diff --git a/README.md b/README.md index 1ca2e82e22..75af9acb78 100644 --- a/README.md +++ b/README.md @@ -145,7 +145,7 @@ Join our discord community via [this invite link](https://discord.gg/bxgXW8jJGh) | [instance\_profile\_path](#input\_instance\_profile\_path) | The path that will be added to the instance\_profile, if not set the environment name will be used. | `string` | `null` | no | | [instance\_target\_capacity\_type](#input\_instance\_target\_capacity\_type) | Default lifecycle used for runner instances, can be either `spot` or `on-demand`. | `string` | `"spot"` | no | | [instance\_termination\_watcher](#input\_instance\_termination\_watcher) | Configuration for the instance termination watcher. This feature is Beta, changes will not trigger a major release as long in beta.

`enable`: Enable or disable the spot termination watcher.
'features': Enable or disable features of the termination watcher.
`memory_size`: Memory size limit in MB of the lambda.
`s3_key`: S3 key for syncer lambda function. Required if using S3 bucket to specify lambdas.
`s3_object_version`: S3 object version for syncer lambda function. Useful if S3 versioning is enabled on source bucket.
`timeout`: Time out of the lambda in seconds.
`zip`: File location of the lambda zip file. | 
object({
    enable = optional(bool, false)
    features = optional(object({
      enable_spot_termination_handler              = optional(bool, true)
      enable_spot_termination_notification_watcher = optional(bool, true)
    }), {})
    memory_size       = optional(number, null)
    s3_key            = optional(string, null)
    s3_object_version = optional(string, null)
    timeout           = optional(number, null)
    zip               = optional(string, null)
  })
| `{}` | no | -| [instance\_types](#input\_instance\_types) | List of instance types for the action runner. Defaults are based on runner\_os (al2023 for linux and Windows Server Core for win). | `list(string)` | 
[
  "m5.large",
  "c5.large"
]
| no | +| [instance\_types](#input\_instance\_types) | List of instance types for the action runner. Defaults are based on runner\_os (al2023 for linux, macOS Sequoia for osx, Windows Server Core for win). | `list(string)` | 
[
  "m5.large",
  "c5.large"
]
| no | | [job\_queue\_retention\_in\_seconds](#input\_job\_queue\_retention\_in\_seconds) | The number of seconds the job is held in the queue before it is purged. | `number` | `86400` | no | | [job\_retry](#input\_job\_retry) | Experimental! Can be removed / changed without trigger a major release.Configure job retries. The configuration enables job retries (for ephemeral runners). After creating the instances a message will be published to a job retry queue. The job retry check lambda is checking after a delay if the job is queued. If not the message will be published again on the scale-up (build queue). Using this feature can impact the rate limit of the GitHub app.

`enable`: Enable or disable the job retry feature.
`delay_in_seconds`: The delay in seconds before the job retry check lambda will check the job status.
`delay_backoff`: The backoff factor for the delay.
`lambda_memory_size`: Memory size limit in MB for the job retry check lambda.
`lambda_timeout`: Time out of the job retry check lambda in seconds.
`max_attempts`: The maximum number of attempts to retry the job. | 
object({
    enable             = optional(bool, false)
    delay_in_seconds   = optional(number, 300)
    delay_backoff      = optional(number, 2)
    lambda_memory_size = optional(number, 256)
    lambda_timeout     = optional(number, 30)
    max_attempts       = optional(number, 1)
  })
| `{}` | no | | [key\_name](#input\_key\_name) | Key pair name | `string` | `null` | no | @@ -201,7 +201,7 @@ Join our discord community via [this invite link](https://discord.gg/bxgXW8jJGh) | [runner\_log\_files](#input\_runner\_log\_files) | (optional) Replaces the module default cloudwatch log config. See https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/CloudWatch-Agent-Configuration-File-Details.html for details. | 
list(object({
    log_group_name   = string
    prefix_log_group = bool
    file_path        = string
    log_stream_name  = string
  }))
| `null` | no | | [runner\_metadata\_options](#input\_runner\_metadata\_options) | Metadata options for the ec2 runner instances. By default, the module uses metadata tags for bootstrapping the runner, only disable `instance_metadata_tags` when using custom scripts for starting the runner. | `map(any)` | 
{
  "http_endpoint": "enabled",
  "http_put_response_hop_limit": 1,
  "http_tokens": "required",
  "instance_metadata_tags": "enabled"
}
| no | | [runner\_name\_prefix](#input\_runner\_name\_prefix) | The prefix used for the GitHub runner name. The prefix will be used in the default start script to prefix the instance name when register the runner in GitHub. The value is available via an EC2 tag 'ghr:runner\_name\_prefix'. | `string` | `""` | no | -| [runner\_os](#input\_runner\_os) | The EC2 Operating System type to use for action runner instances (linux,windows). | `string` | `"linux"` | no | +| [runner\_os](#input\_runner\_os) | The EC2 Operating System type to use for action runner instances (linux, osx, windows). | `string` | `"linux"` | no | | [runner\_placement](#input\_runner\_placement) | The placement options for the instance. See https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/launch_template#placement for details. | 
object({
    affinity                = optional(string)
    availability_zone       = optional(string)
    group_id                = optional(string)
    group_name              = optional(string)
    host_id                 = optional(string)
    host_resource_group_arn = optional(number)
    spread_domain           = optional(string)
    tenancy                 = optional(string)
    partition_number        = optional(number)
  })
| `null` | no | | [runner\_run\_as](#input\_runner\_run\_as) | Run the GitHub actions agent as user. | `string` | `"ec2-user"` | no | | [runners\_ebs\_optimized](#input\_runners\_ebs\_optimized) | Enable EBS optimization for the runner instances. | `bool` | `false` | no | diff --git a/examples/prebuilt/README.md b/examples/prebuilt/README.md index b24f47a01d..c2d66139d4 100644 --- a/examples/prebuilt/README.md +++ b/examples/prebuilt/README.md @@ -112,7 +112,7 @@ terraform output webhook_secret | [aws\_region](#input\_aws\_region) | AWS region. | `string` | `"eu-west-1"` | no | | [environment](#input\_environment) | Environment name, used as prefix. | `string` | `null` | no | | [github\_app](#input\_github\_app) | GitHub for API usages. | 
object({
    id         = string
    key_base64 = string
  })
| n/a | yes | -| [runner\_os](#input\_runner\_os) | The EC2 Operating System type to use for action runner instances (linux,windows). | `string` | `"linux"` | no | +| [runner\_os](#input\_runner\_os) | The EC2 Operating System type to use for action runner instances (linux, osx, windows). | `string` | `"linux"` | no | ## Outputs diff --git a/examples/prebuilt/variables.tf b/examples/prebuilt/variables.tf index 643072a163..11670a5d2e 100644 --- a/examples/prebuilt/variables.tf +++ b/examples/prebuilt/variables.tf @@ -22,7 +22,7 @@ variable "aws_region" { } variable "runner_os" { - description = "The EC2 Operating System type to use for action runner instances (linux,windows)." + description = "The EC2 Operating System type to use for action runner instances (linux, osx, windows)." type = string default = "linux" diff --git a/lambdas/functions/control-plane/src/aws/runners.ts b/lambdas/functions/control-plane/src/aws/runners.ts index d95dc99fa4..a45c1db3a2 100644 --- a/lambdas/functions/control-plane/src/aws/runners.ts +++ b/lambdas/functions/control-plane/src/aws/runners.ts @@ -207,6 +207,7 @@ async function processFleetResult( 'MaxSpotInstanceCountExceeded', 'MaxSpotFleetRequestCountExceeded', 'InsufficientInstanceCapacity', + 'InsufficientCapacityOnHost', ]; const failedCount = countScaleErrors(errors, scaleErrors); diff --git a/modules/multi-runner/README.md b/modules/multi-runner/README.md index 7920092afa..7a66b85d0b 100644 --- a/modules/multi-runner/README.md +++ b/modules/multi-runner/README.md @@ -150,7 +150,7 @@ module "multi-runner" { | [logging\_retention\_in\_days](#input\_logging\_retention\_in\_days) | Specifies the number of days you want to retain log events for the lambda log group. Possible values are: 0, 1, 3, 5, 7, 14, 30, 60, 90, 120, 150, 180, 365, 400, 545, 731, 1827, and 3653. | `number` | `180` | no | | [matcher\_config\_parameter\_store\_tier](#input\_matcher\_config\_parameter\_store\_tier) | The tier of the parameter store for the matcher configuration. Valid values are `Standard`, and `Advanced`. | `string` | `"Standard"` | no | | [metrics](#input\_metrics) | Configuration for metrics created by the module, by default metrics are disabled to avoid additional costs. When metrics are enable all metrics are created unless explicit configured otherwise. | 
object({
    enable    = optional(bool, false)
    namespace = optional(string, "GitHub Runners")
    metric = optional(object({
      enable_github_app_rate_limit    = optional(bool, true)
      enable_job_retry                = optional(bool, true)
      enable_spot_termination_warning = optional(bool, true)
    }), {})
  })
| `{}` | no | -| [multi\_runner\_config](#input\_multi\_runner\_config) | multi\_runner\_config = {
runner\_config: {
runner\_os: "The EC2 Operating System type to use for action runner instances (linux,windows)."
runner\_architecture: "The platform architecture of the runner instance\_type."
runner\_metadata\_options: "(Optional) Metadata options for the ec2 runner instances."
ami: "(Optional) AMI configuration for the action runner instances. This object allows you to specify all AMI-related settings in one place."
create\_service\_linked\_role\_spot: (Optional) create the serviced linked role for spot instances that is required by the scale-up lambda.
credit\_specification: "(Optional) The credit specification of the runner instance\_type. Can be unset, `standard` or `unlimited`.
delay\_webhook\_event: "The number of seconds the event accepted by the webhook is invisible on the queue before the scale up lambda will receive the event."
disable\_runner\_autoupdate: "Disable the auto update of the github runner agent. Be aware there is a grace period of 30 days, see also the [GitHub article](https://github.blog/changelog/2022-02-01-github-actions-self-hosted-runners-can-now-disable-automatic-updates/)"
ebs\_optimized: "The EC2 EBS optimized configuration."
enable\_ephemeral\_runners: "Enable ephemeral runners, runners will only be used once."
enable\_job\_queued\_check: "Enables JIT configuration for creating runners instead of registration token based registraton. JIT configuration will only be applied for ephemeral runners. By default JIT configuration is enabled for ephemeral runners an can be disabled via this override. When running on GHES without support for JIT configuration this variable should be set to true for ephemeral runners."
enable\_on\_demand\_failover\_for\_errors: "Enable on-demand failover. For example to fall back to on demand when no spot capacity is available the variable can be set to `InsufficientInstanceCapacity`. When not defined the default behavior is to retry later."
enable\_organization\_runners: "Register runners to organization, instead of repo level"
enable\_runner\_binaries\_syncer: "Option to disable the lambda to sync GitHub runner distribution, useful when using a pre-build AMI."
enable\_ssm\_on\_runners: "Enable to allow access the runner instances for debugging purposes via SSM. Note that this adds additional permissions to the runner instances."
enable\_userdata: "Should the userdata script be enabled for the runner. Set this to false if you are using your own prebuilt AMI."
instance\_allocation\_strategy: "The allocation strategy for spot instances. AWS recommends to use `capacity-optimized` however the AWS default is `lowest-price`."
instance\_max\_spot\_price: "Max price price for spot instances per hour. This variable will be passed to the create fleet as max spot price for the fleet."
instance\_target\_capacity\_type: "Default lifecycle used for runner instances, can be either `spot` or `on-demand`."
instance\_types: "List of instance types for the action runner. Defaults are based on runner\_os (al2023 for linux and Windows Server Core for win)."
job\_queue\_retention\_in\_seconds: "The number of seconds the job is held in the queue before it is purged"
minimum\_running\_time\_in\_minutes: "The time an ec2 action runner should be running at minimum before terminated if not busy."
pool\_runner\_owner: "The pool will deploy runners to the GitHub org ID, set this value to the org to which you want the runners deployed. Repo level is not supported."
runner\_additional\_security\_group\_ids: "List of additional security groups IDs to apply to the runner. If added outside the multi\_runner\_config block, the additional security group(s) will be applied to all runner configs. If added inside the multi\_runner\_config, the additional security group(s) will be applied to the individual runner."
runner\_as\_root: "Run the action runner under the root user. Variable `runner_run_as` will be ignored."
runner\_boot\_time\_in\_minutes: "The minimum time for an EC2 runner to boot and register as a runner."
runner\_disable\_default\_labels: "Disable default labels for the runners (os, architecture and `self-hosted`). If enabled, the runner will only have the extra labels provided in `runner_extra_labels`. In case you on own start script is used, this configuration parameter needs to be parsed via SSM."
runner\_extra\_labels: "Extra (custom) labels for the runners (GitHub). Separate each label by a comma. Labels checks on the webhook can be enforced by setting `multi_runner_config.matcherConfig.exactMatch`. GitHub read-only labels should not be provided."
runner\_group\_name: "Name of the runner group."
runner\_name\_prefix: "Prefix for the GitHub runner name."
runner\_run\_as: "Run the GitHub actions agent as user."
runners\_maximum\_count: "The maximum number of runners that will be created. Setting the variable to `-1` desiables the maximum check."
scale\_down\_schedule\_expression: "Scheduler expression to check every x for scale down."
scale\_up\_reserved\_concurrent\_executions: "Amount of reserved concurrent executions for the scale-up lambda function. A value of 0 disables lambda from being triggered and -1 removes any concurrency limitations."
userdata\_template: "Alternative user-data template, replacing the default template. By providing your own user\_data you have to take care of installing all required software, including the action runner. Variables userdata\_pre/post\_install are ignored."
enable\_jit\_config "Overwrite the default behavior for JIT configuration. By default JIT configuration is enabled for ephemeral runners and disabled for non-ephemeral runners. In case of GHES check first if the JIT config API is available. In case you are upgrading from 3.x to 4.x you can set `enable_jit_config` to `false` to avoid a breaking change when having your own AMI."
enable\_runner\_detailed\_monitoring: "Should detailed monitoring be enabled for the runner. Set this to true if you want to use detailed monitoring. See https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/using-cloudwatch-new.html for details."
enable\_cloudwatch\_agent: "Enabling the cloudwatch agent on the ec2 runner instances, the runner contains default config. Configuration can be overridden via `cloudwatch_config`."
cloudwatch\_config: "(optional) Replaces the module default cloudwatch log config. See https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/CloudWatch-Agent-Configuration-File-Details.html for details."
userdata\_pre\_install: "Script to be ran before the GitHub Actions runner is installed on the EC2 instances"
userdata\_post\_install: "Script to be ran after the GitHub Actions runner is installed on the EC2 instances"
runner\_hook\_job\_started: "Script to be ran in the runner environment at the beginning of every job"
runner\_hook\_job\_completed: "Script to be ran in the runner environment at the end of every job"
runner\_ec2\_tags: "Map of tags that will be added to the launch template instance tag specifications."
runner\_iam\_role\_managed\_policy\_arns: "Attach AWS or customer-managed IAM policies (by ARN) to the runner IAM role"
vpc\_id: "The VPC for security groups of the action runners. If not set uses the value of `var.vpc_id`."
subnet\_ids: "List of subnets in which the action runners will be launched, the subnets needs to be subnets in the `vpc_id`. If not set, uses the value of `var.subnet_ids`."
idle\_config: "List of time period that can be defined as cron expression to keep a minimum amount of runners active instead of scaling down to 0. By defining this list you can ensure that in time periods that match the cron expression within 5 seconds a runner is kept idle."
runner\_log\_files: "(optional) Replaces the module default cloudwatch log config. See https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/CloudWatch-Agent-Configuration-File-Details.html for details."
block\_device\_mappings: "The EC2 instance block device configuration. Takes the following keys: `device_name`, `delete_on_termination`, `volume_type`, `volume_size`, `encrypted`, `iops`, `throughput`, `kms_key_id`, `snapshot_id`."
job\_retry: "Experimental! Can be removed / changed without trigger a major release. Configure job retries. The configuration enables job retries (for ephemeral runners). After creating the instances a message will be published to a job retry queue. The job retry check lambda is checking after a delay if the job is queued. If not the message will be published again on the scale-up (build queue). Using this feature can impact the rate limit of the GitHub app."
pool\_config: "The configuration for updating the pool. The `pool_size` to adjust to by the events triggered by the `schedule_expression`. For example you can configure a cron expression for week days to adjust the pool to 10 and another expression for the weekend to adjust the pool to 1. Use `schedule_expression_timezone` to override the schedule time zone (defaults to UTC)."
}
matcherConfig: {
labelMatchers: "The list of list of labels supported by the runner configuration. `[[self-hosted, linux, x64, example]]`"
exactMatch: "If set to true all labels in the workflow job must match the GitHub labels (os, architecture and `self-hosted`). When false if __any__ workflow label matches it will trigger the webhook."
priority: "If set it defines the priority of the matcher, the matcher with the lowest priority will be evaluated first. Default is 999, allowed values 0-999."
}
redrive\_build\_queue: "Set options to attach (optional) a dead letter queue to the build queue, the queue between the webhook and the scale up lambda. You have the following options. 1. Disable by setting `enabled` to false. 2. Enable by setting `enabled` to `true`, `maxReceiveCount` to a number of max retries."
} | 
map(object({
    runner_config = object({
      runner_os           = string
      runner_architecture = string
      runner_metadata_options = optional(map(any), {
        instance_metadata_tags      = "enabled"
        http_endpoint               = "enabled"
        http_tokens                 = "required"
        http_put_response_hop_limit = 1
      })
      ami = optional(object({
        filter               = optional(map(list(string)), { state = ["available"] })
        owners               = optional(list(string), ["amazon"])
        id_ssm_parameter_arn = optional(string, null)
        kms_key_arn          = optional(string, null)
      }), null)
      create_service_linked_role_spot         = optional(bool, false)
      credit_specification                    = optional(string, null)
      delay_webhook_event                     = optional(number, 30)
      disable_runner_autoupdate               = optional(bool, false)
      ebs_optimized                           = optional(bool, false)
      enable_ephemeral_runners                = optional(bool, false)
      enable_job_queued_check                 = optional(bool, null)
      enable_on_demand_failover_for_errors    = optional(list(string), [])
      enable_organization_runners             = optional(bool, false)
      enable_runner_binaries_syncer           = optional(bool, true)
      enable_ssm_on_runners                   = optional(bool, false)
      enable_userdata                         = optional(bool, true)
      instance_allocation_strategy            = optional(string, "lowest-price")
      instance_max_spot_price                 = optional(string, null)
      instance_target_capacity_type           = optional(string, "spot")
      instance_types                          = list(string)
      job_queue_retention_in_seconds          = optional(number, 86400)
      minimum_running_time_in_minutes         = optional(number, null)
      pool_runner_owner                       = optional(string, null)
      runner_as_root                          = optional(bool, false)
      runner_boot_time_in_minutes             = optional(number, 5)
      runner_disable_default_labels           = optional(bool, false)
      runner_extra_labels                     = optional(list(string), [])
      runner_group_name                       = optional(string, "Default")
      runner_name_prefix                      = optional(string, "")
      runner_run_as                           = optional(string, "ec2-user")
      runners_maximum_count                   = number
      runner_additional_security_group_ids    = optional(list(string), [])
      scale_down_schedule_expression          = optional(string, "cron(*/5 * * * ? *)")
      scale_up_reserved_concurrent_executions = optional(number, 1)
      userdata_template                       = optional(string, null)
      userdata_content                        = optional(string, null)
      enable_jit_config                       = optional(bool, null)
      enable_runner_detailed_monitoring       = optional(bool, false)
      enable_cloudwatch_agent                 = optional(bool, true)
      cloudwatch_config                       = optional(string, null)
      userdata_pre_install                    = optional(string, "")
      userdata_post_install                   = optional(string, "")
      runner_hook_job_started                 = optional(string, "")
      runner_hook_job_completed               = optional(string, "")
      runner_ec2_tags                         = optional(map(string), {})
      runner_iam_role_managed_policy_arns     = optional(list(string), [])
      vpc_id                                  = optional(string, null)
      subnet_ids                              = optional(list(string), null)
      idle_config = optional(list(object({
        cron             = string
        timeZone         = string
        idleCount        = number
        evictionStrategy = optional(string, "oldest_first")
      })), [])
      cpu_options = optional(object({
        core_count       = number
        threads_per_core = number
      }), null)
      placement = optional(object({
        affinity                = optional(string)
        availability_zone       = optional(string)
        group_id                = optional(string)
        group_name              = optional(string)
        host_id                 = optional(string)
        host_resource_group_arn = optional(number)
        spread_domain           = optional(string)
        tenancy                 = optional(string)
        partition_number        = optional(number)
      }), null)
      runner_log_files = optional(list(object({
        log_group_name   = string
        prefix_log_group = bool
        file_path        = string
        log_stream_name  = string
      })), null)
      block_device_mappings = optional(list(object({
        delete_on_termination = optional(bool, true)
        device_name           = optional(string, "/dev/xvda")
        encrypted             = optional(bool, true)
        iops                  = optional(number)
        kms_key_id            = optional(string)
        snapshot_id           = optional(string)
        throughput            = optional(number)
        volume_size           = number
        volume_type           = optional(string, "gp3")
        })), [{
        volume_size = 30
      }])
      pool_config = optional(list(object({
        schedule_expression          = string
        schedule_expression_timezone = optional(string)
        size                         = number
      })), [])
      job_retry = optional(object({
        enable             = optional(bool, false)
        delay_in_seconds   = optional(number, 300)
        delay_backoff      = optional(number, 2)
        lambda_memory_size = optional(number, 256)
        lambda_timeout     = optional(number, 30)
        max_attempts       = optional(number, 1)
      }), {})
    })
    matcherConfig = object({
      labelMatchers = list(list(string))
      exactMatch    = optional(bool, false)
      priority      = optional(number, 999)
    })
    redrive_build_queue = optional(object({
      enabled         = bool
      maxReceiveCount = number
      }), {
      enabled         = false
      maxReceiveCount = null
    })
  }))
| n/a | yes | +| [multi\_runner\_config](#input\_multi\_runner\_config) | multi\_runner\_config = {
runner\_config: {
runner\_os: "The EC2 Operating System type to use for action runner instances (linux, osx, windows)."
runner\_architecture: "The platform architecture of the runner instance\_type."
runner\_metadata\_options: "(Optional) Metadata options for the ec2 runner instances."
ami: "(Optional) AMI configuration for the action runner instances. This object allows you to specify all AMI-related settings in one place."
create\_service\_linked\_role\_spot: (Optional) create the serviced linked role for spot instances that is required by the scale-up lambda.
credit\_specification: "(Optional) The credit specification of the runner instance\_type. Can be unset, `standard` or `unlimited`.
delay\_webhook\_event: "The number of seconds the event accepted by the webhook is invisible on the queue before the scale up lambda will receive the event."
disable\_runner\_autoupdate: "Disable the auto update of the github runner agent. Be aware there is a grace period of 30 days, see also the [GitHub article](https://github.blog/changelog/2022-02-01-github-actions-self-hosted-runners-can-now-disable-automatic-updates/)"
ebs\_optimized: "The EC2 EBS optimized configuration."
enable\_ephemeral\_runners: "Enable ephemeral runners, runners will only be used once."
enable\_job\_queued\_check: "Enables JIT configuration for creating runners instead of registration token based registraton. JIT configuration will only be applied for ephemeral runners. By default JIT configuration is enabled for ephemeral runners an can be disabled via this override. When running on GHES without support for JIT configuration this variable should be set to true for ephemeral runners."
enable\_on\_demand\_failover\_for\_errors: "Enable on-demand failover. For example to fall back to on demand when no spot capacity is available the variable can be set to `InsufficientInstanceCapacity`. When not defined the default behavior is to retry later."
enable\_organization\_runners: "Register runners to organization, instead of repo level"
enable\_runner\_binaries\_syncer: "Option to disable the lambda to sync GitHub runner distribution, useful when using a pre-build AMI."
enable\_ssm\_on\_runners: "Enable to allow access the runner instances for debugging purposes via SSM. Note that this adds additional permissions to the runner instances."
enable\_userdata: "Should the userdata script be enabled for the runner. Set this to false if you are using your own prebuilt AMI."
instance\_allocation\_strategy: "The allocation strategy for spot instances. AWS recommends to use `capacity-optimized` however the AWS default is `lowest-price`."
instance\_max\_spot\_price: "Max price price for spot instances per hour. This variable will be passed to the create fleet as max spot price for the fleet."
instance\_target\_capacity\_type: "Default lifecycle used for runner instances, can be either `spot` or `on-demand`."
description: "List of instance types for the action runner. Defaults are based on runner\_os (al2023 for linux, macOS Sequoia for osx, Windows Server Core for win)."
job\_queue\_retention\_in\_seconds: "The number of seconds the job is held in the queue before it is purged"
minimum\_running\_time\_in\_minutes: "The time an ec2 action runner should be running at minimum before terminated if not busy."
pool\_runner\_owner: "The pool will deploy runners to the GitHub org ID, set this value to the org to which you want the runners deployed. Repo level is not supported."
runner\_additional\_security\_group\_ids: "List of additional security groups IDs to apply to the runner. If added outside the multi\_runner\_config block, the additional security group(s) will be applied to all runner configs. If added inside the multi\_runner\_config, the additional security group(s) will be applied to the individual runner."
runner\_as\_root: "Run the action runner under the root user. Variable `runner_run_as` will be ignored."
runner\_boot\_time\_in\_minutes: "The minimum time for an EC2 runner to boot and register as a runner."
runner\_disable\_default\_labels: "Disable default labels for the runners (os, architecture and `self-hosted`). If enabled, the runner will only have the extra labels provided in `runner_extra_labels`. In case you on own start script is used, this configuration parameter needs to be parsed via SSM."
runner\_extra\_labels: "Extra (custom) labels for the runners (GitHub). Separate each label by a comma. Labels checks on the webhook can be enforced by setting `multi_runner_config.matcherConfig.exactMatch`. GitHub read-only labels should not be provided."
runner\_group\_name: "Name of the runner group."
runner\_name\_prefix: "Prefix for the GitHub runner name."
runner\_run\_as: "Run the GitHub actions agent as user."
runners\_maximum\_count: "The maximum number of runners that will be created. Setting the variable to `-1` desiables the maximum check."
scale\_down\_schedule\_expression: "Scheduler expression to check every x for scale down."
scale\_up\_reserved\_concurrent\_executions: "Amount of reserved concurrent executions for the scale-up lambda function. A value of 0 disables lambda from being triggered and -1 removes any concurrency limitations."
userdata\_template: "Alternative user-data template, replacing the default template. By providing your own user\_data you have to take care of installing all required software, including the action runner. Variables userdata\_pre/post\_install are ignored."
enable\_jit\_config "Overwrite the default behavior for JIT configuration. By default JIT configuration is enabled for ephemeral runners and disabled for non-ephemeral runners. In case of GHES check first if the JIT config API is available. In case you are upgrading from 3.x to 4.x you can set `enable_jit_config` to `false` to avoid a breaking change when having your own AMI."
enable\_runner\_detailed\_monitoring: "Should detailed monitoring be enabled for the runner. Set this to true if you want to use detailed monitoring. See https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/using-cloudwatch-new.html for details."
enable\_cloudwatch\_agent: "Enabling the cloudwatch agent on the ec2 runner instances, the runner contains default config. Configuration can be overridden via `cloudwatch_config`."
cloudwatch\_config: "(optional) Replaces the module default cloudwatch log config. See https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/CloudWatch-Agent-Configuration-File-Details.html for details."
userdata\_pre\_install: "Script to be ran before the GitHub Actions runner is installed on the EC2 instances"
userdata\_post\_install: "Script to be ran after the GitHub Actions runner is installed on the EC2 instances"
runner\_hook\_job\_started: "Script to be ran in the runner environment at the beginning of every job"
runner\_hook\_job\_completed: "Script to be ran in the runner environment at the end of every job"
runner\_ec2\_tags: "Map of tags that will be added to the launch template instance tag specifications."
runner\_iam\_role\_managed\_policy\_arns: "Attach AWS or customer-managed IAM policies (by ARN) to the runner IAM role"
vpc\_id: "The VPC for security groups of the action runners. If not set uses the value of `var.vpc_id`."
subnet\_ids: "List of subnets in which the action runners will be launched, the subnets needs to be subnets in the `vpc_id`. If not set, uses the value of `var.subnet_ids`."
idle\_config: "List of time period that can be defined as cron expression to keep a minimum amount of runners active instead of scaling down to 0. By defining this list you can ensure that in time periods that match the cron expression within 5 seconds a runner is kept idle."
runner\_log\_files: "(optional) Replaces the module default cloudwatch log config. See https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/CloudWatch-Agent-Configuration-File-Details.html for details."
block\_device\_mappings: "The EC2 instance block device configuration. Takes the following keys: `device_name`, `delete_on_termination`, `volume_type`, `volume_size`, `encrypted`, `iops`, `throughput`, `kms_key_id`, `snapshot_id`."
job\_retry: "Experimental! Can be removed / changed without trigger a major release. Configure job retries. The configuration enables job retries (for ephemeral runners). After creating the instances a message will be published to a job retry queue. The job retry check lambda is checking after a delay if the job is queued. If not the message will be published again on the scale-up (build queue). Using this feature can impact the rate limit of the GitHub app."
pool\_config: "The configuration for updating the pool. The `pool_size` to adjust to by the events triggered by the `schedule_expression`. For example you can configure a cron expression for week days to adjust the pool to 10 and another expression for the weekend to adjust the pool to 1. Use `schedule_expression_timezone` to override the schedule time zone (defaults to UTC)."
}
matcherConfig: {
labelMatchers: "The list of list of labels supported by the runner configuration. `[[self-hosted, linux, x64, example]]`"
exactMatch: "If set to true all labels in the workflow job must match the GitHub labels (os, architecture and `self-hosted`). When false if __any__ workflow label matches it will trigger the webhook."
priority: "If set it defines the priority of the matcher, the matcher with the lowest priority will be evaluated first. Default is 999, allowed values 0-999."
}
redrive\_build\_queue: "Set options to attach (optional) a dead letter queue to the build queue, the queue between the webhook and the scale up lambda. You have the following options. 1. Disable by setting `enabled` to false. 2. Enable by setting `enabled` to `true`, `maxReceiveCount` to a number of max retries."
} | 
map(object({
    runner_config = object({
      runner_os           = string
      runner_architecture = string
      runner_metadata_options = optional(map(any), {
        instance_metadata_tags      = "enabled"
        http_endpoint               = "enabled"
        http_tokens                 = "required"
        http_put_response_hop_limit = 1
      })
      ami = optional(object({
        filter               = optional(map(list(string)), { state = ["available"] })
        owners               = optional(list(string), ["amazon"])
        id_ssm_parameter_arn = optional(string, null)
        kms_key_arn          = optional(string, null)
      }), null)
      create_service_linked_role_spot         = optional(bool, false)
      credit_specification                    = optional(string, null)
      delay_webhook_event                     = optional(number, 30)
      disable_runner_autoupdate               = optional(bool, false)
      ebs_optimized                           = optional(bool, false)
      enable_ephemeral_runners                = optional(bool, false)
      enable_job_queued_check                 = optional(bool, null)
      enable_on_demand_failover_for_errors    = optional(list(string), [])
      enable_organization_runners             = optional(bool, false)
      enable_runner_binaries_syncer           = optional(bool, true)
      enable_ssm_on_runners                   = optional(bool, false)
      enable_userdata                         = optional(bool, true)
      instance_allocation_strategy            = optional(string, "lowest-price")
      instance_max_spot_price                 = optional(string, null)
      instance_target_capacity_type           = optional(string, "spot")
      instance_types                          = list(string)
      job_queue_retention_in_seconds          = optional(number, 86400)
      minimum_running_time_in_minutes         = optional(number, null)
      pool_runner_owner                       = optional(string, null)
      runner_as_root                          = optional(bool, false)
      runner_boot_time_in_minutes             = optional(number, 5)
      runner_disable_default_labels           = optional(bool, false)
      runner_extra_labels                     = optional(list(string), [])
      runner_group_name                       = optional(string, "Default")
      runner_name_prefix                      = optional(string, "")
      runner_run_as                           = optional(string, "ec2-user")
      runners_maximum_count                   = number
      runner_additional_security_group_ids    = optional(list(string), [])
      scale_down_schedule_expression          = optional(string, "cron(*/5 * * * ? *)")
      scale_up_reserved_concurrent_executions = optional(number, 1)
      userdata_template                       = optional(string, null)
      userdata_content                        = optional(string, null)
      enable_jit_config                       = optional(bool, null)
      enable_runner_detailed_monitoring       = optional(bool, false)
      enable_cloudwatch_agent                 = optional(bool, true)
      cloudwatch_config                       = optional(string, null)
      userdata_pre_install                    = optional(string, "")
      userdata_post_install                   = optional(string, "")
      runner_hook_job_started                 = optional(string, "")
      runner_hook_job_completed               = optional(string, "")
      runner_ec2_tags                         = optional(map(string), {})
      runner_iam_role_managed_policy_arns     = optional(list(string), [])
      vpc_id                                  = optional(string, null)
      subnet_ids                              = optional(list(string), null)
      idle_config = optional(list(object({
        cron             = string
        timeZone         = string
        idleCount        = number
        evictionStrategy = optional(string, "oldest_first")
      })), [])
      cpu_options = optional(object({
        core_count       = number
        threads_per_core = number
      }), null)
      placement = optional(object({
        affinity                = optional(string)
        availability_zone       = optional(string)
        group_id                = optional(string)
        group_name              = optional(string)
        host_id                 = optional(string)
        host_resource_group_arn = optional(number)
        spread_domain           = optional(string)
        tenancy                 = optional(string)
        partition_number        = optional(number)
      }), null)
      runner_log_files = optional(list(object({
        log_group_name   = string
        prefix_log_group = bool
        file_path        = string
        log_stream_name  = string
      })), null)
      block_device_mappings = optional(list(object({
        delete_on_termination = optional(bool, true)
        device_name           = optional(string, "/dev/xvda")
        encrypted             = optional(bool, true)
        iops                  = optional(number)
        kms_key_id            = optional(string)
        snapshot_id           = optional(string)
        throughput            = optional(number)
        volume_size           = number
        volume_type           = optional(string, "gp3")
        })), [{
        volume_size = 30
      }])
      pool_config = optional(list(object({
        schedule_expression          = string
        schedule_expression_timezone = optional(string)
        size                         = number
      })), [])
      job_retry = optional(object({
        enable             = optional(bool, false)
        delay_in_seconds   = optional(number, 300)
        delay_backoff      = optional(number, 2)
        lambda_memory_size = optional(number, 256)
        lambda_timeout     = optional(number, 30)
        max_attempts       = optional(number, 1)
      }), {})
    })
    matcherConfig = object({
      labelMatchers = list(list(string))
      exactMatch    = optional(bool, false)
      priority      = optional(number, 999)
    })
    redrive_build_queue = optional(object({
      enabled         = bool
      maxReceiveCount = number
      }), {
      enabled         = false
      maxReceiveCount = null
    })
  }))
| n/a | yes | | [pool\_lambda\_reserved\_concurrent\_executions](#input\_pool\_lambda\_reserved\_concurrent\_executions) | Amount of reserved concurrent executions for the scale-up lambda function. A value of 0 disables lambda from being triggered and -1 removes any concurrency limitations. | `number` | `1` | no | | [pool\_lambda\_timeout](#input\_pool\_lambda\_timeout) | Time out for the pool lambda in seconds. | `number` | `60` | no | | [prefix](#input\_prefix) | The prefix used for naming resources | `string` | `"github-actions"` | no | diff --git a/modules/multi-runner/variables.tf b/modules/multi-runner/variables.tf index 0ca473ecf2..2262918a37 100644 --- a/modules/multi-runner/variables.tf +++ b/modules/multi-runner/variables.tf @@ -185,7 +185,7 @@ variable "multi_runner_config" { description = < [role\_path](#input\_role\_path) | The path that will be added to the role, if not set the environment name will be used. | `string` | `null` | no | | [role\_permissions\_boundary](#input\_role\_permissions\_boundary) | Permissions boundary that will be added to the created role for the lambda. | `string` | `null` | no | | [runner\_architecture](#input\_runner\_architecture) | The platform architecture of the runner instance\_type. | `string` | `"x64"` | no | -| [runner\_os](#input\_runner\_os) | The EC2 Operating System type to use for action runner instances (linux,windows). | `string` | `"linux"` | no | +| [runner\_os](#input\_runner\_os) | The EC2 Operating System type to use for action runner instances (linux, osx, windows). | `string` | `"linux"` | no | | [s3\_logging\_bucket](#input\_s3\_logging\_bucket) | Bucket for action runner distribution bucket access logging. | `string` | `null` | no | | [s3\_logging\_bucket\_prefix](#input\_s3\_logging\_bucket\_prefix) | Bucket prefix for action runner distribution bucket access logging. | `string` | `null` | no | | [s3\_tags](#input\_s3\_tags) | Map of tags that will be added to the S3 bucket. Note these are additional tags to the default tags. | `map(string)` | `{}` | no | diff --git a/modules/runner-binaries-syncer/main.tf b/modules/runner-binaries-syncer/main.tf index c1cbf382af..2b1e212262 100644 --- a/modules/runner-binaries-syncer/main.tf +++ b/modules/runner-binaries-syncer/main.tf @@ -1,5 +1,5 @@ locals { - action_runner_distribution_object_key = "actions-runner-${var.runner_os}.${var.runner_os == "linux" ? "tar.gz" : "zip"}" + action_runner_distribution_object_key = "actions-runner-${var.runner_os}.${var.runner_os == "windows" ? "zip" : "tar.gz"}" } resource "aws_s3_bucket" "action_dist" { diff --git a/modules/runner-binaries-syncer/variables.tf b/modules/runner-binaries-syncer/variables.tf index dd16a7c3ee..970f8641ff 100644 --- a/modules/runner-binaries-syncer/variables.tf +++ b/modules/runner-binaries-syncer/variables.tf @@ -99,13 +99,13 @@ variable "role_path" { } variable "runner_os" { - description = "The EC2 Operating System type to use for action runner instances (linux,windows)." + description = "The EC2 Operating System type to use for action runner instances (linux, osx, windows)." type = string default = "linux" validation { - condition = contains(["linux", "windows"], var.runner_os) - error_message = "Valid values for runner_os are (linux, windows)." + condition = contains(["linux", "osx", "windows"], var.runner_os) + error_message = "Valid values for runner_os are (linux, osx, windows)." } } diff --git a/modules/runners/README.md b/modules/runners/README.md index 9bb3a6f4e6..aa4ba8df17 100644 --- a/modules/runners/README.md +++ b/modules/runners/README.md @@ -168,7 +168,7 @@ yarn run dist | [instance\_max\_spot\_price](#input\_instance\_max\_spot\_price) | Max price price for spot instances per hour. This variable will be passed to the create fleet as max spot price for the fleet. | `string` | `null` | no | | [instance\_profile\_path](#input\_instance\_profile\_path) | The path that will be added to the instance\_profile, if not set the prefix will be used. | `string` | `null` | no | | [instance\_target\_capacity\_type](#input\_instance\_target\_capacity\_type) | Default lifecycle used runner instances, can be either `spot` or `on-demand`. | `string` | `"spot"` | no | -| [instance\_types](#input\_instance\_types) | List of instance types for the action runner. Defaults are based on runner\_os (al2023 for linux and Windows Server Core for win). | `list(string)` | `null` | no | +| [instance\_types](#input\_instance\_types) | List of instance types for the action runner. Defaults are based on runner\_os (al2023 for linux, macOS Sequoia for osx, Windows Server Core for win). | `list(string)` | `null` | no | | [job\_retry](#input\_job\_retry) | Configure job retries. The configuration enables job retries (for ephemeral runners). After creating the instances a message will be published to a job retry queue. The job retry check lambda is checking after a delay if the job is queued. If not the message will be published again on the scale-up (build queue). Using this feature can impact the rate limit of the GitHub app.

`enable`: Enable or disable the job retry feature.
`delay_in_seconds`: The delay in seconds before the job retry check lambda will check the job status.
`delay_backoff`: The backoff factor for the delay.
`lambda_memory_size`: Memory size limit in MB for the job retry check lambda.
'lambda\_reserved\_concurrent\_executions': Amount of reserved concurrent executions for the job retry check lambda function. A value of 0 disables lambda from being triggered and -1 removes any concurrency limitations.
`lambda_timeout`: Time out of the job retry check lambda in seconds.
`max_attempts`: The maximum number of attempts to retry the job. | 
object({
    enable                                = optional(bool, false)
    delay_in_seconds                      = optional(number, 300)
    delay_backoff                         = optional(number, 2)
    lambda_memory_size                    = optional(number, 256)
    lambda_reserved_concurrent_executions = optional(number, 1)

    lambda_timeout = optional(number, 30)

    max_attempts = optional(number, 1)
  })
| `{}` | no | | [key\_name](#input\_key\_name) | Key pair name | `string` | `null` | no | | [kms\_key\_arn](#input\_kms\_key\_arn) | Optional CMK Key ARN to be used for Parameter Store. | `string` | `null` | no | @@ -214,7 +214,7 @@ yarn run dist | [runner\_labels](#input\_runner\_labels) | All the labels for the runners (GitHub) including the default one's(e.g: self-hosted, linux, x64, label1, label2). Separate each label by a comma | `list(string)` | n/a | yes | | [runner\_log\_files](#input\_runner\_log\_files) | (optional) List of logfiles to send to CloudWatch, will only be used if `enable_cloudwatch_agent` is set to true. Object description: `log_group_name`: Name of the log group, `prefix_log_group`: If true, the log group name will be prefixed with `/github-self-hosted-runners/`, `file_path`: path to the log file, `log_stream_name`: name of the log stream. | 
list(object({
    log_group_name   = string
    prefix_log_group = bool
    file_path        = string
    log_stream_name  = string
  }))
| `null` | no | | [runner\_name\_prefix](#input\_runner\_name\_prefix) | The prefix used for the GitHub runner name. The prefix will be used in the default start script to prefix the instance name when register the runner in GitHub. The value is available via an EC2 tag 'ghr:runner\_name\_prefix'. | `string` | `""` | no | -| [runner\_os](#input\_runner\_os) | The EC2 Operating System type to use for action runner instances (linux,windows). | `string` | `"linux"` | no | +| [runner\_os](#input\_runner\_os) | The EC2 Operating System type to use for action runner instances (linux, osx, windows). | `string` | `"linux"` | no | | [runner\_run\_as](#input\_runner\_run\_as) | Run the GitHub actions agent as user. | `string` | `"ec2-user"` | no | | [runners\_lambda\_s3\_key](#input\_runners\_lambda\_s3\_key) | S3 key for runners lambda function. Required if using S3 bucket to specify lambdas. | `string` | `null` | no | | [runners\_lambda\_s3\_object\_version](#input\_runners\_lambda\_s3\_object\_version) | S3 object version for runners lambda function. Useful if S3 versioning is enabled on source bucket. | `string` | `null` | no | diff --git a/modules/runners/main.tf b/modules/runners/main.tf index 5522c5fb45..7e2017d8b5 100644 --- a/modules/runners/main.tf +++ b/modules/runners/main.tf @@ -20,21 +20,25 @@ locals { default_ami = { "windows" = { name = ["Windows_Server-2022-English-Full-ECS_Optimized-*"] } "linux" = var.runner_architecture == "arm64" ? { name = ["al2023-ami-2023.*-kernel-6.*-arm64"] } : { name = ["al2023-ami-2023.*-kernel-6.*-x86_64"] } + "osx" = var.runner_architecture == "arm64" ? { name = ["amzn-ec2-macos-15.*-arm64"] } : { name = ["amzn-ec2-macos-15.*"] } } default_userdata_template = { "windows" = "${path.module}/templates/user-data.ps1" "linux" = "${path.module}/templates/user-data.sh" + "osx" = "${path.module}/templates/user-data-osx.sh" } userdata_install_runner = { "windows" = "${path.module}/templates/install-runner.ps1" "linux" = "${path.module}/templates/install-runner.sh" + "osx" = "${path.module}/templates/install-runner-osx.sh" } userdata_start_runner = { "windows" = "${path.module}/templates/start-runner.ps1" "linux" = "${path.module}/templates/start-runner.sh" + "osx" = "${path.module}/templates/start-runner-osx.sh" } # Handle AMI configuration diff --git a/modules/runners/scale-down-state-diagram.md b/modules/runners/scale-down-state-diagram.md index b4f260eb2a..64e32bc141 100644 --- a/modules/runners/scale-down-state-diagram.md +++ b/modules/runners/scale-down-state-diagram.md @@ -117,7 +117,7 @@ stateDiagram-v2 note right of CheckMinimumTime Minimum running time in minutes - (Linux: 5min, Windows: 15min) + (Linux: 5min, Windows: 15min, OSX: 20min) end note note right of CheckBootTime @@ -145,6 +145,6 @@ stateDiagram-v2 ## Configuration Parameters - **Cron Schedule**: `cron(*/5 * * * ? *)` (every 5 minutes) -- **Minimum Runtime**: Linux 5min, Windows 15min +- **Minimum Runtime**: Linux 5min, Windows 15min, OSX 20min - **Boot Timeout**: Configurable via `runner_boot_time_in_minutes` - **Idle Config**: Per-environment configuration for desired idle runners diff --git a/modules/runners/scale-down.tf b/modules/runners/scale-down.tf index d274e3d4f1..f3df5141b3 100644 --- a/modules/runners/scale-down.tf +++ b/modules/runners/scale-down.tf @@ -1,8 +1,11 @@ locals { # Windows Runners can take their sweet time to do anything + # For an AWS vended AMI with a x86 Mac instance or a Apple silicon Mac instance, + # the launch time can range from approximately 6 minutes to 20 minutes. min_runtime_defaults = { "windows" = 15 "linux" = 5 + "osx" = 20 } } resource "aws_lambda_function" "scale_down" { diff --git a/modules/runners/templates/install-runner-osx.sh b/modules/runners/templates/install-runner-osx.sh new file mode 100644 index 0000000000..893a62b098 --- /dev/null +++ b/modules/runners/templates/install-runner-osx.sh @@ -0,0 +1,60 @@ +# shellcheck shell=bash + +## install the runner (macOS) + +s3_location=${S3_LOCATION_RUNNER_DISTRIBUTION} +architecture=${RUNNER_ARCHITECTURE} + +if [ -z "$RUNNER_TARBALL_URL" ] && [ -z "$s3_location" ]; then + echo "Neither RUNNER_TARBALL_URL or s3_location are set" + exit 1 +fi + +file_name="actions-runner.tar.gz" + +echo "Setting up GH Actions runner tool cache" +mkdir -p /opt/hostedtoolcache + +echo "Creating actions-runner directory for the GH Action installation" +sudo mkdir -p /opt/actions-runner +cd /opt/actions-runner || exit 1 + +if [[ -n "$RUNNER_TARBALL_URL" ]]; then + echo "Downloading the GH Action runner from $RUNNER_TARBALL_URL to $file_name" + curl -s -o "$file_name" -L "$RUNNER_TARBALL_URL" +else + echo "Retrieving REGION from AWS API" + token="$(curl -s -f -X PUT "http://169.254.169.254/latest/api/token" \ + -H "X-aws-ec2-metadata-token-ttl-seconds: 180")" + + region="$(curl -s -f -H "X-aws-ec2-metadata-token: $token" \ + http://169.254.169.254/latest/dynamic/instance-identity/document | jq -r .region)" + echo "Retrieved REGION from AWS API ($region)" + + echo "Downloading the GH Action runner from s3 bucket $s3_location" + aws s3 cp "$s3_location" "$file_name" --region "$region" --no-progress +fi + +echo "Un-tar action runner" +tar xzf "./$file_name" +echo "Delete tar file" +rm -rf "$file_name" + +os_name=$(sw_vers -productName 2>/dev/null || echo "macOS") +os_version=$(sw_vers -productVersion 2>/dev/null || echo "unknown") +arch_name=$(uname -m) + +echo "OS: $os_name $os_version ($arch_name)" + +if ! command -v brew >/dev/null 2>&1; then + echo "Homebrew not found; skipping dependency installation via brew" +else + echo "Homebrew detected; install any macOS-specific dependencies here if needed" + # Example: brew install jq awscli +fi + +user_name="${RUNNER_USER:-ec2-user}" + +echo "Set file ownership of action runner" +sudo chown -R "$user_name":"$user_name" /opt/actions-runner +sudo chown -R "$user_name":"$user_name" /opt/hostedtoolcache diff --git a/modules/runners/templates/start-runner-osx.sh b/modules/runners/templates/start-runner-osx.sh new file mode 100644 index 0000000000..20541a2312 --- /dev/null +++ b/modules/runners/templates/start-runner-osx.sh @@ -0,0 +1,196 @@ +#!/bin/bash + +set -euo pipefail + +# macOS variant of start-runner.sh + +tag_instance_with_runner_id() { + echo "Checking for .runner file to extract agent ID" + + if [[ ! -f "/opt/actions-runner/.runner" ]]; then + echo "Warning: .runner file not found" + return 0 + fi + + echo "Found .runner file, extracting agent ID" + local agent_id + agent_id=$(jq -r '.agentId' /opt/actions-runner/.runner 2>/dev/null || echo "") + + if [[ -z "$agent_id" || "$agent_id" == "null" ]]; then + echo "Warning: Could not extract agent ID from .runner file" + return 0 + fi + + echo "Tagging instance with GitHub runner agent ID: $agent_id" + if aws ec2 create-tags \ + --region "$region" \ + --resources "$instance_id" \ + --tags Key=ghr:github_runner_id,Value="$agent_id"; then + echo "Successfully tagged instance with agent ID: $agent_id" + return 0 + else + echo "Warning: Failed to tag instance with agent ID" + return 0 + fi +} + +cleanup() { + local exit_code="$1" + + if [ "$exit_code" -ne 0 ]; then + echo "ERROR: runner-start-failed with exit code $exit_code" + fi + + if [ "$agent_mode" = "ephemeral" ] || [ "$exit_code" -ne 0 ]; then + echo "Terminating instance" + aws ec2 terminate-instances \ + --instance-ids "$instance_id" \ + --region "$region" || true + fi +} + +trap 'cleanup $?' EXIT + +echo "Retrieving TOKEN from AWS API" +token=$(curl -f -X PUT "http://169.254.169.254/latest/api/token" \ + -H "X-aws-ec2-metadata-token-ttl-seconds: 180" || true) +if [ -z "$token" ]; then + retrycount=0 + until [ -n "$token" ]; do + echo "Failed to retrieve token. Retrying in 5 seconds." + sleep 5 + token=$(curl -f -X PUT "http://169.254.169.254/latest/api/token" \ + -H "X-aws-ec2-metadata-token-ttl-seconds: 180" || true) + retrycount=$((retrycount + 1)) + if [ $retrycount -gt 40 ]; then + break + fi + done +fi + +region=$(curl -f -H "X-aws-ec2-metadata-token: $token" \ + http://169.254.169.254/latest/dynamic/instance-identity/document | jq -r .region) +echo "Retrieved REGION from AWS API ($region)" + +instance_id=$(curl -f -H "X-aws-ec2-metadata-token: $token" \ + http://169.254.169.254/latest/meta-data/instance-id) +echo "Retrieved INSTANCE_ID from AWS API ($instance_id)" + +availability_zone=$(curl -f -H "X-aws-ec2-metadata-token: $token" \ + http://169.254.169.254/latest/meta-data/placement/availability-zone) + +environment=$(curl -f -H "X-aws-ec2-metadata-token: $token" \ + http://169.254.169.254/latest/meta-data/tags/instance/ghr:environment || echo "") +ssm_config_path=$(curl -f -H "X-aws-ec2-metadata-token: $token" \ + http://169.254.169.254/latest/meta-data/tags/instance/ghr:ssm_config_path || echo "") +runner_name_prefix=$(curl -f -H "X-aws-ec2-metadata-token: $token" \ + http://169.254.169.254/latest/meta-data/tags/instance/ghr:runner_name_prefix || echo "") + +echo "Retrieved ghr:environment tag - ($environment)" +echo "Retrieved ghr:ssm_config_path tag - ($ssm_config_path)" +echo "Retrieved ghr:runner_name_prefix tag - ($runner_name_prefix)" + +parameters=$(aws ssm get-parameters-by-path \ + --path "$ssm_config_path" \ + --region "$region" \ + --query "Parameters[*].{Name:Name,Value:Value}") +echo "Retrieved parameters from AWS SSM ($parameters)" + +run_as=$(echo "$parameters" | jq -r '.[] | select(.Name == "'$ssm_config_path'/run_as") | .Value') +echo "Retrieved /$ssm_config_path/run_as parameter - ($run_as)" + +agent_mode=$(echo "$parameters" | jq -r '.[] | select(.Name == "'$ssm_config_path'/agent_mode") | .Value') +echo "Retrieved /$ssm_config_path/agent_mode parameter - ($agent_mode)" + +disable_default_labels=$(echo "$parameters" | jq -r '.[] | select(.Name == "'$ssm_config_path'/disable_default_labels") | .Value') +echo "Retrieved /$ssm_config_path/disable_default_labels parameter - ($disable_default_labels)" + +enable_jit_config=$(echo "$parameters" | jq -r '.[] | select(.Name == "'$ssm_config_path'/enable_jit_config") | .Value') +echo "Retrieved /$ssm_config_path/enable_jit_config parameter - ($enable_jit_config)" + +token_path=$(echo "$parameters" | jq -r '.[] | select(.Name == "'$ssm_config_path'/token_path") | .Value') +echo "Retrieved /$ssm_config_path/token_path parameter - ($token_path)" + +echo "Get GH Runner config from AWS SSM" +config=$(aws ssm get-parameter \ + --name "$token_path"/"$instance_id" \ + --with-decryption \ + --region "$region" | jq -r ".Parameter | .Value") + +while [[ -z "$config" ]]; do + echo "Waiting for GH Runner config to become available in AWS SSM" + sleep 1 + config=$(aws ssm get-parameter \ + --name "$token_path"/"$instance_id" \ + --with-decryption \ + --region "$region" | jq -r ".Parameter | .Value") +done + +echo "Delete GH Runner token from AWS SSM" +aws ssm delete-parameter --name "$token_path"/"$instance_id" --region "$region" + +if [ -z "$run_as" ]; then + echo "No user specified, using default ec2-user account" + run_as="ec2-user" +fi + +if [[ "$run_as" == "root" ]]; then + echo "run_as is set to root - export RUNNER_ALLOW_RUNASROOT=1" + export RUNNER_ALLOW_RUNASROOT=1 +fi + +sudo chown -R "$run_as" /opt/actions-runner + +info_arch=$(uname -m) +info_os=$(sw_vers -productName 2>/dev/null || echo "macOS") +info_ver=$(sw_vers -productVersion 2>/dev/null || echo "unknown") + +tee /opt/actions-runner/.setup_info < >(tee /var/log/user-data.log | logger -t user-data -s 2>/dev/console) 2>&1 + +# macOS user-data for GitHub Actions runners + +set +x + +%{ if enable_debug_logging } +set -x +%{ endif } + +${pre_install} + +# On macOS we don't use dnf; assume base image has required tools +# Optionally use brew here if needed +if command -v brew >/dev/null 2>&1; then + echo "Homebrew detected; you can install extra dependencies via brew if needed" +fi + +user_name=ec2-user + +${install_runner} + +${post_install} + +# Register runner job hooks +# Ref: https://docs.github.com/en/actions/hosting-your-own-runners/managing-self-hosted-runners/running-scripts-before-or-after-a-job +%{ if hook_job_started != "" } +cat > /opt/actions-runner/hook_job_started.sh <<'EOF' +${hook_job_started} +EOF +echo ACTIONS_RUNNER_HOOK_JOB_STARTED=/opt/actions-runner/hook_job_started.sh | tee -a /opt/actions-runner/.env +%{ endif } + +%{ if hook_job_completed != "" } +cat > /opt/actions-runner/hook_job_completed.sh <<'EOF' +${hook_job_completed} +EOF +echo ACTIONS_RUNNER_HOOK_JOB_COMPLETED=/opt/actions-runner/hook_job_completed.sh | tee -a /opt/actions-runner/.env +%{ endif } + +${start_runner} diff --git a/modules/runners/variables.tf b/modules/runners/variables.tf index 8c8a1a136c..39f94a4248 100644 --- a/modules/runners/variables.tf +++ b/modules/runners/variables.tf @@ -119,18 +119,18 @@ variable "instance_max_spot_price" { } variable "runner_os" { - description = "The EC2 Operating System type to use for action runner instances (linux,windows)." + description = "The EC2 Operating System type to use for action runner instances (linux, osx, windows)." type = string default = "linux" validation { - condition = contains(["linux", "windows"], var.runner_os) - error_message = "Valid values for runner_os are (linux, windows)." + condition = contains(["linux", "osx", "windows"], var.runner_os) + error_message = "Valid values for runner_os are (linux, osx, windows)." } } variable "instance_types" { - description = "List of instance types for the action runner. Defaults are based on runner_os (al2023 for linux and Windows Server Core for win)." + description = "List of instance types for the action runner. Defaults are based on runner_os (al2023 for linux, macOS Sequoia for osx, Windows Server Core for win)." type = list(string) default = null } diff --git a/variables.tf b/variables.tf index bc9d3abac5..1bd7f3bc15 100644 --- a/variables.tf +++ b/variables.tf @@ -542,7 +542,7 @@ variable "instance_max_spot_price" { } variable "instance_types" { - description = "List of instance types for the action runner. Defaults are based on runner_os (al2023 for linux and Windows Server Core for win)." + description = "List of instance types for the action runner. Defaults are based on runner_os (al2023 for linux, macOS Sequoia for osx, Windows Server Core for win)." type = list(string) default = ["m5.large", "c5.large"] } @@ -658,13 +658,13 @@ variable "enable_managed_runner_security_group" { } variable "runner_os" { - description = "The EC2 Operating System type to use for action runner instances (linux,windows)." + description = "The EC2 Operating System type to use for action runner instances (linux, osx, windows)." type = string default = "linux" validation { - condition = contains(["linux", "windows"], var.runner_os) - error_message = "Valid values for runner_os are (linux, windows)." + condition = contains(["linux", "osx", "windows"], var.runner_os) + error_message = "Valid values for runner_os are (linux, osx, windows)." } }