# Auto Scaling groups with multiple instance types and purchase options Use multiple instance types and purchase options You can launch and automatically scale a fleet of On-Demand Instances and Spot Instances within a single Auto Scaling group. In addition to receiving discounts for using Spot Instances, you can use Reserved Instances or a Savings Plans to receive discounts on the regular On-Demand Instance pricing. These factors help you optimize your cost savings for EC2 instances and get the desired scale and performance for your application. Spot Instances are spare capacity available at steep discounts compared to the EC2 On-Demand price. Spot Instances are a cost-effective choice if you can be flexible about when your applications run and if your applications can be interrupted. They can be used for various fault-tolerant and flexible applications. Examples include stateless web servers, API endpoints, big data and analytics applications, containerized workloads, CI/CD pipelines, high performance and high throughput computing (HPC/HTC), rendering workloads, and other flexible workloads. For more information, see [Instance purchasing options](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/instance-purchasing-options.html) in the *Amazon EC2 User Guide*. **Topics** + [ # Setup overview for creating a mixed instances group ](mixed-instances-groups-set-up-overview.md) + [ # Allocation strategies for multiple instance types ](allocation-strategies.md) + [ # Create mixed instances group using attribute-based instance type selection ](create-mixed-instances-group-attribute-based-instance-type-selection.md) + [ # Create a mixed instances group by manually choosing instance types ](create-mixed-instances-group-manual-instance-type-selection.md) + [ # Configure an Auto Scaling group to use instance weights ](ec2-auto-scaling-mixed-instances-groups-instance-weighting.md) + [ # Use multiple launch templates ](ec2-auto-scaling-mixed-instances-groups-launch-template-overrides.md) # Setup overview for creating a mixed instances group This topic provides an overview and best practices for creating an Auto Scaling mixed instances group. **Topics** + [ ## Overview ](#mixed-instances-groups-overview) + [ ## Instance type flexibility ](#mixed-instances-group-instance-flexibility) + [ ## Availability Zone flexibility ](#mixed-instances-group-az-flexibility) + [ ## Spot max price ](#mixed-instances-group-spot-max-price) + [ ## Proactive capacity rebalancing ](#use-capacity-rebalancing) + [ ## Scaling behavior ](#mixed-instances-group-scaling-behavior) + [ ## Regional availability of instance types ](#setup-overview-regional-availability-of-instance-types) + [ ## Related resources ](#setup-overview-related-resources) + [ ## Limitations ](#setup-overview-limitations) ## Overview To create a mixed instances group, you have two options: + [Attribute-based instance type selection](create-mixed-instances-group-attribute-based-instance-type-selection.md) – Define your compute requirements to choose your instance types automatically based on their specific instance attributes. + [Manual instance type selection](create-mixed-instances-group-manual-instance-type-selection.md) – Manually choose the instance types that suit your workload. ------ #### [ Manual selection ] The following steps describe how to create a mixed instances group by manually choosing instance types: 1. Choose a launch template that has the parameters to launch an EC2 instance. Parameters in launch templates are optional, but Amazon EC2 Auto Scaling can't launch an instance if the amilong; (AMI) ID is missing from the launch template. 1. Choose the option to override the launch template. 1. Manually choose the instance types that suit your workload. 1. Specify the percentages of On-Demand Instances and Spot Instances to launch. 1. Choose allocation strategies that determine how Amazon EC2 Auto Scaling fulfills your On-Demand and Spot capacities from the possible instance types. 1. Choose the Availability Zones and VPC subnets to launch your instances in. 1. Specify the initial size of the group (the desired capacity) and the minimum and maximum size of the group. Overrides are necessary to override the instance type declared in the launch template and use multiple instances types that are embedded in the Auto Scaling group's own resource definition. For more information about the instance types that are available, see [Instance types](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/instance-types.html) in the *Amazon EC2 User Guide*. You can also configure the following optional parameters for each instance type: + `LaunchTemplateSpecification` – You can assign a different launch template to an instance type as needed. This option is currently not available from the console. For more information, see [Use multiple launch templates](ec2-auto-scaling-mixed-instances-groups-launch-template-overrides.md). + `WeightedCapacity` – You decide how much the instance counts toward the desired capacity relative to the rest of the instances in your group. If you specify a `WeightedCapacity` value for one instance type, you must specify a `WeightedCapacity` value for all of them. By default, each instance counts as one toward your desired capacity. For more information, see [Configure an Auto Scaling group to use instance weights](ec2-auto-scaling-mixed-instances-groups-instance-weighting.md). ------ #### [ Attribute-based selection ] To let Amazon EC2 Auto Scaling choose your instance types automatically based on their specific instance attributes, use the following steps to create a mixed instances group by specifying your compute requirements: 1. Choose a launch template that has the parameters to launch an EC2 instance. Parameters in launch templates are optional, but Amazon EC2 Auto Scaling can't launch an instance if the amilong; (AMI) ID is missing from the launch template. 1. Choose the option to override the launch template. 1. Specify instance attributes that match your compute requirements, such as vCPUs and memory requirements. 1. Specify the percentages of On-Demand Instances and Spot Instances to launch. 1. Choose allocation strategies that determine how Amazon EC2 Auto Scaling fulfills your On-Demand and Spot capacities from the possible instance types. 1. Choose the Availability Zones and VPC subnets to launch your instances in. 1. Specify the initial size of the group (the desired capacity) and the minimum and maximum size of the group. Overrides are necessary to override the instance type declared in the launch template and use a set of instance attributes that describe your compute requirements. For supported attributes, see [InstanceRequirements](https://docs.aws.amazon.com/autoscaling/ec2/APIReference/API_InstanceRequirements.html) in the *Amazon EC2 Auto Scaling API Reference*. Alternatively, you can use a launch template that already has your instance attributes definition. You can also configure the `LaunchTemplateSpecification` parameter within the overrides structure to assign a different launch template to a set of instance requirements as needed. This option is currently not available from the console. For more information, see [LaunchTemplateOverrides](https://docs.aws.amazon.com/autoscaling/ec2/APIReference/API_LaunchTemplateOverrides.html) in the *Amazon EC2 Auto Scaling API Reference*. By default, you set the number of instances as the desired capacity of your Auto Scaling group. Alternatively, you can set the value for desired capacity to the number of vCPUs or the amount of memory. To do so, use the `DesiredCapacityType` property in the `CreateAutoScalingGroup` API operation or the **Desired capacity type** dropdown field in the AWS Management Console. This is a useful alternative to [instance weights](ec2-auto-scaling-mixed-instances-groups-instance-weighting.md). ------ ## Instance type flexibility To enhance availability, deploy your application across multiple instance types. It's a best practice to use multiple instance types to satisfy capacity requirements. This way, Amazon EC2 Auto Scaling can launch another instance type if there is insufficient instance capacity in your chosen Availability Zones. If there is insufficient instance capacity with Spot Instances, Amazon EC2 Auto Scaling keeps trying to launch from other Spot Instance pools. (The pools it uses are determined by your choice of instance types and allocation strategy.) Amazon EC2 Auto Scaling helps you leverage the cost savings of Spot Instances by launching them instead of On-Demand Instances. We recommend being flexible across at least 10 instance types for each workload. When choosing your instance types, don't limit yourself to the most popular new instance types. Choosing earlier generation instance types tends to result in fewer Spot interruptions because they are less in demand from On-Demand customers. ## Availability Zone flexibility We strongly recommend that you span your Auto Scaling group across multiple Availability Zones. With multiple Availability Zones, you can design applications that automatically fail over between zones for greater resiliency. As an added benefit, you can access a deeper Amazon EC2 capacity pool when compared to groups in a single Availability Zone. Because capacity fluctuates independently for each instance type in each Availability Zone, you can often get more compute capacity with flexibility for both the instance type and the Availability Zone. For more information about using multiple Availability Zones, see [Example: Distribute instances across Availability Zones](auto-scaling-benefits.md#arch-AutoScalingMultiAZ). ## Spot max price When you create your Auto Scaling group using the AWS CLI or an SDK, you can specify the `SpotMaxPrice` parameter. The `SpotMaxPrice` parameter determines the maximum price that you're willing to pay for a Spot Instance hour. When you specify the `WeightedCapacity` parameter in your overrides (or `"DesiredCapacityType": "vcpu"` or `"DesiredCapacityType": "memory-mib"` at the group level), the maximum price represents the maximum unit price, not the maximum price for a whole instance. We strongly recommend that you do not specify a maximum price. Your application might not run if you do not receive any Spot Instances, such as when your maximum price is too low. If you don't specify a maximum price, the default maximum price is the On-Demand price. You pay only the Spot price for the Spot Instances that you launch. You still receive the steep discounts provided by Spot Instances. These discounts are possible because of the stable Spot pricing that's available with the [Spot pricing model](https://aws.amazon.com/blogs/compute/new-amazon-ec2-spot-pricing/). For more information, see [Pricing and savings](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/using-spot-instances.html#spot-pricing) in the *Amazon EC2 User Guide*. ## Proactive capacity rebalancing If your use case allows, we recommend Capacity Rebalancing. Capacity Rebalancing helps you maintain your workload availability by proactively replacing Spot Instances at risk of interruption. When Capacity Rebalancing is enabled, Amazon EC2 Auto Scaling attempts to proactively replace Spot Instances that have received an EC2 instance rebalance recommendation. This provides an opportunity to rebalance your workload to new Spot Instances that are not at an elevated risk of interruption. For more information, see [Capacity Rebalancing in Auto Scaling to replace at-risk Spot Instances](ec2-auto-scaling-capacity-rebalancing.md). ## Scaling behavior When you create a mixed instances group, it uses On-Demand Instances by default. To use Spot Instances, you must modify the percentage of the group to be launched as On-Demand Instances. You can specify any number from 0 to 100 for the On-Demand percentage. Optionally, you can also designate a base number of On-Demand Instances to start with. If you do so, Amazon EC2 Auto Scaling waits to launch Spot Instances until after it launches the base capacity of On-Demand Instances when the group scales out. Anything beyond the base capacity uses the On-Demand percentage to determine how many On-Demand Instances and Spot Instances to launch. Amazon EC2 Auto Scaling converts the percentage to the equivalent number of instances. If the result creates a fractional number, it rounds up to the next integer in favor of On-Demand Instances. The following table demonstrates the behavior of the Auto Scaling group as it increases and decreases in size. **Example: Scaling behavior** | Purchase options | Group size and number of running instances across purchase options | | --- |--- | | | **10** | **20** | **30** | **40** | | --- |--- |--- |--- |--- | | **Example 1**: base of 10, 50/50% On-Demand/Spot | | | | | | On-Demand Instances (base amount) | 10 | 10 | 10 | 10 | | On-Demand Instances | 0 | 5 | 10 | 15 | | Spot Instances | 0 | 5 | 10 | 15 | | **Example 2**: base of 0, 0/100% On-Demand/Spot | | | | | | On-Demand Instances (base amount) | 0 | 0 | 0 | 0 | | On-Demand Instances | 0 | 0 | 0 | 0 | | Spot Instances | 10 | 20 | 30 | 40 | | **Example 3**: base of 0, 60/40% On-Demand/Spot | | | | | | On-Demand Instances (base amount) | 0 | 0 | 0 | 0 | | On-Demand Instances | 6 | 12 | 18 | 24 | | Spot Instances | 4 | 8 | 12 | 16 | | **Example 4**: base of 0, 100/0% On-Demand/Spot | | | | | | On-Demand Instances (base amount) | 0 | 0 | 0 | 0 | | On-Demand Instances | 10 | 20 | 30 | 40 | | Spot Instances | 0 | 0 | 0 | 0 | | **Example 5**: base of 12, 0/100% On-Demand/Spot | | | | | | On-Demand Instances (base amount) | 10 | 12 | 12 | 12 | | On-Demand Instances | 0 | 0 | 0 | 0 | | Spot Instances | 0 | 8 | 18 | 28 | When the size of the group *increases*, Amazon EC2 Auto Scaling attempts to balance your capacity evenly across your specified Availability Zones. Then, it launches instance types according to the specified allocation strategy. When the size of the group *decreases*, Amazon EC2 Auto Scaling first identifies which of the two types (Spot or On-Demand) should be terminated. Then, it tries to terminate instances in a balanced way across your specified Availability Zones. It also favors terminating instances in a way that aligns closer to your allocation strategies. For information about termination policies, see [Configure termination policies for Amazon EC2 Auto Scaling](ec2-auto-scaling-termination-policies.md). ## Regional availability of instance types The availability of EC2 instance types varies depending on your AWS Region. For example, the newest generation instance types might not yet be available in a given Region. Due to the variances in instance availability across Regions, you might encounter issues when making programmatic requests if multiple instance types in your overrides are not available in your Region. Using multiple instance types that are not available in your Region might cause the request to fail entirely. To solve the issue, retry the request with different instance types, making sure that each instance type is available in the Region. To search for instance types offered by location, use the [describe-instance-type-offerings](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/ec2/describe-instance-type-offerings.html) command. For more information, see [Finding an Amazon EC2 instance type](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/instance-discovery.html) in the *Amazon EC2 User Guide*. ## Related resources For more best practices for Spot Instances, see [Best practices for EC2 Spot](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/spot-best-practices.html) in the *Amazon EC2 User Guide*. ## Limitations After you add overrides to an Auto Scaling group using a [mixed instances policy](https://docs.aws.amazon.com/autoscaling/ec2/APIReference/API_MixedInstancesPolicy.html), you can update the overrides with the `UpdateAutoScalingGroup` API call but not delete them. To completely remove the overrides, you must first switch the Auto Scaling group to use a launch template or launch configuration instead of a mixed instances policy. Then, you can add a mixed instances policy again without any overrides. # Allocation strategies for multiple instance types When you use multiple instance types, you manage how Amazon EC2 Auto Scaling fulfills your On-Demand and Spot capacities from the possible instance types. To do this, you specify allocation strategies. To review the best practices for a mixed instances group, see [Setup overview for creating a mixed instances group](mixed-instances-groups-set-up-overview.md). **Topics** + [ ## Spot Instances ](#spot-allocation-strategy) + [ ## On-Demand Instances ](#on-demand-allocation-strategy) + [ ## How the allocation strategies work with weights ](#lowest-price-allocation-strategy) ## Spot Instances Amazon EC2 Auto Scaling provides the following allocation strategies for Spot Instances: `price-capacity-optimized` (recommended) The price and capacity optimized allocation strategy looks at both price and capacity to select the Spot Instance pools that are the least likely to be interrupted and have the lowest possible price. We recommend this strategy when you're getting started. For more information, see [Introducing the price-capacity-optimized allocation strategy for EC2 Spot Instances](https://aws.amazon.com/blogs/compute/introducing-price-capacity-optimized-allocation-strategy-for-ec2-spot-instances/) in the AWS blog. `capacity-optimized` Amazon EC2 Auto Scaling requests your Spot Instance from the pool with optimal capacity for the number of instances that are launching. With Spot Instances, the pricing changes slowly over time based on long-term trends in supply and demand. However, capacity fluctuates in real time. The `capacity-optimized` strategy automatically launches Spot Instances into the most available pools by looking at real-time capacity data and predicting which are the most available. This helps to minimize possible disruptions for workloads that might have a higher cost of interruption associated with restarting work and checkpointing. To give certain instance types a higher chance of launching first, use `capacity-optimized-prioritized`. `capacity-optimized-prioritized` You set the order of instance types for the launch template overrides from highest to lowest priority (from first to last in the list). Amazon EC2 Auto Scaling honors the instance type priorities on a best-effort basis but optimizes for capacity first. This is a good option for workloads where the possibility of disruption must be minimized, but the preference for certain instance types matters, too. If the On-Demand allocation strategy is set to `prioritized`, the same priority is applied when fulfilling On-Demand capacity. `lowest-price` (not recommended) We don't recommend the `lowest-price` strategy because it has the highest risk of interruption for your Spot Instances. Amazon EC2 Auto Scaling requests your Spot Instances using the lowest priced pools within an Availability Zone, across the N number of Spot pools that you specify for the **Lowest priced pools** setting. For example, if you specify four instance types and four Availability Zones, your Auto Scaling group can access up to 16 Spot pools. (Four in each Availability Zone.) If you specify two Spot pools (N=2) for the allocation strategy, your Auto Scaling group can draw on the two lowest priced pools per Availability Zone to fulfill your Spot capacity. The `lowest-price` strategy is only available when using the AWS CLI. Amazon EC2 Auto Scaling makes an effort to draw Spot Instances from the N number of pools that you specify. However, if a pool runs out of Spot capacity before fulfilling your desired capacity, Amazon EC2 Auto Scaling continues to fulfill your request by drawing from the next lowest priced pool. To meet your desired capacity, you might receive Spot Instances from more pools than your specified N number. Likewise, if most of the pools have no Spot capacity, you might receive your full desired capacity from fewer pools than your specified N number. **Note** If you configure your Spot Instances to launch with [AMD SEV-SNP](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/sev-snp.html) turned on, you are charged an additional hourly usage fee that is equivalent to 10% of the [On-Demand hourly rate](https://aws.amazon.com/ec2/pricing/on-demand/) of the selected instance type. If the allocation strategy uses price as an input, Amazon EC2 Auto Scaling does not include this additional fee; only the Spot price is used. ## On-Demand Instances Amazon EC2 Auto Scaling provides the following allocation strategies that can be used for On-Demand Instances: `lowest-price` Amazon EC2 Auto Scaling automatically deploys the lowest priced instance type in each Availability Zone based on the current On-Demand price. To meet your desired capacity, you might receive On-Demand Instances of more than one instance type in each Availability Zone. This depends on how much capacity you request. `prioritized` When fulfilling On-Demand capacity, Amazon EC2 Auto Scaling determines which instance type to use first based on the order of instance types in the list of launch template overrides. For example, let's say that you specify three launch template overrides in the following order: `c5.large`, `c4.large`, and `c3.large`. When your On-Demand Instances launch, the Auto Scaling group fulfills On-Demand capacity in the following order: `c5.large`, `c4.large`, and then `c3.large`. Consider the following when managing the priority order of your On-Demand Instances: + You can pay for usage upfront to get significant discounts for On-Demand Instances by using Savings Plans or Reserved Instances. For more information, see the [Amazon EC2 pricing](https://aws.amazon.com/ec2/pricing/) page. + With Reserved Instances, your discounted rate of the regular On-Demand Instance pricing applies if Amazon EC2 Auto Scaling launches matching instance types. Therefore, if you have unused Reserved Instances for `c4.large`, you can set the instance type priority to give the highest priority for your Reserved Instances to a `c4.large` instance type. When a `c4.large` instance launches, you receive the Reserved Instance pricing. + With Savings Plans, your discounted rate of the regular On-Demand Instance pricing applies when using Amazon EC2 Instance Savings Plans or Compute Savings Plans. With Savings Plans, you have more flexibility when prioritizing your instance types. As long as you use instance types that are covered by your Savings Plans, you can set them in any priority order. You can also occasionally change the entire order of your instance types, while still receiving the Savings Plans discounted rate. For more information about Savings Plans, see the [Savings Plans User Guide](https://docs.aws.amazon.com/savingsplans/latest/userguide/). ## How the allocation strategies work with weights When you specify the `WeightedCapacity` parameter in your overrides (or `"DesiredCapacityType": "vcpu"` or `"DesiredCapacityType": "memory-mib"` at the group level), the allocation strategies work exactly like they do for other Auto Scaling groups. Suppose you have an Auto Scaling group with several instance types that have varying amounts of vCPUs. You use `lowest-price` for your Spot and On-Demand allocation strategies. If you choose to assign weights based on the vCPU count of each instance type, Amazon EC2 Auto Scaling launches whichever instance types have the lowest price per your assigned weight values (for example, per vCPU) at the time of fulfillment. If it's a Spot Instance, then this means the lowest Spot price per vCPU. If it's an On-Demand Instance, then this means the lowest On-Demand price per vCPU. For more information, see [Configure an Auto Scaling group to use instance weights](ec2-auto-scaling-mixed-instances-groups-instance-weighting.md). # Create mixed instances group using attribute-based instance type selection Create a group using attribute-based instance type selection Instead of manually choosing instance types for your mixed instances group, you can specify a set of instance attributes that describe your compute requirements. As Amazon EC2 Auto Scaling launches instances, any instance types used by the Auto Scaling group must match your required instance attributes. This is known as *attribute-based instance type selection*. This approach is ideal for workloads and frameworks that can be flexible about which instance types they use, such as containers, big data, and CI/CD. The following are benefits of attribute-based instance type selection: + **Optimal flexibility for Spot Instances** – Amazon EC2 Auto Scaling can select from a wide range of instance types for launching Spot Instances. This meets the Spot best practice of being flexible about instance types, which gives the Amazon EC2 Spot service a better chance of finding and allocating your required amount of compute capacity. + **Easily use the right instance types** – With so many instance types available, finding the right instance types for your workload can be time consuming. When you specify instance attributes, the instance types will automatically have the required attributes for your workload. + **Automatic use of new instance types** – Your Auto Scaling groups can use newer generation instance types as they're released. Newer generation instance types are automatically used when they match your requirements and align with the allocation strategies you choose for your Auto Scaling group. **Topics** + [ ## How attribute-based instance type selection works ](#how-attribute-based-instance-type-selection-works) + [ ## Price protection ](#understand-price-protection) + [ ## Performance protection ](#understand-performance-protection) + [ ## Prerequisites ](#attribute-based-instance-type-selection-prerequisites) + [ ## Create a mixed instances group with attribute-based instance type selection (console) ](#attribute-based-instance-type-selection-console) + [ ## Create a mixed instances group with attribute-based instance type selection (AWS CLI) ](#attribute-based-instance-type-selection-aws-cli) + [ ## Example configuration ](#attribute-based-instance-type-selection-example-configurations) + [ ## Preview your instance types ](#attribute-based-instance-type-selection-preview) + [ ## Related resources ](#attribute-based-instance-type-selection-related-resources) ## How attribute-based instance type selection works With attribute-based instance type selection, instead of providing a list of specific instance types, you provide a list of instance attributes that your instances require, such as: + **vCPU count** – The minimum and maximum number of vCPUs per instance. + **Memory** – The minimum and maximum GiBs of memory per instance. + **Local storage** – Whether to use EBS or instance store volumes for local storage. + **Burstable performance** – Whether to use the T instance family, including T4g, T3a, T3, and T2 types. There are many options available for defining your instance requirements. For a description of each option and the default values, see [InstanceRequirements](https://docs.aws.amazon.com/autoscaling/ec2/APIReference/API_InstanceRequirements.html) in the *Amazon EC2 Auto Scaling API Reference*. When your Auto Scaling group needs to launch an instance, it will search for instance types that match your specified attributes and are available in that Availability Zone. The allocation strategy then determines which of the matching instance types to launch. By default, attribute-based instance type selection has a price protection feature enabled to prevent your Auto Scaling group from launching instance types that exceed your budget thresholds. By default, you use the number of instances as the unit of measurement when setting the desired capacity of your Auto Scaling group, meaning each instance counts as one unit. Alternatively, you can set the value for desired capacity to the number of vCPUs or the amount of memory. To do so, use the **Desired capacity type** dropdown field in the AWS Management Console or the `DesiredCapacityType` property in the `CreateAutoScalingGroup` or `UpdateAutoScalingGroup` API operation. Amazon EC2 Auto Scaling then launches the number of instances required to meet the desired vCPU or memory capacity. For example, if you use vCPUs as the desired capacity type and use instances with 2 vCPUs each, a desired capacity of 10 vCPUs would launch 5 instances. This is a useful alternative to [instance weights](ec2-auto-scaling-mixed-instances-groups-instance-weighting.md). ## Price protection With price protection, you can specify the maximum price you are willing to pay for EC2 instances launched by your Auto Scaling group. Price protection is a feature that prevents your Auto Scaling group from using instance types that you would consider too expensive even if they happen to fit the attributes that you specified. Price protection is enabled by default and has separate price thresholds for On-Demand Instances and Spot Instances. When Amazon EC2 Auto Scaling needs to launch new instances, any instance types priced above the relevant threshold are not launched. **Topics** + [ ### On-Demand price protection ](#on-demand-price-price-protection) + [ ### Spot price protection ](#spot-price-price-protection) + [ ### Customize price protection ](#customize-price-price-protection) ### On-Demand price protection For On-Demand Instances, you define the maximum On-Demand price you're willing to pay as a percentage higher than an identified On-Demand price. The identified On-Demand price is the price of the lowest priced current generation C, M, or R instance type with your specified attributes. If an On-Demand price protection value is not explicitly defined, a default maximum On-Demand price of 20 percent higher than the identified On-Demand price will be used. ### Spot price protection By default, Amazon EC2 Auto Scaling will automatically apply optimal Spot Instance price protection to consistently select from a wide range of instance types. You can also manually set the price protection yourself. However, letting Amazon EC2 Auto Scaling do it for you can improve the likelihood that your Spot capacity is fulfilled. You can manually specify the price protection using one of the following options. If you manually set the price protection, we recommend using the first option. + **A percentage of an identified *On-Demand price*** – The identified On-Demand price is the price of the lowest priced current generation C, M, or R instance type with your specified attributes. + **A percentage higher than an identified *Spot price*** – The identified Spot price is the price of the lowest priced current generation C, M, or R instance type with your specified attributes. We do not recommend using this option because Spot prices can fluctuate, and therefore your price protection threshold might also fluctuate. ### Customize price protection You can customize the price protection thresholds in the Amazon EC2 Auto Scaling console or using the AWS CLI or SDKs. + In the console, use the **On-Demand price protection** and **Spot price protection** settings in **Additional instance attributes**. + In the [InstanceRequirements](https://docs.aws.amazon.com/autoscaling/ec2/APIReference/API_InstanceRequirements.html) structure, to specify the On-Demand Instance price protection threshold, use the `OnDemandMaxPricePercentageOverLowestPrice` property. To specify the Spot Instance price protection threshold, use either the `MaxSpotPriceAsPercentageOfOptimalOnDemandPrice` or the `SpotMaxPricePercentageOverLowestPrice` property. If you set **Desired capacity type** (`DesiredCapacityType`) to **vCPUs** or **Memory GiB**, the price protection applies based on the per vCPU or per memory price instead of the per instance price. You can also turn off price protection. To indicate no price protection threshold, specify a high percentage value, such as `999999`. **Note** If no current generation C, M, or R instance types match your specified attributes, price protection is still applicable. When no match is found, the identified price is from the lowest priced current generation instance types, or failing that, the lowest priced previous generation instance types, that match your attributes. ## Performance protection *Performance protection* is a feature that ensures your Auto Scaling group uses instance types that are similar to or exceed a specified performance baseline. To use performance protection, you specify an instance family as a baseline reference. The capabilities of the specified instance family establish the lowest acceptable level of performance. When Auto Scaling selects instance types, it considers your specified attributes and the performance baseline. Instance types that fall below the performance baseline are automatically excluded from selection, even if they match your other specified attributes. This ensures that all selected instance types offer performance similar to or better than the baseline established by the specified instance family. Auto Scaling uses this baseline to guide instance type selection, but there is no guarantee that the selected instance types will always exceed the baseline for every application. Currently, this feature only supports CPU performance as a baseline performance factor. The CPU performance of the specified instance family serves as the performance baseline, ensuring that selected instance types are similar to or exceed this baseline. Instance families with the same CPU processors lead to the same filtering results, even if their network or disk performance differs. For example, specifying either `c6in` or `c6i` as the baseline reference would produce identical performance-based filtering results because both instance families use the same CPU processor. **Unsupported instance families** The following instance families are not supported for performance protection: + `c1` + `g3` \$1 `g3s` + `hpc7g` + `m1` \$1 `m2` + `mac1` \$1 `mac2` \$1 `mac2-m1ultra` \$1 `mac2-m2` \$1 `mac2-m2pro` + `p3dn` \$1 `p4d` \$1 `p5` + `t1` + `u-12tb1` \$1 `u-18tb1` \$1 `u-24tb1` \$1 `u-3tb1` \$1 `u-6tb1` \$1 `u-9tb1` \$1 `u7i-12tb` \$1 `u7in-16tb` \$1 `u7in-24tb` \$1 `u7in-32tb` If you enable performance protection by specifying a supported instance family, the returned instance types will exclude the above unsupported instance families. **Example: Set a CPU performance baseline** In the following example, the instance requirement is to launch with instance types that have CPU cores that are as performant as the `c6i` instance family. This will filter out instance types with less performant CPU processors, even if they meet your other specified instance requirements such as the number of vCPUs. For example, if your specified instance attributes include 4 vCPUs and 16 GB of memory, an instance type with these attributes but with lower CPU performance than `c6i` will be excluded from selection. ``` "BaselinePerformanceFactors": { "Cpu": { "References": [ { "InstanceFamily": "c6i" } ] } ``` **Considerations** Consider the following when using performance protection: + You can specify either instance types or instance attributes, but not both at the same time. + You can specify a maximum of four `InstanceRequirements` structures in a request configuration. ## Prerequisites + Create a launch template. For more information, see [Create a launch template for an Auto Scaling group](create-launch-template.md). + Verify that the launch template doesn't already request Spot Instances. ## Create a mixed instances group with attribute-based instance type selection (console) Use the following procedure to create a mixed instances group by using attribute-based instance type selection. To help you move through the steps efficiently, some optional sections are skipped. For most general purpose workloads, it's enough to specify the number of vCPUs and memory that you need. For advanced use cases, you can specify attributes like storage type, network interfaces, CPU manufacturer, and accelerator type. To review the best practices for a mixed instances group, see [Setup overview for creating a mixed instances group](mixed-instances-groups-set-up-overview.md). **To create a mixed instances group** 1. Open the Amazon EC2 console at [https://console.aws.amazon.com/ec2/](https://console.aws.amazon.com/ec2/), and choose **Auto Scaling Groups** from the navigation pane. 1. On the navigation bar at the top of the screen, choose the same AWS Region that you used when you created the launch template. 1. Choose **Create an Auto Scaling group**. 1. On the **Choose launch template or configuration** page, for **Auto Scaling group name**, enter a name for your Auto Scaling group. 1. To choose your launch template, do the following: 1. For **Launch template**, choose an existing launch template. 1. For **Launch template version**, choose whether the Auto Scaling group uses the default, the latest, or a specific version of the launch template when scaling out. 1. Verify that your launch template supports all of the options that you are planning to use, and then choose **Next**. 1. On the **Choose instance launch options** page, do the following: 1. For **Instance type requirements**, choose **Override launch template**. **Note** If you chose a launch template that already contains a set of instance attributes, such as vCPUs and memory, then the instance attributes are displayed. These attributes are added to the Auto Scaling group properties, where you can update them from the Amazon EC2 Auto Scaling console at any time. 1. Under **Specify instance attributes**, start by entering your vCPUs and memory requirements. + For **vCPUs**, enter the desired minimum and maximum number of vCPUs. To specify no limit, select **No minimum**, **No maximum**, or both. + For **Memory (GiB)**, enter the desired minimum and maximum amount of memory. To specify no limit, select **No minimum**, **No maximum**, or both. 1. (Optional) For **Additional instance attributes**, you can optionally specify one or more attributes to express your compute requirements in more detail. Each additional attribute adds further constraints to your request. 1. Expand **Preview matching instance types** to view the instance types that have your specified attributes. 1. Under **Instance purchase options**, for **Instances distribution**, specify the percentages of the group to launch as On-Demand Instances and as Spot Instances. If your application is stateless, fault tolerant, and can handle an instance being interrupted, you can specify a higher percentage of Spot Instances. 1. (Optional) When you specify a percentage for Spot Instances, select **Include On-Demand base capacity** and then specify the minimum amount of the Auto Scaling group's initial capacity that must be fulfilled by On-Demand Instances. Anything beyond the base capacity uses the **Instances distribution** settings to determine how many On-Demand Instances and Spot Instances to launch. 1. Under **Allocation strategies**, **Lowest price** is automatically selected for the **On-Demand allocation strategy** and cannot be changed. 1. For **Spot allocation strategy**, choose an allocation strategy. **Price capacity optimized** is selected by default. 1. For **Capacity Rebalancing**, choose whether to enable or disable Capacity Rebalancing. Use Capacity Rebalancing to automatically respond when your Spot Instances approach termination from a Spot interruption. For more information, see [Capacity Rebalancing in Auto Scaling to replace at-risk Spot Instances](ec2-auto-scaling-capacity-rebalancing.md). 1. Under **Network**, for **VPC**, choose a VPC. The Auto Scaling group must be created in the same VPC as the security group you specified in your launch template. 1. For **Availability Zones and subnets**, choose one or more subnets in the specified VPC. Use subnets in multiple Availability Zones for high availability. For more information, see [Considerations when choosing VPC subnets](asg-in-vpc.md#as-vpc-considerations). 1. Choose **Next**, **Next**. 1. For the **Configure group size and scaling policies** step, do the following: 1. To measure your desired capacity in units other than instances, choose the appropriate option for **Group size**, **Desired capacity type**. **Units**, **vCPUs**, and **Memory GiB** are supported. By default, Amazon EC2 Auto Scaling specifies **Units**, which translates into number of instances. 1. For **Desired capacity**, the initial size of your Auto Scaling group. 1. In the **Scaling** section, under **Scaling limits**, if your new value for **Desired capacity** is greater than **Min desired capacity** and **Max desired capacity**, the **Max desired capacity** is automatically increased to the new desired capacity value. You can change these limits as needed. For more information, see [Set scaling limits for your Auto Scaling group](asg-capacity-limits.md). 1. Choose **Skip to review**. 1. On the **Review** page, choose **Create Auto Scaling group**. ## Create a mixed instances group with attribute-based instance type selection (AWS CLI) **To create a mixed instances group using the command line** Use one of the following commands: + [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) (AWS CLI) + [New-ASAutoScalingGroup](https://docs.aws.amazon.com/powershell/latest/reference/items/New-ASAutoScalingGroup.html) (AWS Tools for Windows PowerShell) ## Example configuration To create an Auto Scaling group with attribute-based instance type selection by using the AWS CLI, use the following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command. The following instance attributes are specified: + `VCpuCount` – The instance types must have a minimum of four vCPUs and a maximum of eight vCPUs. + `MemoryMiB` – The instance types must have a minimum of 16,384 MiB of memory. + `CpuManufacturers` – The instance types must have an Intel manufactured CPU. ### JSON ``` aws autoscaling create-auto-scaling-group --cli-input-json file://~/config.json ``` The following is an example `config.json` file. ``` { "AutoScalingGroupName": "my-asg", "DesiredCapacityType": "units", "MixedInstancesPolicy": { "LaunchTemplate": { "LaunchTemplateSpecification": { "LaunchTemplateName": "my-launch-template", "Version": "$Default" }, "Overrides": [{ "InstanceRequirements": { "VCpuCount": {"Min": 4, "Max": 8}, "MemoryMiB": {"Min": 16384}, "CpuManufacturers": ["intel"] } }] }, "InstancesDistribution": { "OnDemandPercentageAboveBaseCapacity": 50, "SpotAllocationStrategy": "price-capacity-optimized" } }, "MinSize": 0, "MaxSize": 100, "DesiredCapacity": 4, "DesiredCapacityType": "units", "VPCZoneIdentifier": "subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782" } ``` To set the value for desired capacity as the number of vCPUs or the amount of memory, specify `"DesiredCapacityType": "vcpu"` or `"DesiredCapacityType": "memory-mib"` in the file. The default desired capacity type is `units`, which sets the value for desired capacity as the number of instances. ### YAML Alternatively, you can use the following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command to create the Auto Scaling group. This references a YAML file as the sole parameter for your Auto Scaling group. ``` aws autoscaling create-auto-scaling-group --cli-input-yaml file://~/config.yaml ``` The following is an example `config.yaml` file. ``` --- AutoScalingGroupName: my-asg DesiredCapacityType: units MixedInstancesPolicy: LaunchTemplate: LaunchTemplateSpecification: LaunchTemplateName: my-launch-template Version: $Default Overrides: - InstanceRequirements: VCpuCount: Min: 2 Max: 4 MemoryMiB: Min: 2048 CpuManufacturers: - intel InstancesDistribution: OnDemandPercentageAboveBaseCapacity: 50 SpotAllocationStrategy: price-capacity-optimized MinSize: 0 MaxSize: 100 DesiredCapacity: 4 DesiredCapacityType: units VPCZoneIdentifier: subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782 ``` To set the value for desired capacity as the number of vCPUs or the amount of memory, specify `DesiredCapacityType: vcpu` or `DesiredCapacityType: memory-mib` in the file. The default desired capacity type is `units`, which sets the value for desired capacity as the number of instances. For an example of how to use multiple launch templates with attribute-based instance type selection, see [Use multiple launch templates](ec2-auto-scaling-mixed-instances-groups-launch-template-overrides.md). ## Preview your instance types You can preview the instance types that match your compute requirements without launching them and adjust your requirements if necessary. When creating your Auto Scaling group in the Amazon EC2 Auto Scaling console, a preview of the instance types appears in the **Preview matching instance types** section on the **Choose instance launch options** page. Alternatively, you can preview the instance types by making an Amazon EC2 [GetInstanceTypesFromInstanceRequirements](https://docs.aws.amazon.com/AWSEC2/latest/APIReference/API_GetInstanceTypesFromInstanceRequirements.html) API call using the AWS CLI or an SDK. Pass the `InstanceRequirements` parameters in the request in the exact format that you would use to create or update an Auto Scaling group. For more information, see [Preview instance types with specified attributes](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ec2-fleet-attribute-based-instance-type-selection.html#ec2fleet-get-instance-types-from-instance-requirements) in the *Amazon EC2 User Guide*. ## Related resources To learn more about attribute-based instance type selection, see [Attribute-Based Instance Type Selection for EC2 Auto Scaling and EC2 Fleet](https://aws.amazon.com/blogs/aws/new-attribute-based-instance-type-selection-for-ec2-auto-scaling-and-ec2-fleet/) on the AWS Blog. You can declare attribute-based instance type selection when you create an Auto Scaling group using CloudFormation. For more information, see the example snippet in the [Auto scaling template snippets](https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/quickref-autoscaling.html#scenario-mixed-instances-group-template-examples) section of the *CloudFormation User Guide*. # Create a mixed instances group by manually choosing instance types Create a group using manual instance type selection This topic shows you how to launch multiple instance types in a single Auto Scaling group by manually choosing your instance types. If you prefer to use instance attributes as criteria for selecting instance types, see [Create mixed instances group using attribute-based instance type selection](create-mixed-instances-group-attribute-based-instance-type-selection.md). **Topics** + [ ## Prerequisites ](#manual-instance-type-selection-prerequisites) + [ ## Create a mixed instances group (console) ](#manual-instance-type-selection-console) + [ ## Create a mixed instances group (AWS CLI) ](#manual-instance-type-selection-aws-cli) + [ ## Example configurations ](#manual-instance-type-selection-example-configurations) ## Prerequisites + Create a launch template. For more information, see [Create a launch template for an Auto Scaling group](create-launch-template.md). + Verify that the launch template doesn't already request Spot Instances. ## Create a mixed instances group (console) Use the following procedure to create a mixed instances group by manually choosing which instance types your group can launch. To help you move through the steps efficiently, some optional sections are skipped. To review the best practices for a mixed instances group, see [Setup overview for creating a mixed instances group](mixed-instances-groups-set-up-overview.md). **To create a mixed instances group** 1. Open the Amazon EC2 console at [https://console.aws.amazon.com/ec2/](https://console.aws.amazon.com/ec2/), and choose **Auto Scaling Groups** from the navigation pane. 1. On the navigation bar at the top of the screen, choose the same AWS Region that you used when you created the launch template. 1. Choose **Create an Auto Scaling group**. 1. On the **Choose launch template or configuration** page, for **Auto Scaling group name**, enter a name for your Auto Scaling group. 1. To choose your launch template, do the following: 1. For **Launch template**, choose an existing launch template. 1. For **Launch template version**, choose whether the Auto Scaling group uses the default, the latest, or a specific version of the launch template when scaling out. 1. Verify that your launch template supports all of the options that you plan to use, and then choose **Next**. 1. On the **Choose instance launch options** page, do the following: 1. For **Instance type requirements**, choose **Override launch template**, and then choose **Manually add instance types**. 1. Choose your instance types. You can use our recommendations as a starting point. **Family and generation flexible** is selected by default. + To change the order of the instance types, use the arrows. If you choose an allocation strategy that supports prioritization, the instance type order sets their launch priority. + To remove an instance type, choose **X**. + (Optional) For the boxes in the **Weight** column, assign each instance type a relative weight. To do so, enter the number of units that an instance of that type counts toward the desired capacity of the group. Doing so might be useful if the instance types offer different vCPU, memory, storage, or network bandwidth capabilities. For more information, see [Configure an Auto Scaling group to use instance weights](ec2-auto-scaling-mixed-instances-groups-instance-weighting.md). Note that if you choose to use **Size flexible** recommendations, then all instance types that are part of this section automatically have a weight value. If you don't want to specify any weights, clear the boxes in the **Weight** column for all instance types. 1. Under **Instance purchase options**, for **Instances distribution**, specify the percentages of the group to be launched as On-Demand Instances and Spot Instances respectively. If your application is stateless, fault tolerant, and can handle an instance being interrupted, you can specify a higher percentage of Spot Instances. 1. (Optional) When you specify a percentage for Spot Instances, select **Include On-Demand base capacity** and then specify the minimum amount of the Auto Scaling group's initial capacity that must be fulfilled by On-Demand Instances. Anything beyond the base capacity uses the **Instances distribution** settings to determine how many On-Demand Instances and Spot Instances to launch. 1. Under **Allocation strategies**, for **On-Demand allocation strategy**, choose an allocation strategy. When you manually choose your instance types, **Prioritized** is selected by default. 1. For **Spot allocation strategy**, choose an allocation strategy. **Price capacity optimized** is selected by default. If you choose **Capacity optimized**, you can optionally check the **Prioritize instance types** box to let Amazon EC2 Auto Scaling choose which instance type to launch first based on the order your instance types are listed in. 1. For **Capacity Rebalancing**, choose whether to enable or disable Capacity Rebalancing. Use Capacity Rebalancing to automatically respond when your Spot Instances approach termination from a Spot interruption. For more information, see [Capacity Rebalancing in Auto Scaling to replace at-risk Spot Instances](ec2-auto-scaling-capacity-rebalancing.md). 1. Under **Network**, for **VPC**, choose a VPC. The Auto Scaling group must be created in the same VPC as the security group you specified in your launch template. 1. For **Availability Zones and subnets**, choose one or more subnets in the specified VPC. Use subnets in multiple Availability Zones for high availability. For more information, see [Considerations when choosing VPC subnets](asg-in-vpc.md#as-vpc-considerations). 1. Choose **Next**, **Next**. 1. For the **Configure group size and scaling policies** step, do the following: 1. Under **Group size**, for **Desired capacity**, enter the initial number of instances to launch. By default, the desired capacity is expressed as the number of instances. If you assigned weights to your instance types, you must convert this value to the same unit of measurement that you used to assign weights, such as the number of vCPUs. 1. In the **Scaling** section, under **Scaling limits**, if your new value for **Desired capacity** is greater than **Min desired capacity** and **Max desired capacity**, the **Max desired capacity** is automatically increased to the new desired capacity value. You can change these limits as needed. For more information, see [Set scaling limits for your Auto Scaling group](asg-capacity-limits.md). 1. Choose **Skip to review**. 1. On the **Review** page, choose **Create Auto Scaling group**. ## Create a mixed instances group (AWS CLI) **To create a mixed instances group using the command line** Use one of the following commands: + [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) (AWS CLI) + [New-ASAutoScalingGroup](https://docs.aws.amazon.com/powershell/latest/reference/items/New-ASAutoScalingGroup.html) (AWS Tools for Windows PowerShell) ## Example configurations The following example configurations show how to create mixed instances groups using different Spot allocation strategies. **Note** These examples show how to use a configuration file formatted in JSON or YAML. If you use AWS CLI version 1, you must specify a JSON-formatted configuration file. If you use AWS CLI version 2, you can specify a configuration file formatted in either YAML or JSON. **Topics** + [ ### Example 1: Launch Spot Instances using the `capacity-optimized` allocation strategy ](#capacity-optimized-aws-cli) + [ ### Example 2: Launch Spot Instances using the `capacity-optimized-prioritized` allocation strategy ](#capacity-optimized-prioritized-aws-cli) + [ ### Example 3: Launch Spot Instances using the `lowest-price` allocation strategy diversified over two pools ](#lowest-price-aws-cli) + [ ### Example 4: Launch Spot Instances using the `price-capacity-optimized` allocation strategy ](#price-capacity-optimized-aws-cli) ### Example 1: Launch Spot Instances using the `capacity-optimized` allocation strategy The following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command creates an Auto Scaling group that specifies the following: + The percentage of the group to launch as On-Demand Instances (`0`) and a base number of On-Demand Instances to start with (`1`). + The instance types to launch in priority order (`c5.large`, `c5a.large`, `m5.large`, `m5a.large`, `c4.large`, `m4.large`, `c3.large`, `m3.large`) . + The subnets in which to launch the instances (`subnet-5ea0c127`, `subnet-6194ea3b`, `subnet-c934b782`). Each corresponds to a different Availability Zone. + The launch template (`my-launch-template`) and the launch template version (`$Default`). When Amazon EC2 Auto Scaling attempts to fulfill your On-Demand capacity, it launches the `c5.large` instance type first. The Spot Instances come from the optimal Spot pool in each Availability Zone based on Spot Instance capacity. #### JSON ``` aws autoscaling create-auto-scaling-group --cli-input-json file://~/config.json ``` The `config.json` file contains the following content. ``` { "AutoScalingGroupName": "my-asg", "MixedInstancesPolicy": { "LaunchTemplate": { "LaunchTemplateSpecification": { "LaunchTemplateName": "my-launch-template", "Version": "$Default" }, "Overrides": [ { "InstanceType": "c5.large" }, { "InstanceType": "c5a.large" }, { "InstanceType": "m5.large" }, { "InstanceType": "m5a.large" }, { "InstanceType": "c4.large" }, { "InstanceType": "m4.large" }, { "InstanceType": "c3.large" }, { "InstanceType": "m3.large" } ] }, "InstancesDistribution": { "OnDemandBaseCapacity": 1, "OnDemandPercentageAboveBaseCapacity": 0, "SpotAllocationStrategy": "capacity-optimized" } }, "MinSize": 1, "MaxSize": 5, "DesiredCapacity": 3, "VPCZoneIdentifier": "subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782" } ``` #### YAML Alternatively, you can use the following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command to create the Auto Scaling group. This references a YAML file as the sole parameter for your Auto Scaling group. ``` aws autoscaling create-auto-scaling-group --cli-input-yaml file://~/config.yaml ``` The `config.yaml` file contains the following content. ``` --- AutoScalingGroupName: my-asg MixedInstancesPolicy: LaunchTemplate: LaunchTemplateSpecification: LaunchTemplateName: my-launch-template Version: $Default Overrides: - InstanceType: c5.large - InstanceType: c5a.large - InstanceType: m5.large - InstanceType: m5a.large - InstanceType: c4.large - InstanceType: m4.large - InstanceType: c3.large - InstanceType: m3.large InstancesDistribution: OnDemandBaseCapacity: 1 OnDemandPercentageAboveBaseCapacity: 0 SpotAllocationStrategy: capacity-optimized MinSize: 1 MaxSize: 5 DesiredCapacity: 3 VPCZoneIdentifier: subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782 ``` ### Example 2: Launch Spot Instances using the `capacity-optimized-prioritized` allocation strategy The following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command creates an Auto Scaling group that specifies the following: + The percentage of the group to launch as On-Demand Instances (`0`) and a base number of On-Demand Instances to start with (`1`). + The instance types to launch in priority order (`c5.large`, `c5a.large`, `m5.large`, `m5a.large`, `c4.large`, `m4.large`, `c3.large`, `m3.large`) . + The subnets in which to launch the instances (`subnet-5ea0c127`, `subnet-6194ea3b`, `subnet-c934b782`). Each corresponds to a different Availability Zone. + The launch template (`my-launch-template`) and the launch template version (`$Latest`). When Amazon EC2 Auto Scaling attempts to fulfill your On-Demand capacity, it launches the `c5.large` instance type first. When Amazon EC2 Auto Scaling attempts to fulfill your Spot capacity, it honors the instance type priorities on a best-effort basis. However, it optimizes for capacity first. #### JSON ``` aws autoscaling create-auto-scaling-group --cli-input-json file://~/config.json ``` The `config.json` file contains the following content. ``` { "AutoScalingGroupName": "my-asg", "MixedInstancesPolicy": { "LaunchTemplate": { "LaunchTemplateSpecification": { "LaunchTemplateName": "my-launch-template", "Version": "$Latest" }, "Overrides": [ { "InstanceType": "c5.large" }, { "InstanceType": "c5a.large" }, { "InstanceType": "m5.large" }, { "InstanceType": "m5a.large" }, { "InstanceType": "c4.large" }, { "InstanceType": "m4.large" }, { "InstanceType": "c3.large" }, { "InstanceType": "m3.large" } ] }, "InstancesDistribution": { "OnDemandBaseCapacity": 1, "OnDemandPercentageAboveBaseCapacity": 0, "SpotAllocationStrategy": "capacity-optimized-prioritized" } }, "MinSize": 1, "MaxSize": 5, "DesiredCapacity": 3, "VPCZoneIdentifier": "subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782" } ``` #### YAML Alternatively, you can use the following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command to create the Auto Scaling group. This references a YAML file as the sole parameter for your Auto Scaling group. ``` aws autoscaling create-auto-scaling-group --cli-input-yaml file://~/config.yaml ``` The `config.yaml` file contains the following content. ``` --- AutoScalingGroupName: my-asg MixedInstancesPolicy: LaunchTemplate: LaunchTemplateSpecification: LaunchTemplateName: my-launch-template Version: $Default Overrides: - InstanceType: c5.large - InstanceType: c5a.large - InstanceType: m5.large - InstanceType: m5a.large - InstanceType: c4.large - InstanceType: m4.large - InstanceType: c3.large - InstanceType: m3.large InstancesDistribution: OnDemandBaseCapacity: 1 OnDemandPercentageAboveBaseCapacity: 0 SpotAllocationStrategy: capacity-optimized-prioritized MinSize: 1 MaxSize: 5 DesiredCapacity: 3 VPCZoneIdentifier: subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782 ``` ### Example 3: Launch Spot Instances using the `lowest-price` allocation strategy diversified over two pools The following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command creates an Auto Scaling group that specifies the following: + The percentage of the group to launch as On-Demand Instances (`50`). (This does not specify a base number of On-Demand Instances to start with.) + The instance types to launch in priority order (`c5.large`, `c5a.large`, `m5.large`, `m5a.large`, `c4.large`, `m4.large`, `c3.large`, `m3.large`). + The subnets in which to launch the instances (`subnet-5ea0c127`, `subnet-6194ea3b`, `subnet-c934b782`). Each corresponds to a different Availability Zone. + The launch template (`my-launch-template`) and the launch template version (`$Latest`). When Amazon EC2 Auto Scaling attempts to fulfill your On-Demand capacity, it launches the `c5.large` instance type first. For your Spot capacity, Amazon EC2 Auto Scaling attempts to launch the Spot Instances evenly across the two lowest priced pools in each Availability Zone. #### JSON ``` aws autoscaling create-auto-scaling-group --cli-input-json file://~/config.json ``` The `config.json` file contains the following content. ``` { "AutoScalingGroupName": "my-asg", "MixedInstancesPolicy": { "LaunchTemplate": { "LaunchTemplateSpecification": { "LaunchTemplateName": "my-launch-template", "Version": "$Latest" }, "Overrides": [ { "InstanceType": "c5.large" }, { "InstanceType": "c5a.large" }, { "InstanceType": "m5.large" }, { "InstanceType": "m5a.large" }, { "InstanceType": "c4.large" }, { "InstanceType": "m4.large" }, { "InstanceType": "c3.large" }, { "InstanceType": "m3.large" } ] }, "InstancesDistribution": { "OnDemandPercentageAboveBaseCapacity": 50, "SpotAllocationStrategy": "lowest-price", "SpotInstancePools": 2 } }, "MinSize": 1, "MaxSize": 5, "DesiredCapacity": 3, "VPCZoneIdentifier": "subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782" } ``` #### YAML Alternatively, you can use the following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command to create the Auto Scaling group. This references a YAML file as the sole parameter for your Auto Scaling group. ``` aws autoscaling create-auto-scaling-group --cli-input-yaml file://~/config.yaml ``` The `config.yaml` file contains the following content. ``` --- AutoScalingGroupName: my-asg MixedInstancesPolicy: LaunchTemplate: LaunchTemplateSpecification: LaunchTemplateName: my-launch-template Version: $Default Overrides: - InstanceType: c5.large - InstanceType: c5a.large - InstanceType: m5.large - InstanceType: m5a.large - InstanceType: c4.large - InstanceType: m4.large - InstanceType: c3.large - InstanceType: m3.large InstancesDistribution: OnDemandPercentageAboveBaseCapacity: 50 SpotAllocationStrategy: lowest-price SpotInstancePools: 2 MinSize: 1 MaxSize: 5 DesiredCapacity: 3 VPCZoneIdentifier: subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782 ``` ### Example 4: Launch Spot Instances using the `price-capacity-optimized` allocation strategy The following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command creates an Auto Scaling group that specifies the following: + The percentage of the group to launch as On-Demand Instances (`30`). (This does not specify a base number of On-Demand Instances to start with.) + The instance types to launch in priority order (`c5.large`, `c5a.large`, `m5.large`, `m5a.large`, `c4.large`, `m4.large`, `c3.large`, `m3.large`). + The subnets in which to launch the instances (`subnet-5ea0c127`, `subnet-6194ea3b`, `subnet-c934b782`). Each corresponds to a different Availability Zone. + The launch template (`my-launch-template`) and the launch template version (`$Latest`). When Amazon EC2 Auto Scaling attempts to fulfill your On-Demand capacity, it launches the `c5.large` instance type first. For your Spot capacity, Amazon EC2 Auto Scaling attempts to launch the Spot Instances from Spot Instance pools with the lowest price possible, but also with optimal capacity for the number of instances that are launching. #### JSON ``` aws autoscaling create-auto-scaling-group --cli-input-json file://~/config.json ``` The `config.json` file contains the following content. ``` { "AutoScalingGroupName": "my-asg", "MixedInstancesPolicy": { "LaunchTemplate": { "LaunchTemplateSpecification": { "LaunchTemplateName": "my-launch-template", "Version": "$Latest" }, "Overrides": [ { "InstanceType": "c5.large" }, { "InstanceType": "c5a.large" }, { "InstanceType": "m5.large" }, { "InstanceType": "m5a.large" }, { "InstanceType": "c4.large" }, { "InstanceType": "m4.large" }, { "InstanceType": "c3.large" }, { "InstanceType": "m3.large" } ] }, "InstancesDistribution": { "OnDemandPercentageAboveBaseCapacity": 30, "SpotAllocationStrategy": "price-capacity-optimized" } }, "MinSize": 1, "MaxSize": 5, "DesiredCapacity": 3, "VPCZoneIdentifier": "subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782" } ``` #### YAML Alternatively, you can use the following [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command to create the Auto Scaling group. This references a YAML file as the sole parameter for your Auto Scaling group. ``` aws autoscaling create-auto-scaling-group --cli-input-yaml file://~/config.yaml ``` The `config.yaml` file contains the following content. ``` --- AutoScalingGroupName: my-asg MixedInstancesPolicy: LaunchTemplate: LaunchTemplateSpecification: LaunchTemplateName: my-launch-template Version: $Default Overrides: - InstanceType: c5.large - InstanceType: c5a.large - InstanceType: m5.large - InstanceType: m5a.large - InstanceType: c4.large - InstanceType: m4.large - InstanceType: c3.large - InstanceType: m3.large InstancesDistribution: OnDemandPercentageAboveBaseCapacity: 30 SpotAllocationStrategy: price-capacity-optimized MinSize: 1 MaxSize: 5 DesiredCapacity: 3 VPCZoneIdentifier: subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782 ``` # Configure an Auto Scaling group to use instance weights Use instance weights When you use multiple instance types, you can specify how many units to associate with each instance type, and then specify the capacity of your group with the same unit of measurement. This capacity specification option is known as weights. For example, let's say that you run a compute-intensive application that performs best with at least 8 vCPUs and 15 GiB of RAM. If you use `c5.2xlarge` as your base unit, any of the following EC2 instance types would meet your application needs. **Instance types example** | Instance type | vCPU | Memory (GiB) | | --- | --- | --- | | c5.2xlarge | 8 | 16 | | c5.4xlarge | 16 | 32 | | c5.12xlarge | 48 | 96 | | c5.18xlarge | 72 | 144 | | c5.24xlarge | 96 | 192 | By default, all instance types have equal weight regardless of size. In other words, whether Amazon EC2 Auto Scaling launches a large or small instance type, each instance counts the same toward the desired capacity of the Auto Scaling group. With weights, however, you assign a number value that specifies how many units to associate with each instance type. For example, if the instances are of different sizes, a `c5.2xlarge` instance could have the weight of 2, and a `c5.4xlarge` (which is two times bigger) could have the weight of 4, and so on. Then, when Amazon EC2 Auto Scaling scales the group, these weights translate into the number of units that each instance counts toward your desired capacity. The weights do not change which instance types Amazon EC2 Auto Scaling chooses to launch; instead, the allocation strategies do that. For more information, see [Allocation strategies for multiple instance types](allocation-strategies.md). **Important** To configure an Auto Scaling group to fulfill its desired capacity using the number of vCPUs or the amount of memory of each instance type, we recommend using attribute-based instance type selection. Setting the `DesiredCapacityType` parameter automatically specifies the number of units to associate with each instance type based on the value that you set for this parameter. For more information, see [Create mixed instances group using attribute-based instance type selection](create-mixed-instances-group-attribute-based-instance-type-selection.md). **Topics** + [ ## Considerations ](#weights-considerations) + [ ## Instance weight behaviors ](#instance-weighting-behaviors) + [ # Configure an Auto Scaling group to use weights ](configue-auto-scaling-group-to-use-weights.md) + [ # Spot price per unit hour example ](weights-spot-price-per-unit-hour-example.md) ## Considerations This section discusses key considerations for effectively implementing weights. + Choose a few instance types that match your application's performance needs. Decide the weight each instance type should count toward the desired capacity of your Auto Scaling group based on its capabilities. These weights apply to current and future instances. + Avoid large ranges between weights. For example, don't specify a weight of 1 for an instance type when the next larger instance type has a weight of 200. The difference between the smallest and largest weights shouldn't be extreme, either. Extreme weight differences can negatively impact cost-performance optimization. + Specify the group's desired capacity in units, not instances. For example, if you use vCPU-based weights, set your desired number of cores and also the minimum and maximum. + Set your weights and desired capacity so that the desired capacity is at least two to three times larger than your largest weight. Note the following when updating existing groups: + When you add weights to an existing group, include weights for all instance types currently in use. + When you add or change weights, Amazon EC2 Auto Scaling will launch or terminate instances to reach the desired capacity based on the new weight values. + If you remove an instance type, running instances of that type keep their last weight, even if no longer defined. ## Instance weight behaviors When you use instance weights, Amazon EC2 Auto Scaling behaves in the following way: + Current capacity will either be at the desired capacity or above it. Current capacity can exceed the desired capacity if instances launched that exceed the remaining desired capacity units. For example, suppose that you specify two instance types, `c5.2xlarge` and `c5.12xlarge`, and you assign instance weights of 2 for `c5.2xlarge` and 12 for `c5.12xlarge`. If there are five units remaining to fulfill the desired capacity, and Amazon EC2 Auto Scaling provisions a `c5.12xlarge`, the desired capacity is exceeded by seven units. + When launching instances, Amazon EC2 Auto Scaling prioritizes distributing capacity across Availability Zones and respecting allocation strategies over exceeding the desired capacity. + Amazon EC2 Auto Scaling can exceed the maximum capacity limit to maintain balance across Availability Zones, using your preferred allocation strategies. The hard limit enforced by Amazon EC2 Auto Scaling is your desired capacity plus your largest weight. # Configure an Auto Scaling group to use weights You can configure an Auto Scaling group to use weights, as shown in the following AWS CLI examples. For instructions on using the console, see [Create a mixed instances group by manually choosing instance types](create-mixed-instances-group-manual-instance-type-selection.md). **To configure a new Auto Scaling group to use weights (AWS CLI)** Use the [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command. For example, the following command creates a new Auto Scaling group and assigns weights by specifying the following: + The percentage of the group to launch as On-Demand Instances (`0`) + The allocation strategy for Spot Instances in each Availability Zone (`capacity-optimized`) + The instance types to launch in priority order (`m4.16xlarge`, `m5.24xlarge`) + The instance weights that correspond to the relative size difference (vCPUs) between instance types (`16`, `24`) + The subnets in which to launch the instances (`subnet-5ea0c127`, `subnet-6194ea3b`, `subnet-c934b782`), each corresponding to a different Availability Zone + The launch template (`my-launch-template`) and the launch template version (`$Latest`) ``` aws autoscaling create-auto-scaling-group --cli-input-json file://~/config.json ``` The `config.json` file contains the following content. ``` { "AutoScalingGroupName": "my-asg", "MixedInstancesPolicy": { "LaunchTemplate": { "LaunchTemplateSpecification": { "LaunchTemplateName": "my-launch-template", "Version": "$Latest" }, "Overrides": [ { "InstanceType": "m4.16xlarge", "WeightedCapacity": "16" }, { "InstanceType": "m5.24xlarge", "WeightedCapacity": "24" } ] }, "InstancesDistribution": { "OnDemandPercentageAboveBaseCapacity": 0, "SpotAllocationStrategy": "capacity-optimized" } }, "MinSize": 160, "MaxSize": 720, "DesiredCapacity": 480, "VPCZoneIdentifier": "subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782", "Tags": [] } ``` **To configure an existing Auto Scaling group to use weights (AWS CLI)** Use the [update-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/update-auto-scaling-group.html) command. For example, the following command assigns weights to instance types in an existing Auto Scaling group by specifying the following: + The instance types to launch in priority order (`c5.18xlarge`, `c5.24xlarge`, `c5.2xlarge`, `c5.4xlarge`) + The instance weights that correspond to the relative size difference (vCPUs) between instance types (`18`, `24`, `2`, `4`) + The new, increased desired capacity, which is larger than the largest weight ``` aws autoscaling update-auto-scaling-group --cli-input-json file://~/config.json ``` The `config.json` file contains the following content. ``` { "AutoScalingGroupName": "my-existing-asg", "MixedInstancesPolicy": { "LaunchTemplate": { "Overrides": [ { "InstanceType": "c5.18xlarge", "WeightedCapacity": "18" }, { "InstanceType": "c5.24xlarge", "WeightedCapacity": "24" }, { "InstanceType": "c5.2xlarge", "WeightedCapacity": "2" }, { "InstanceType": "c5.4xlarge", "WeightedCapacity": "4" } ] } }, "MinSize": 0, "MaxSize": 100, "DesiredCapacity": 100 } ``` **To verify the weights using the command line** Use one of the following commands: + [describe-auto-scaling-groups](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/describe-auto-scaling-groups.html) (AWS CLI) + [Get-ASAutoScalingGroup](https://docs.aws.amazon.com/powershell/latest/reference/items/Get-ASAutoScalingGroup.html) (AWS Tools for Windows PowerShell) # Spot price per unit hour example The following table compares the hourly price for Spot Instances in different Availability Zones in US East (Northern Virginia) with the price for On-Demand Instances in the same Region. The prices shown are example pricing and not current pricing. These are your costs per *instance* hour. **Example: Spot pricing per instance hour** | Instance type | us-east-1a | us-east-1b | us-east-1c | On-Demand pricing | | --- | --- | --- | --- | --- | | c5.2xlarge | \$10.180 | \$10.191 | \$10.170 | \$10.34 | | c5.4xlarge | \$10.341 | \$10.361 | \$10.318 | \$10.68 | | c5.12xlarge | \$10.779 | \$10.777 | \$10.777 | \$12.04 | | c5.18xlarge | \$11.207 | \$11.475 | \$11.357 | \$13.06 | | c5.24xlarge | \$11.555 | \$11.555 | \$11.555 | \$14.08 | With instance weights, you can evaluate your costs based on what you use per *unit* hour. You can determine the price per unit hour by dividing your price for an instance type by the number of units that it represents. For On-Demand Instances, the price per unit hour is the same when deploying one instance type as it is when deploying a different size of the same instance type. In contrast, however, the Spot price per unit hour varies by Spot pool. The following example shows how the Spot price per unit hour calculation works with instance weights. For ease of calculation, let's say you want to launch Spot Instances only in `us-east-1a`. The per unit hour price is captured in the following table. **Example: Spot Price per unit hour** | Instance type | us-east-1a | Instance weight | Price per unit hour | | --- | --- | --- | --- | | c5.2xlarge | \$10.180 | 2 | \$10.090 | | c5.4xlarge | \$10.341 | 4 | \$10.085 | | c5.12xlarge | \$10.779 | 12 | \$10.065 | | c5.18xlarge | \$11.207 | 18 | \$10.067 | | c5.24xlarge | \$11.555 | 24 | \$10.065 | # Use multiple launch templates In addition to using multiple instance types, you can also use multiple launch templates. For example, say that you configure an Auto Scaling group for compute-intensive applications and want to include a mix of C5, C5a, and C6g instance types. However, C6g instances feature an AWS Graviton processor based on 64-bit Arm architecture, while the C5 and C5a instances run on 64-bit Intel x86 processors. The AMIs for C5 and C5a instances both work on each of those instances, but not on C6g instances. To solve this problem, use a different launch template for C6g instances. You can still use the same launch template for C5 and C5a instances. This section contains procedures for using the AWS CLI to perform tasks related to using multiple launch templates. Currently, this feature is available only if you use the AWS CLI or an SDK, and is not available from the console. **Topics** + [ ## Configure an Auto Scaling group to use multiple launch templates ](#configue-auto-scaling-group-to-use-multiple-launch-templates) + [ ## Related resources ](#multiple-launch-templates-related-resources) ## Configure an Auto Scaling group to use multiple launch templates You can configure an Auto Scaling group to use multiple launch templates, as shown in the following examples. **To configure a new Auto Scaling group to use multiple launch templates (AWS CLI)** Use the [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command. For example, the following command creates a new Auto Scaling group. It specifies the `c5.large`, `c5a.large`, and `c6g.large` instance types and defines a new launch template for the `c6g.large` instance type to ensure that an appropriate AMI is used to launch Arm instances. Amazon EC2 Auto Scaling uses the order of instance types to determine which instance type to use first when fulfilling On-Demand capacity. ``` aws autoscaling create-auto-scaling-group --cli-input-json file://~/config.json ``` The `config.json` file contains the following content. ``` { "AutoScalingGroupName":"my-asg", "MixedInstancesPolicy":{ "LaunchTemplate":{ "LaunchTemplateSpecification":{ "LaunchTemplateName":"my-launch-template-for-x86", "Version":"$Latest" }, "Overrides":[ { "InstanceType":"c6g.large", "LaunchTemplateSpecification": { "LaunchTemplateName": "my-launch-template-for-arm", "Version": "$Latest" } }, { "InstanceType":"c5.large" }, { "InstanceType":"c5a.large" } ] }, "InstancesDistribution":{ "OnDemandBaseCapacity": 1, "OnDemandPercentageAboveBaseCapacity": 50, "SpotAllocationStrategy": "capacity-optimized" } }, "MinSize":1, "MaxSize":5, "DesiredCapacity":3, "VPCZoneIdentifier":"subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782", "Tags":[ ] } ``` **To configure an existing Auto Scaling group to use multiple launch templates (AWS CLI)** Use the [update-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/update-auto-scaling-group.html) command. For example, the following command assigns the launch template named `my-launch-template-for-arm` to the `c6g.large` instance type for the Auto Scaling group named *`my-asg`*. ``` aws autoscaling update-auto-scaling-group --cli-input-json file://~/config.json ``` The `config.json` file contains the following content. ``` { "AutoScalingGroupName":"my-asg", "MixedInstancesPolicy":{ "LaunchTemplate":{ "Overrides":[ { "InstanceType":"c6g.large", "LaunchTemplateSpecification": { "LaunchTemplateName": "my-launch-template-for-arm", "Version": "$Latest" } }, { "InstanceType":"c5.large" }, { "InstanceType":"c5a.large" } ] } } } ``` **To configure a new Auto Scaling group to use multiple launch templates with attribute-based instance type selection (AWS CLI)** Use the [create-auto-scaling-group](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/create-auto-scaling-group.html) command. For example, the following command creates a new Auto Scaling group by specifying a launch template for AWS Graviton instances with an ARM AMI and an additional launch template for AMD or Intel based instances with an x86 AMI. Then, it uses [attribute-based instance selection](create-mixed-instances-group-attribute-based-instance-type-selection.md) twice to select from a wide range of instance types for each CPU architecture. You can add a similar configuration to an existing Auto Scaling group with the [update-autoscaling-group](https://docs.aws.amazon.com/cli/latest/reference/autoscaling/update-auto-scaling-group.html) command. ``` aws autoscaling create-auto-scaling-group --cli-input-json file://~/config.json ``` The `config.json` file contains the following content. ``` { "AutoScalingGroupName":"my-asg", "MixedInstancesPolicy":{ "LaunchTemplate":{ "LaunchTemplateSpecification":{ "LaunchTemplateName":"my-launch-template-for-arm", "Version":"$Latest" }, "Overrides":[ { "InstanceRequirements": { "VCpuCount": {"Min": 2}, "MemoryMiB": {"Min": 2048}, "CpuManufacturers": ["amazon-web-services"] } }, { "InstanceRequirements": { "VCpuCount": {"Min": 2}, "MemoryMiB": {"Min": 2048}, "CpuManufacturers": ["intel", "amd"] }, "LaunchTemplateSpecification": { "LaunchTemplateName": "my-launch-template-for-x86", "Version": "$Latest" } } ] }, "InstancesDistribution":{ "OnDemandPercentageAboveBaseCapacity": 0, "SpotAllocationStrategy": "price-capacity-optimized" } }, "MinSize":1, "MaxSize":10, "DesiredCapacity":6, "VPCZoneIdentifier":"subnet-5ea0c127,subnet-6194ea3b,subnet-c934b782", "Tags":[ ] } ``` **To verify the launch templates for an Auto Scaling group** Use one of the following commands: + [describe-auto-scaling-groups](https://awscli.amazonaws.com/v2/documentation/api/latest/reference/autoscaling/describe-auto-scaling-groups.html) (AWS CLI) + [Get-ASAutoScalingGroup](https://docs.aws.amazon.com/powershell/latest/reference/items/Get-ASAutoScalingGroup.html) (AWS Tools for Windows PowerShell) ## Related resources You can find an example of specifying multiple launch templates using attribute-based instance type selection in a CloudFormation template on [AWS re:Post](https://repost.aws/articles/ARQeKDQX68TcqipYaaisl6bA/cloudformation-auto-scaling-group-sample-template-for-mixed-x86-intel-amd-and-aws-graviton-instances).