AWS Glue for Ray job

aws Workeraws-glue-ray-job

Managed Ray job for distributed Python data processing on Glue workers.

Provider: aws
Kind: Worker
Category: Compute
Official docs: https://docs.aws.amazon.com/glue/latest/dg/author-job-ray.html

Properties · 5 set by this component

property	type	values	required	kind default	this component
accelerator	enum	none \| gpu \| asic	no	none	▸ none
concurrency	integer	—	no	—	inherited
fanoutStrategy · advanced	enum	none \| push \| pull \| hybrid	no	none	▸ none
fanoutThreshold · advanced	integer	—	no	—	inherited
idempotent	boolean	—	no	false	inherited
instances	integer	—	no	—	inherited
managed	boolean	—	no	false	▸ true
numberOfWorkers · component property	integer	—	no	—	inherited
rateControl · advanced	enum	none \| fixed \| aimd	no	none	▸ none
rayRuntime · component property	string	—	no	Ray2.4	inherited
schedule	string	—	no	—	inherited
trigger	enum	queue \| stream \| schedule \| batch \| manual	yes	—	▸ batch
workerType · component property	enum	z-2x	no	z-2x	inherited

Rules that apply · 3

Every rule whose selector covers this component's kind or its category. Whether a rule fires depends on the design: its properties, its edges and its groups.

Push fan-out without a celebrity thresholdwarning

push-fanout-without-celebrity-path

Pure push fan-out over a social graph is brutal at the tail: one post from an account with millions of followers becomes millions of cache writes. A follower threshold above which readers merge at read time bounds that write amplification.

— all of these hold:
— the node's `fanoutStrategy` is "push"
— the node's `fanoutThreshold` has no value
— the design contains a node of kind graph-database

Shared worker pool across queueswarning

shared-worker-pool-across-queues

One worker pool consuming several queues lets a slow or poisoned queue starve the others. That is precisely what the design's per-queue isolation requirement forbids.

— all of these hold:
— the node has more than 1 incoming edge from a queue
— the design requirement `isolation.perQueue` is true

Stateful node without zone redundancyerror

stateful-node-without-zone-redundancy

A stateful compute node pinned to exactly one availability zone disappears with that zone, and its state goes with it. Nodes inside a cell are exempt, because there the cell is deliberately single-zone and redundancy comes from the other cells.

— all of these hold:
— the node's `stateless` is false
— the node sits inside exactly 1 availability-zone group
— this does not hold:
— the node sits inside at least one cell group

Use it in a design

nodes:
  - { id: "glue-ray-job", component: "aws-glue-ray-job" }

Other Worker products · 82

SchedulerXalibaba-schedulerxalibabacloudworker Amazon AppFlowaws-appflowawsworker AWS B2B Data Interchangeaws-b2b-data-interchangeawsworker AWS Batchaws-batchawsworker AWS Glue DataBrewaws-databrewawsworker AWS DataSyncaws-datasyncawsworker AWS Deadline Cloudaws-deadline-cloudawsworker AWS Entity Resolutionaws-entity-resolutionawsworker Amazon EventBridge Scheduleraws-eventbridge-schedulerawsworker AWS Fault Injection Serviceaws-fisawsworker AWS Glue crawleraws-glue-crawlerawsworker AWS Application Migration Serviceaws-mgnawsworker Amazon MSK Connectaws-msk-connectawsworker Amazon SageMaker AI Training Jobaws-sagemaker-training-jobawsworker AWS Systems Manageraws-systems-managerawsworker AWS Thinkbox Deadline 10aws-thinkbox-deadlineawsworker AWS Glue jobglue-jobawsworker Azure Automationazure-automationazureworker Azure Batchazure-batchazureworker Azure Chaos Studioazure-chaos-studioazureworker Azure Data Factoryazure-data-factoryazureworker Azure Data Shareazure-data-shareazureworker Azure Database Migration Serviceazure-database-migration-serviceazureworker Microsoft Entra Connect Syncazure-entra-connect-syncazureworker Microsoft Graph Data Connectazure-graph-data-connectazureworker Azure Storage Moverazure-storage-moverazureworker Azure Update Managerazure-update-managerazureworker Browser Runcloudflare-browser-runcloudflareworker Cloudflare Snippetscloudflare-snippetscloudflareworker Cloudflare Zarazcloudflare-zarazcloudflareworker Airbyteairbyteworker Airflow DAGairflow-dagworker Chaos Meshchaos-meshworker Chunkrchunkrworker Container workercontainer-workerworker Cron jobcron-jobworker Daytonadaytonaworker E2Be2bworker Embulkembulkworker Fan-out daemonfanout-daemonworker

and 42 more → Worker page

Source

1# yaml-language-server: $schema=../../schema/component.schema.json
2# AWS Glue for Ray job: managed Ray job run on the same AWS Glue job machinery as Spark ETL, "designed for
3# running the same script on a recurring cadence" ("Working with Ray jobs in AWS Glue"). Maintenance: closed to
4# new customers since 31 March 2026; existing jobs stay placeable until end of support. Ray jobs have access to
5# one worker type, Z.2X, and the runtime is chosen from the Runtime field, whose only valid value is Ray2.4.
6schemaVersion: "1"
7type: "component"
8id: "aws-glue-ray-job"
9kind: "worker"
10provider: "aws"
11docs: "https://docs.aws.amazon.com/glue/latest/dg/author-job-ray.html"
12defaults:
13  managed: true
14  trigger: "batch"
15  accelerator: "none"
16  rateControl: "none"
17  fanoutStrategy: "none"
18properties:
19  workerType:
20    type: "enum"
21    values: ["z-2x"]
22    default: "z-2x"
23  rayRuntime:
24    type: "string"
25    default: "Ray2.4"
26  numberOfWorkers:
27    type: "integer"

The same page as markdown, for an agent: /catalogue/components/aws-glue-ray-job.md