AWS::Batch::ComputeEnvironment
The AWS::Batch::ComputeEnvironment
resource defines your Amazon Batch compute environment. You can define MANAGED
or
UNMANAGED
compute environments. MANAGED
compute environments
can use Amazon EC2 or Amazon Fargate resources. UNMANAGED
compute environments can only use EC2 resources. For more information, see Compute
Environments in the
Amazon Batch User Guide
.
In a managed compute environment, Amazon Batch manages the capacity and instance types of the compute resources within the environment. This is based on the compute resource specification that you define or the launch template that you specify when you create the compute environment. You can choose either to use EC2 On-Demand Instances and EC2 Spot Instances, or to use Fargate and Fargate Spot capacity in your managed compute environment. You can optionally set a maximum price so that Spot Instances only launch when the Spot Instance price is below a specified percentage of the On-Demand price.
Note
Multi-node parallel jobs are not supported on Spot Instances.
In an unmanaged compute environment, you can manage your own EC2 compute resources and have a lot of flexibility with how you configure your compute resources. For example, you can use custom AMI. However, you need to verify that your AMI meets the Amazon ECS container instance AMI specification. For more information, see container instance AMIs in the Amazon Elastic Container Service Developer Guide. After you have created your unmanaged compute environment, you can use the DescribeComputeEnvironments operation to find the Amazon ECS cluster that is associated with it. Then, manually launch your container instances into that Amazon ECS cluster. For more information, see Launching an Amazon ECS container instance in the Amazon Elastic Container Service Developer Guide.
Note
To create a compute environment that uses EKS resources, the caller must have
permissions to call eks:DescribeCluster
.
Note
Amazon Batch doesn't upgrade the AMIs in a compute environment after it's created except under specific conditions. For example, it doesn't automatically update the AMIs when a newer version of the Amazon ECS optimized AMI is available. Therefore, you're responsible for the management of the guest operating system (including updates and security patches) and any additional application software or utilities that you install on the compute resources. There are two ways to use a new AMI for your Amazon Batch jobs. The original method is to complete these steps:
-
Create a new compute environment with the new AMI.
-
Add the compute environment to an existing job queue.
-
Remove the earlier compute environment from your job queue.
-
Delete the earlier compute environment.
In April 2022, Amazon Batch added enhanced support for updating compute
environments. For example, the UpdateComputeEnvironent
API lets you use the
ReplaceComputeEnvironment
property to dynamically update compute
environment parameters such as the launch template or instance type without replacement.
For more information, see Updating compute
environments in the
Amazon Batch User Guide.
To use the enhanced updating of compute environments to update AMIs, follow these rules:
-
Either do not set the ServiceRole property or set it to the AWSServiceRoleForBatch service-linked role.
-
Set the AllocationStrategy property to
BEST_FIT_PROGRESSIVE
orSPOT_CAPACITY_OPTIMIZED
. -
Set the ReplaceComputeEnvironment property to
false
.Important
Set the
ReplaceComputeEnvironment
property totrue
if the compute environment uses theBEST_FIT
allocation strategy.Note
If the
ReplaceComputeEnvironment
property is set tofalse
, you might receive an error message when you update the CFN template for a compute environment. This issue occurs if the updateddesiredvcpus
value is less than the currentdesiredvcpus
value. As a workaround, delete thedesiredvcpus
value from the updated template or use theminvcpus
property to manage the number of vCPUs. For information, see Error message when you update theDesiredvCpus
setting. -
Set the UpdateToLatestImageVersion property to
true
. This property is used when you update a compute environment. The UpdateToLatestImageVersion property is ignored when you create a compute environment. -
Either do not specify an image ID in ImageId or ImageIdOverride properties, or in the launch template identified by the Launch Template property. In that case Amazon Batch will select the latest Amazon ECS optimized AMI supported by Amazon Batch at the time the infrastructure update is initiated. Alternatively you can specify the AMI ID in the
ImageId
orImageIdOverride
properties, or the launch template identified by theLaunchTemplate
properties. Changing any of these properties will trigger an infrastructure update.
If these rules are followed, any update that triggers an infrastructure update will
cause the AMI ID to be re-selected. If the Version property of the LaunchTemplateSpecification is set to $Latest
or
$Default
, the latest or default version of the launch template will be
evaluated up at the time of the infrastructure update, even if the
LaunchTemplateSpecification
was not updated.
Syntax
To declare this entity in your Amazon CloudFormation template, use the following syntax:
JSON
{ "Type" : "AWS::Batch::ComputeEnvironment", "Properties" : { "ComputeEnvironmentName" :
String
, "ComputeResources" :ComputeResources
, "Context" :String
, "EksConfiguration" :EksConfiguration
, "ReplaceComputeEnvironment" :Boolean
, "ServiceRole" :String
, "State" :String
, "Tags" :{
, "Type" :Key
:Value
, ...}String
, "UnmanagedvCpus" :Integer
, "UpdatePolicy" :UpdatePolicy
} }
YAML
Type: AWS::Batch::ComputeEnvironment Properties: ComputeEnvironmentName:
String
ComputeResources:ComputeResources
Context:String
EksConfiguration:EksConfiguration
ReplaceComputeEnvironment:Boolean
ServiceRole:String
State:String
Tags:Type:
Key
:Value
String
UnmanagedvCpus:Integer
UpdatePolicy:UpdatePolicy
Properties
ComputeEnvironmentName
-
The name for your compute environment. It can be up to 128 characters long. It can contain uppercase and lowercase letters, numbers, hyphens (-), and underscores (_).
Required: No
Type: String
Update requires: Replacement
ComputeResources
-
The ComputeResources property type specifies details of the compute resources managed by the compute environment. This parameter is required for managed compute environments. For more information, see Compute Environments in the Amazon Batch User Guide .
Required: No
Type: ComputeResources
Update requires: No interruption
Context
-
Reserved.
Required: No
Type: String
Minimum:
0
Maximum:
256
Update requires: No interruption
EksConfiguration
-
The details for the Amazon EKS cluster that supports the compute environment.
Required: No
Type: EksConfiguration
Update requires: Replacement
ReplaceComputeEnvironment
-
Specifies whether the compute environment is replaced if an update is made that requires replacing the instances in the compute environment. The default value is
true
. To enable more properties to be updated, set this property tofalse
. When changing the value of this property tofalse
, do not change any other properties at the same time. If other properties are changed at the same time, and the change needs to be rolled back but it can't, it's possible for the stack to go into theUPDATE_ROLLBACK_FAILED
state. You can't update a stack that is in theUPDATE_ROLLBACK_FAILED
state. However, if you can continue to roll it back, you can return the stack to its original settings and then try to update it again. For more information, see Continue rolling back an update in the Amazon CloudFormation User Guide.The properties that can't be changed without replacing the compute environment are in the
ComputeResources
property type:AllocationStrategy
,BidPercentage
,Ec2Configuration
,Ec2KeyPair
,Ec2KeyPair
,ImageId
,InstanceRole
,InstanceTypes
,LaunchTemplate
,MaxvCpus
,MinvCpus
,PlacementGroup
,SecurityGroupIds
,Subnets
, Tags,Type
, andUpdateToLatestImageVersion
.Required: No
Type: Boolean
Update requires: No interruption
ServiceRole
-
The full Amazon Resource Name (ARN) of the IAM role that allows Amazon Batch to make calls to other Amazon services on your behalf. For more information, see Amazon Batch service IAM role in the Amazon Batch User Guide.
Important
If your account already created the Amazon Batch service-linked role, that role is used by default for your compute environment unless you specify a different role here. If the Amazon Batch service-linked role doesn't exist in your account, and no role is specified here, the service attempts to create the Amazon Batch service-linked role in your account.
If your specified role has a path other than
/
, then you must specify either the full role ARN (recommended) or prefix the role name with the path. For example, if a role with the namebar
has a path of/foo/
, specify/foo/bar
as the role name. For more information, see Friendly names and paths in the IAM User Guide.Note
Depending on how you created your Amazon Batch service role, its ARN might contain the
service-role
path prefix. When you only specify the name of the service role, Amazon Batch assumes that your ARN doesn't use theservice-role
path prefix. Because of this, we recommend that you specify the full ARN of your service role when you create compute environments.Required: No
Type: String
Update requires: No interruption
State
-
The state of the compute environment. If the state is
ENABLED
, then the compute environment accepts jobs from a queue and can scale out automatically based on queues.If the state is
ENABLED
, then the Amazon Batch scheduler can attempt to place jobs from an associated job queue on the compute resources within the environment. If the compute environment is managed, then it can scale its instances out or in automatically, based on the job queue demand.If the state is
DISABLED
, then the Amazon Batch scheduler doesn't attempt to place jobs within the environment. Jobs in aSTARTING
orRUNNING
state continue to progress normally. Managed compute environments in theDISABLED
state don't scale out.Note
Compute environments in a
DISABLED
state may continue to incur billing charges. To prevent additional charges, turn off and then delete the compute environment. For more information, see State in the Amazon Batch User Guide.When an instance is idle, the instance scales down to the
minvCpus
value. However, the instance size doesn't change. For example, consider ac5.8xlarge
instance with aminvCpus
value of4
and adesiredvCpus
value of36
. This instance doesn't scale down to ac5.large
instance.Required: No
Type: String
Allowed values:
ENABLED | DISABLED
Update requires: No interruption
-
The tags applied to the compute environment.
Required: No
Type: Object of String
Pattern:
.*
Update requires: Replacement
Type
-
The type of the compute environment:
MANAGED
orUNMANAGED
. For more information, see Compute Environments in the Amazon Batch User Guide.Required: Yes
Type: String
Allowed values:
MANAGED | UNMANAGED
Update requires: Replacement
UnmanagedvCpus
-
The maximum number of vCPUs for an unmanaged compute environment. This parameter is only used for fair share scheduling to reserve vCPU capacity for new share identifiers. If this parameter isn't provided for a fair share job queue, no vCPU capacity is reserved.
Note
This parameter is only supported when the
type
parameter is set toUNMANAGED
.Required: No
Type: Integer
Update requires: No interruption
UpdatePolicy
-
Specifies the infrastructure update policy for the compute environment. For more information about infrastructure updates, see Updating compute environments in the Amazon Batch User Guide.
Required: No
Type: UpdatePolicy
Update requires: No interruption
Return values
Ref
When you pass the logical ID of this resource to the intrinsic Ref
function, Ref
returns the compute environment ARN, such as
arn:aws-cn:batch:us-east-1:555555555555:compute-environment/M4OnDemand
.
For more information about using the Ref
function, see Ref
.
Fn::GetAtt
The Fn::GetAtt
intrinsic function returns a value for a specified attribute of this type. The following are the available attributes and sample return values.
For more information about using the Fn::GetAtt
intrinsic function, see Fn::GetAtt
.
ComputeEnvironmentArn
-
Returns the compute environment ARN, such as
arn:aws-cn:batch:us-east-1:111122223333:compute-environment/ComputeEnvironmentName
.
Examples
Managed Compute Environment
The following example creates a managed compute environment called
C4OnDemand
that uses C4 On-Demand instances and a custom AMI.
JSON
{ "ComputeEnvironment": { "Type": "AWS::Batch::ComputeEnvironment", "Properties": { "Type": "MANAGED", "arn:aws-cn:iam::111122223333:role/aws-service-role/batch.amazonaws.com/AWSServiceRoleForBatch", "ComputeEnvironmentName": "C4OnDemand", "ComputeResources": { "MaxvCpus": 128, "SecurityGroupIds": [ "sg-abcd1234" ], "Type": "EC2", "Subnets": [ "subnet-aaaaaaaa", "subnet-bbbbbbbb", "subnet-cccccccc" ], "MinvCpus": 0, "ImageId": "ami-a1b2c3d4", "InstanceRole": "ecsInstanceRole", "InstanceTypes": [ "c4.large", "c4.xlarge", "c4.2xlarge", "c4.4xlarge", "c4.8xlarge" ], "Ec2KeyPair": "id_rsa", "Tags": { "Name": "Batch Instance - C4OnDemand" }, "DesiredvCpus": 48 }, "State": "ENABLED" } } }
YAML
ComputeEnvironment: Type: 'AWS::Batch::ComputeEnvironment' Properties: Type: MANAGED ServiceRole: - arn:aws-cn:iam::111122223333:role/aws-service-role/batch.amazonaws.com/AWSServiceRoleForBatch ComputeEnvironmentName: C4OnDemand ComputeResources: MaxvCpus: 128 SecurityGroupIds: - sg-abcd1234 Type: EC2 Subnets: - subnet-aaaaaaaa - subnet-bbbbbbbb - subnet-cccccccc MinvCpus: 0 ImageId: ami-a1b2c3d4 InstanceRole: ecsInstanceRole InstanceTypes: - c4.large - c4.xlarge - c4.2xlarge - c4.4xlarge - c4.8xlarge Ec2KeyPair: id_rsa Tags: Name: Batch Instance - C4OnDemand DesiredvCpus: 48 State: ENABLED
See also
-
Compute Environments in the Amazon Batch User Guide .