GitHub Action that when used together with reusable workflows makes it easier to workaround the limit of 256 jobs in a matrix.
This project is part of our comprehensive "SweetOps" approach towards DevOps.
It's 100% Open Source and licensed under the APACHE2.
GitHub Actions matrix have limit to 256 items There is workaround to extend the limit with reusable workflows This GitHub Action outputs a JSON structure for up to 3 levels deep of nested matrixes. In theory run 256 ^ 3 (i.e., 16 777 216) jobs per workflow run!
Matrix max nested level | Total jobs count limit |
---|---|
1 | 256 |
2 | 65 536 |
3 | 16 777 216 |
If nested-matrices-count
input is 1
, the output matrix
would be JSON formatted string with the following structure
{
"include": [matrix items]
}
If nested-matrices-count
input is 2
output matrix
whould be a JSON formatted string with the following structure
{
"include": [{
"name": "group name",
"items": {
"include": [matrix items]
} ## serialized as string
}]
}
If nested-matrices-count
input is 3
output matrix
would be a JSON formatted string with the following structure
{
"include": [{
"name": "group name",
"items": [{
"name": "chunk 256 range name",
"include": [
"items": {
"include": [matrix items] ## serialized as string
}
]
}] ## serialized as string
} ## serialized as string
}]
}
Warning
Make sure you restrict the concurrency of your jobs to avoid DDOS'ing the GitHub Actions API, which might cause restrictions to be applied to your account.
Matrix max nested level | First Matrix Concurrency | Second Matrix Concurrency | Third Matrix Concurrency |
---|---|---|---|
1 | x | - | - |
2 | 1 | x | - |
3 | 1 | 1 | x |
The action have 3 modes depends of how many nested levels you want. The settings affect to reusable workflows count and usage pattern.
.github/workflows/matrices-1.yml
name: Pull Request
on:
pull_request:
branches: [ 'main' ]
types: [opened, synchronize, reopened, closed, labeled, unlabeled]
jobs:
matrix-builder:
runs-on: self-hosted
name: Affected stacks
outputs:
matrix: ${{ steps.extend.outputs.matrix }}
steps:
- id: setup-matrix
uses: druzsan/setup-matrix@v1
with:
matrix: |
os: ubuntu-latest windows-latest macos-latest,
python-version: 3.8 3.9 3.10
arch: arm64 amd64
- uses: cloudposse/github-action-matrix-extended@main
id: extend
with:
matrix: ${{ steps.setup-matrix.outputs.matrix }}
sort-by: '[.python-version, .os, .arch] | join("-")'
group-by: '.arch'
nested-matrices-count: '1'
operation:
if: ${{ needs.matrix-builder.outputs.matrix != '{"include":[]}' }}
needs:
- matrix-builder
strategy:
max-parallel: 10
fail-fast: false # Don't fail fast to avoid locking TF State
matrix: ${{ fromJson(needs.matrix-builder.outputs.matrix) }}
name: Do (${{ matrix.arch }})
runs-on: self-hosted
steps:
- shell: bash
run: |
echo "Do real work - ${{ matrix.os }} - ${{ matrix.arch }} - ${{ matrix.python-version }}"
.github/workflows/matrices-1.yml
name: Pull Request
on:
pull_request:
branches: [ 'main' ]
types: [opened, synchronize, reopened, closed, labeled, unlabeled]
jobs:
matrix-builder:
runs-on: self-hosted
name: Affected stacks
outputs:
matrix: ${{ steps.extend.outputs.matrix }}
steps:
- id: setup-matrix
uses: druzsan/setup-matrix@v1
with:
matrix: |
os: ubuntu-latest windows-latest macos-latest,
python-version: 3.8 3.9 3.10
arch: arm64 amd64
- uses: cloudposse/github-action-matrix-extended@main
id: extend
with:
sort-by: '[.python-version, .os, .arch] | join("-")'
group-by: '.arch'
nested-matrices-count: '1'
matrix: ${{ steps.setup-matrix.outputs.matrix }}
operation:
if: ${{ needs.matrix-builder.outputs.matrix != '{"include":[]}' }}
uses: ./.github/workflows/matrices-2.yml
needs:
- matrix-builder
strategy:
max-parallel: 1 # This is important to avoid ddos GHA API
fail-fast: false # Don't fail fast to avoid locking TF State
matrix: ${{ fromJson(needs.matrix-builder.outputs.matrix) }}
name: Group (${{ matrix.name }})
with:
items: ${{ matrix.items }}
.github/workflows/matrices-2.yml
name: Reusable workflow for 2 level of nested matrices
on:
workflow_call:
inputs:
items:
description: "Items"
required: true
type: string
jobs:
operation:
if: ${{ inputs.items != '{"include":[]}' }}
strategy:
max-parallel: 10
fail-fast: false # Don't fail fast to avoid locking TF State
matrix: ${{ fromJson(inputs.items) }}
name: Do (${{ matrix.arch }})
runs-on: self-hosted
steps:
- shell: bash
run: |
echo "Do real work - ${{ matrix.os }} - ${{ matrix.arch }} - ${{ matrix.python-version }}"
.github/workflows/matrices-1.yml
name: Pull Request
on:
pull_request:
branches: [ 'main' ]
types: [opened, synchronize, reopened, closed, labeled, unlabeled]
jobs:
matrix-builder:
runs-on: self-hosted
name: Affected stacks
outputs:
matrix: ${{ steps.extend.outputs.matrix }}
steps:
- id: setup-matrix
uses: druzsan/setup-matrix@v1
with:
matrix: |
os: ubuntu-latest windows-latest macos-latest,
python-version: 3.8 3.9 3.10
arch: arm64 amd64
- uses: cloudposse/github-action-matrix-extended@main
id: query
with:
sort-by: '[.python-version, .os, .arch] | join("-")'
group-by: '.arch'
nested-matrices-count: '1'
matrix: ${{ steps.setup-matrix.outputs.matrix }}
operation:
if: ${{ needs.matrix-builder.outputs.matrix != '{"include":[]}' }}
uses: ./.github/workflows/matrices-2.yml
needs:
- matrix-builder
strategy:
max-parallel: 1 # This is important to avoid ddos GHA API
fail-fast: false # Don't fail fast to avoid locking TF State
matrix: ${{ fromJson(needs.matrix-builder.outputs.matrix) }}
name: Group (${{ matrix.name }})
with:
items: ${{ matrix.items }}
.github/workflows/matrices-2.yml
name: Reusable workflow for 2 level of nested matrices
on:
workflow_call:
inputs:
items:
description: "Items"
required: true
type: string
jobs:
operation:
if: ${{ inputs.items != '{"include":[]}' }}
uses: ./.github/workflows/matrices-3.yml
strategy:
max-parallel: 1 # This is important to avoid ddos GHA API
fail-fast: false # Don't fail fast to avoid locking TF State
matrix: ${{ fromJson(inputs.items) }}
name: Group (${{ matrix.name }})
with:
items: ${{ matrix.items }}
.github/workflows/matrices-3.yml
name: Reusable workflow for 3 level of nested matrices
on:
workflow_call:
inputs:
items:
description: "Items"
required: true
type: string
jobs:
operation:
if: ${{ inputs.items != '{"include":[]}' }}
strategy:
max-parallel: 10
fail-fast: false # Don't fail fast to avoid locking TF State
matrix: ${{ fromJson(inputs.items) }}
name: Do (${{ matrix.arch }})
runs-on: self-hosted
steps:
- shell: bash
run: |
echo "Do real work - ${{ matrix.os }} - ${{ matrix.arch }} - ${{ matrix.python-version }}"
Name | Description | Default | Required |
---|---|---|---|
group-by | Group by query | empty | false |
matrix | Matrix inputs (JSON array or object which includes property passed as string or file path) | N/A | true |
nested-matrices-count | Number of nested matrices that should be returned as the output (from 1 to 3) | 1 | false |
sort-by | Sort by query | empty | false |
Name | Description |
---|---|
matrix | A matrix suitable for extending matrix size workaround (see README) |
Like this project? Please give it a ★ on our GitHub! (it helps us a lot)
Are you using this project or any of our other projects? Consider leaving a testimonial. =)
Check out these related projects.
- Setup matrix - GitHub action to create reusable dynamic job matrices for your workflows.
For additional context, refer to some of these links.
- github-action-atmos-affected-stacks - A composite workflow that runs the atmos describe affected command
Got a question? We got answers.
File a GitHub issue, send us an email or join our Slack Community.
We are a DevOps Accelerator. We'll help you build your cloud infrastructure from the ground up so you can own it. Then we'll show you how to operate it and stick around for as long as you need us.
Work directly with our team of DevOps experts via email, slack, and video conferencing.
We deliver 10x the value for a fraction of the cost of a full-time engineer. Our track record is not even funny. If you want things done right and you need it done FAST, then we're your best bet.
- Reference Architecture. You'll get everything you need from the ground up built using 100% infrastructure as code.
- Release Engineering. You'll have end-to-end CI/CD with unlimited staging environments.
- Site Reliability Engineering. You'll have total visibility into your apps and microservices.
- Security Baseline. You'll have built-in governance with accountability and audit logs for all changes.
- GitOps. You'll be able to operate your infrastructure via Pull Requests.
- Training. You'll receive hands-on training so your team can operate what we build.
- Questions. You'll have a direct line of communication between our teams via a Shared Slack channel.
- Troubleshooting. You'll get help to triage when things aren't working.
- Code Reviews. You'll receive constructive feedback on Pull Requests.
- Bug Fixes. We'll rapidly work with you to fix any bugs in our projects.
Join our Open Source Community on Slack. It's FREE for everyone! Our "SweetOps" community is where you get to talk with others who share a similar vision for how to rollout and manage infrastructure. This is the best place to talk shop, ask questions, solicit feedback, and work together as a community to build totally sweet infrastructure.
Participate in our Discourse Forums. Here you'll find answers to commonly asked questions. Most questions will be related to the enormous number of projects we support on our GitHub. Come here to collaborate on answers, find solutions, and get ideas about the products and services we value. It only takes a minute to get started! Just sign in with SSO using your GitHub account.
Sign up for our newsletter that covers everything on our technology radar. Receive updates on what we're up to on GitHub as well as awesome new projects we discover.
Join us every Wednesday via Zoom for our weekly "Lunch & Learn" sessions. It's FREE for everyone!
Please use the issue tracker to report any bugs or file feature requests.
If you are interested in being a contributor and want to get involved in developing this project or help out with our other projects, we would love to hear from you! Shoot us an email.
In general, PRs are welcome. We follow the typical "fork-and-pull" Git workflow.
- Fork the repo on GitHub
- Clone the project to your own machine
- Commit changes to your own branch
- Push your work back up to your fork
- Submit a Pull Request so that we can review your changes
NOTE: Be sure to merge the latest changes from "upstream" before making a pull request!
Copyright © 2017-2023 Cloud Posse, LLC
See LICENSE for full details.
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements. See the NOTICE file
distributed with this work for additional information
regarding copyright ownership. The ASF licenses this file
to you under the Apache License, Version 2.0 (the
"License"); you may not use this file except in compliance
with the License. You may obtain a copy of the License at
https://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing,
software distributed under the License is distributed on an
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
KIND, either express or implied. See the License for the
specific language governing permissions and limitations
under the License.
All other trademarks referenced herein are the property of their respective owners.
This project is maintained and funded by Cloud Posse, LLC. Like it? Please let us know by leaving a testimonial!
We're a DevOps Professional Services company based in Los Angeles, CA. We ❤️ Open Source Software.
We offer paid support on all of our projects.
Check out our other projects, follow us on twitter, apply for a job, or hire us to help with your cloud strategy and implementation.
Igor Rodionov |
---|