc.)
Copy:
- Query
- Variables
- Response
Reuse the same query in your API calls

Example:

BASE='<Pulse_UI_URL>' # or https / other host
curl -sS -X POST "$BASE/graphql" \
-H 'Content-Type: application/json' \
-H 'MonitorGroup: odp_usopp' \
-H 'role: eyJuYW1lIjoiYWRtaW4iLCJjbHVzdGVycyI6W119' \
-H 'Cookie: jwt=PASTE_SIGNED_JWT_COOKIE' \
d @ <<'EOF'
{
"operationName": "OPTIONAL_OR_OMIT",
"query": "PASTE_OR_INLINE_QUERY",
"variables": { }
}

What You Can Build

With these APIs, you can:

Build your own custom dashboards
Export data to BI tools or data lakes
Automate reports and pipelines
Monitor cluster utilization programmatically
Optimize resource usage and costs