-
Notifications
You must be signed in to change notification settings - Fork 679
Closed
Description
Has anyone had any luck using this runner in a matrix? lets say we're running a job that has a matrix of 3 things to run against. I am noticing that the check_run webhook seems to be initiated 3 times but 1 instance only ever spins up. the config i am using is fairly basic:
resource "random_password" "random" {
length = 28
}
module "github-runner" {
source = "philips-labs/github-runner/aws"
version = "0.9.1"
aws_region = var.aws_region
vpc_id = var.vpc_id
subnet_ids = var.subnet_ids
environment = "gh-ci"
github_app = {
key_base64 = "REDACTED"
id = "REDACTED"
client_id = "REDACTED"
client_secret = "REDACTED"
webhook_secret = "REDACTED"
}
block_device_mappings = {
volume_size = 50
}
webhook_lambda_zip = "lambdas-download/webhook.zip"
runner_binaries_syncer_lambda_zip = "lambdas-download/runner-binaries-syncer.zip"
runners_lambda_zip = "lambdas-download/runners.zip"
enable_organization_runners = false
runner_extra_labels = "terraform-runner"
enable_ssm_on_runners = true
ami_filter = { name = ["terraform-runner"] }
ami_owners = ["REDACTED"]
}
and the following workflow:
---
name: gh terraform runner test
on:
push:
branches:
- main
defaults:
run:
shell: bash
# prettier-ignore
jobs:
configure:
runs-on: terraform-runner
strategy:
matrix:
instance: ["one", "two", "three"]
steps:
- name: do some things
run: |
echo "we did it"
sleep 3
i've tried pre-registering 3 offline runners as well to no avail. in the above scenario only one instance spins up and then the "parallel" workflow runs sequentially from there. i would expect that 3 check_runs are sent, 3 instances spin up
any help is appreciated!
Metadata
Metadata
Assignees
Labels
No labels