# frozen_string_literal: true
# WARNING ABOUT GENERATED CODE
#
# This file is generated. See the contributing guide for more information:
# https://github.com/aws/aws-sdk-ruby/blob/version-3/CONTRIBUTING.md
#
# WARNING ABOUT GENERATED CODE
require 'seahorse/client/plugins/content_length.rb'
require 'aws-sdk-core/plugins/credentials_configuration.rb'
require 'aws-sdk-core/plugins/logging.rb'
require 'aws-sdk-core/plugins/param_converter.rb'
require 'aws-sdk-core/plugins/param_validator.rb'
require 'aws-sdk-core/plugins/user_agent.rb'
require 'aws-sdk-core/plugins/helpful_socket_errors.rb'
require 'aws-sdk-core/plugins/retry_errors.rb'
require 'aws-sdk-core/plugins/global_configuration.rb'
require 'aws-sdk-core/plugins/regional_endpoint.rb'
require 'aws-sdk-core/plugins/endpoint_discovery.rb'
require 'aws-sdk-core/plugins/endpoint_pattern.rb'
require 'aws-sdk-core/plugins/response_paging.rb'
require 'aws-sdk-core/plugins/stub_responses.rb'
require 'aws-sdk-core/plugins/idempotency_token.rb'
require 'aws-sdk-core/plugins/invocation_id.rb'
require 'aws-sdk-core/plugins/jsonvalue_converter.rb'
require 'aws-sdk-core/plugins/client_metrics_plugin.rb'
require 'aws-sdk-core/plugins/client_metrics_send_plugin.rb'
require 'aws-sdk-core/plugins/transfer_encoding.rb'
require 'aws-sdk-core/plugins/http_checksum.rb'
require 'aws-sdk-core/plugins/checksum_algorithm.rb'
require 'aws-sdk-core/plugins/request_compression.rb'
require 'aws-sdk-core/plugins/defaults_mode.rb'
require 'aws-sdk-core/plugins/recursion_detection.rb'
require 'aws-sdk-core/plugins/sign.rb'
require 'aws-sdk-core/plugins/protocols/json_rpc.rb'
Aws::Plugins::GlobalConfiguration.add_identifier(:computeoptimizer)
module Aws::ComputeOptimizer
# An API client for ComputeOptimizer. To construct a client, you need to configure a `:region` and `:credentials`.
#
# client = Aws::ComputeOptimizer::Client.new(
# region: region_name,
# credentials: credentials,
# # ...
# )
#
# For details on configuring region and credentials see
# the [developer guide](/sdk-for-ruby/v3/developer-guide/setup-config.html).
#
# See {#initialize} for a full list of supported configuration options.
class Client < Seahorse::Client::Base
include Aws::ClientStubs
@identifier = :computeoptimizer
set_api(ClientApi::API)
add_plugin(Seahorse::Client::Plugins::ContentLength)
add_plugin(Aws::Plugins::CredentialsConfiguration)
add_plugin(Aws::Plugins::Logging)
add_plugin(Aws::Plugins::ParamConverter)
add_plugin(Aws::Plugins::ParamValidator)
add_plugin(Aws::Plugins::UserAgent)
add_plugin(Aws::Plugins::HelpfulSocketErrors)
add_plugin(Aws::Plugins::RetryErrors)
add_plugin(Aws::Plugins::GlobalConfiguration)
add_plugin(Aws::Plugins::RegionalEndpoint)
add_plugin(Aws::Plugins::EndpointDiscovery)
add_plugin(Aws::Plugins::EndpointPattern)
add_plugin(Aws::Plugins::ResponsePaging)
add_plugin(Aws::Plugins::StubResponses)
add_plugin(Aws::Plugins::IdempotencyToken)
add_plugin(Aws::Plugins::InvocationId)
add_plugin(Aws::Plugins::JsonvalueConverter)
add_plugin(Aws::Plugins::ClientMetricsPlugin)
add_plugin(Aws::Plugins::ClientMetricsSendPlugin)
add_plugin(Aws::Plugins::TransferEncoding)
add_plugin(Aws::Plugins::HttpChecksum)
add_plugin(Aws::Plugins::ChecksumAlgorithm)
add_plugin(Aws::Plugins::RequestCompression)
add_plugin(Aws::Plugins::DefaultsMode)
add_plugin(Aws::Plugins::RecursionDetection)
add_plugin(Aws::Plugins::Sign)
add_plugin(Aws::Plugins::Protocols::JsonRpc)
add_plugin(Aws::ComputeOptimizer::Plugins::Endpoints)
# @overload initialize(options)
# @param [Hash] options
# @option options [required, Aws::CredentialProvider] :credentials
# Your AWS credentials. This can be an instance of any one of the
# following classes:
#
# * `Aws::Credentials` - Used for configuring static, non-refreshing
# credentials.
#
# * `Aws::SharedCredentials` - Used for loading static credentials from a
# shared file, such as `~/.aws/config`.
#
# * `Aws::AssumeRoleCredentials` - Used when you need to assume a role.
#
# * `Aws::AssumeRoleWebIdentityCredentials` - Used when you need to
# assume a role after providing credentials via the web.
#
# * `Aws::SSOCredentials` - Used for loading credentials from AWS SSO using an
# access token generated from `aws login`.
#
# * `Aws::ProcessCredentials` - Used for loading credentials from a
# process that outputs to stdout.
#
# * `Aws::InstanceProfileCredentials` - Used for loading credentials
# from an EC2 IMDS on an EC2 instance.
#
# * `Aws::ECSCredentials` - Used for loading credentials from
# instances running in ECS.
#
# * `Aws::CognitoIdentityCredentials` - Used for loading credentials
# from the Cognito Identity service.
#
# When `:credentials` are not configured directly, the following
# locations will be searched for credentials:
#
# * `Aws.config[:credentials]`
# * The `:access_key_id`, `:secret_access_key`, and `:session_token` options.
# * ENV['AWS_ACCESS_KEY_ID'], ENV['AWS_SECRET_ACCESS_KEY']
# * `~/.aws/credentials`
# * `~/.aws/config`
# * EC2/ECS IMDS instance profile - When used by default, the timeouts
# are very aggressive. Construct and pass an instance of
# `Aws::InstanceProfileCredentails` or `Aws::ECSCredentials` to
# enable retries and extended timeouts. Instance profile credential
# fetching can be disabled by setting ENV['AWS_EC2_METADATA_DISABLED']
# to true.
#
# @option options [required, String] :region
# The AWS region to connect to. The configured `:region` is
# used to determine the service `:endpoint`. When not passed,
# a default `:region` is searched for in the following locations:
#
# * `Aws.config[:region]`
# * `ENV['AWS_REGION']`
# * `ENV['AMAZON_REGION']`
# * `ENV['AWS_DEFAULT_REGION']`
# * `~/.aws/credentials`
# * `~/.aws/config`
#
# @option options [String] :access_key_id
#
# @option options [Boolean] :active_endpoint_cache (false)
# When set to `true`, a thread polling for endpoints will be running in
# the background every 60 secs (default). Defaults to `false`.
#
# @option options [Boolean] :adaptive_retry_wait_to_fill (true)
# Used only in `adaptive` retry mode. When true, the request will sleep
# until there is sufficent client side capacity to retry the request.
# When false, the request will raise a `RetryCapacityNotAvailableError` and will
# not retry instead of sleeping.
#
# @option options [Boolean] :client_side_monitoring (false)
# When `true`, client-side metrics will be collected for all API requests from
# this client.
#
# @option options [String] :client_side_monitoring_client_id ("")
# Allows you to provide an identifier for this client which will be attached to
# all generated client side metrics. Defaults to an empty string.
#
# @option options [String] :client_side_monitoring_host ("127.0.0.1")
# Allows you to specify the DNS hostname or IPv4 or IPv6 address that the client
# side monitoring agent is running on, where client metrics will be published via UDP.
#
# @option options [Integer] :client_side_monitoring_port (31000)
# Required for publishing client metrics. The port that the client side monitoring
# agent is running on, where client metrics will be published via UDP.
#
# @option options [Aws::ClientSideMonitoring::Publisher] :client_side_monitoring_publisher (Aws::ClientSideMonitoring::Publisher)
# Allows you to provide a custom client-side monitoring publisher class. By default,
# will use the Client Side Monitoring Agent Publisher.
#
# @option options [Boolean] :convert_params (true)
# When `true`, an attempt is made to coerce request parameters into
# the required types.
#
# @option options [Boolean] :correct_clock_skew (true)
# Used only in `standard` and adaptive retry modes. Specifies whether to apply
# a clock skew correction and retry requests with skewed client clocks.
#
# @option options [String] :defaults_mode ("legacy")
# See {Aws::DefaultsModeConfiguration} for a list of the
# accepted modes and the configuration defaults that are included.
#
# @option options [Boolean] :disable_host_prefix_injection (false)
# Set to true to disable SDK automatically adding host prefix
# to default service endpoint when available.
#
# @option options [Boolean] :disable_request_compression (false)
# When set to 'true' the request body will not be compressed
# for supported operations.
#
# @option options [String, URI::HTTPS, URI::HTTP] :endpoint
# Normally you should not configure the `:endpoint` option
# directly. This is normally constructed from the `:region`
# option. Configuring `:endpoint` is normally reserved for
# connecting to test or custom endpoints. The endpoint should
# be a URI formatted like:
#
# 'http://example.com'
# 'https://example.com'
# 'http://example.com:123'
#
#
# @option options [Integer] :endpoint_cache_max_entries (1000)
# Used for the maximum size limit of the LRU cache storing endpoints data
# for endpoint discovery enabled operations. Defaults to 1000.
#
# @option options [Integer] :endpoint_cache_max_threads (10)
# Used for the maximum threads in use for polling endpoints to be cached, defaults to 10.
#
# @option options [Integer] :endpoint_cache_poll_interval (60)
# When :endpoint_discovery and :active_endpoint_cache is enabled,
# Use this option to config the time interval in seconds for making
# requests fetching endpoints information. Defaults to 60 sec.
#
# @option options [Boolean] :endpoint_discovery (false)
# When set to `true`, endpoint discovery will be enabled for operations when available.
#
# @option options [Boolean] :ignore_configured_endpoint_urls
# Setting to true disables use of endpoint URLs provided via environment
# variables and the shared configuration file.
#
# @option options [Aws::Log::Formatter] :log_formatter (Aws::Log::Formatter.default)
# The log formatter.
#
# @option options [Symbol] :log_level (:info)
# The log level to send messages to the `:logger` at.
#
# @option options [Logger] :logger
# The Logger instance to send log messages to. If this option
# is not set, logging will be disabled.
#
# @option options [Integer] :max_attempts (3)
# An integer representing the maximum number attempts that will be made for
# a single request, including the initial attempt. For example,
# setting this value to 5 will result in a request being retried up to
# 4 times. Used in `standard` and `adaptive` retry modes.
#
# @option options [String] :profile ("default")
# Used when loading credentials from the shared credentials file
# at HOME/.aws/credentials. When not specified, 'default' is used.
#
# @option options [Integer] :request_min_compression_size_bytes (10240)
# The minimum size in bytes that triggers compression for request
# bodies. The value must be non-negative integer value between 0
# and 10485780 bytes inclusive.
#
# @option options [Proc] :retry_backoff
# A proc or lambda used for backoff. Defaults to 2**retries * retry_base_delay.
# This option is only used in the `legacy` retry mode.
#
# @option options [Float] :retry_base_delay (0.3)
# The base delay in seconds used by the default backoff function. This option
# is only used in the `legacy` retry mode.
#
# @option options [Symbol] :retry_jitter (:none)
# A delay randomiser function used by the default backoff function.
# Some predefined functions can be referenced by name - :none, :equal, :full,
# otherwise a Proc that takes and returns a number. This option is only used
# in the `legacy` retry mode.
#
# @see https://www.awsarchitectureblog.com/2015/03/backoff.html
#
# @option options [Integer] :retry_limit (3)
# The maximum number of times to retry failed requests. Only
# ~ 500 level server errors and certain ~ 400 level client errors
# are retried. Generally, these are throttling errors, data
# checksum errors, networking errors, timeout errors, auth errors,
# endpoint discovery, and errors from expired credentials.
# This option is only used in the `legacy` retry mode.
#
# @option options [Integer] :retry_max_delay (0)
# The maximum number of seconds to delay between retries (0 for no limit)
# used by the default backoff function. This option is only used in the
# `legacy` retry mode.
#
# @option options [String] :retry_mode ("legacy")
# Specifies which retry algorithm to use. Values are:
#
# * `legacy` - The pre-existing retry behavior. This is default value if
# no retry mode is provided.
#
# * `standard` - A standardized set of retry rules across the AWS SDKs.
# This includes support for retry quotas, which limit the number of
# unsuccessful retries a client can make.
#
# * `adaptive` - An experimental retry mode that includes all the
# functionality of `standard` mode along with automatic client side
# throttling. This is a provisional mode that may change behavior
# in the future.
#
#
# @option options [String] :sdk_ua_app_id
# A unique and opaque application ID that is appended to the
# User-Agent header as app/sdk_ua_app_id. It should have a
# maximum length of 50. This variable is sourced from environment
# variable AWS_SDK_UA_APP_ID or the shared config profile attribute sdk_ua_app_id.
#
# @option options [String] :secret_access_key
#
# @option options [String] :session_token
#
# @option options [Boolean] :simple_json (false)
# Disables request parameter conversion, validation, and formatting.
# Also disable response data type conversions. This option is useful
# when you want to ensure the highest level of performance by
# avoiding overhead of walking request parameters and response data
# structures.
#
# When `:simple_json` is enabled, the request parameters hash must
# be formatted exactly as the DynamoDB API expects.
#
# @option options [Boolean] :stub_responses (false)
# Causes the client to return stubbed responses. By default
# fake responses are generated and returned. You can specify
# the response data to return or errors to raise by calling
# {ClientStubs#stub_responses}. See {ClientStubs} for more information.
#
# ** Please note ** When response stubbing is enabled, no HTTP
# requests are made, and retries are disabled.
#
# @option options [Aws::TokenProvider] :token_provider
# A Bearer Token Provider. This can be an instance of any one of the
# following classes:
#
# * `Aws::StaticTokenProvider` - Used for configuring static, non-refreshing
# tokens.
#
# * `Aws::SSOTokenProvider` - Used for loading tokens from AWS SSO using an
# access token generated from `aws login`.
#
# When `:token_provider` is not configured directly, the `Aws::TokenProviderChain`
# will be used to search for tokens configured for your profile in shared configuration files.
#
# @option options [Boolean] :use_dualstack_endpoint
# When set to `true`, dualstack enabled endpoints (with `.aws` TLD)
# will be used if available.
#
# @option options [Boolean] :use_fips_endpoint
# When set to `true`, fips compatible endpoints will be used if available.
# When a `fips` region is used, the region is normalized and this config
# is set to `true`.
#
# @option options [Boolean] :validate_params (true)
# When `true`, request parameters are validated before
# sending the request.
#
# @option options [Aws::ComputeOptimizer::EndpointProvider] :endpoint_provider
# The endpoint provider used to resolve endpoints. Any object that responds to `#resolve_endpoint(parameters)` where `parameters` is a Struct similar to `Aws::ComputeOptimizer::EndpointParameters`
#
# @option options [Float] :http_continue_timeout (1)
# The number of seconds to wait for a 100-continue response before sending the
# request body. This option has no effect unless the request has "Expect"
# header set to "100-continue". Defaults to `nil` which disables this
# behaviour. This value can safely be set per request on the session.
#
# @option options [Float] :http_idle_timeout (5)
# The number of seconds a connection is allowed to sit idle before it
# is considered stale. Stale connections are closed and removed from the
# pool before making a request.
#
# @option options [Float] :http_open_timeout (15)
# The default number of seconds to wait for response data.
# This value can safely be set per-request on the session.
#
# @option options [URI::HTTP,String] :http_proxy
# A proxy to send requests through. Formatted like 'http://proxy.com:123'.
#
# @option options [Float] :http_read_timeout (60)
# The default number of seconds to wait for response data.
# This value can safely be set per-request on the session.
#
# @option options [Boolean] :http_wire_trace (false)
# When `true`, HTTP debug output will be sent to the `:logger`.
#
# @option options [Proc] :on_chunk_received
# When a Proc object is provided, it will be used as callback when each chunk
# of the response body is received. It provides three arguments: the chunk,
# the number of bytes received, and the total number of
# bytes in the response (or nil if the server did not send a `content-length`).
#
# @option options [Proc] :on_chunk_sent
# When a Proc object is provided, it will be used as callback when each chunk
# of the request body is sent. It provides three arguments: the chunk,
# the number of bytes read from the body, and the total number of
# bytes in the body.
#
# @option options [Boolean] :raise_response_errors (true)
# When `true`, response errors are raised.
#
# @option options [String] :ssl_ca_bundle
# Full path to the SSL certificate authority bundle file that should be used when
# verifying peer certificates. If you do not pass `:ssl_ca_bundle` or
# `:ssl_ca_directory` the the system default will be used if available.
#
# @option options [String] :ssl_ca_directory
# Full path of the directory that contains the unbundled SSL certificate
# authority files for verifying peer certificates. If you do
# not pass `:ssl_ca_bundle` or `:ssl_ca_directory` the the system
# default will be used if available.
#
# @option options [String] :ssl_ca_store
# Sets the X509::Store to verify peer certificate.
#
# @option options [Float] :ssl_timeout
# Sets the SSL timeout in seconds
#
# @option options [Boolean] :ssl_verify_peer (true)
# When `true`, SSL peer certificates are verified when establishing a connection.
#
def initialize(*args)
super
end
# @!group API Operations
# Deletes a recommendation preference, such as enhanced infrastructure
# metrics.
#
# For more information, see [Activating enhanced infrastructure
# metrics][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/enhanced-infrastructure-metrics.html
#
# @option params [required, String] :resource_type
# The target resource type of the recommendation preference to delete.
#
# The `Ec2Instance` option encompasses standalone instances and
# instances that are part of Auto Scaling groups. The `AutoScalingGroup`
# option encompasses only instances that are part of an Auto Scaling
# group.
#
# The valid values for this parameter are `Ec2Instance` and
# `AutoScalingGroup`.
#
#
#
# @option params [Types::Scope] :scope
# An object that describes the scope of the recommendation preference to
# delete.
#
# You can delete recommendation preferences that are created at the
# organization level (for management accounts of an organization only),
# account level, and resource level. For more information, see
# [Activating enhanced infrastructure metrics][1] in the *Compute
# Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/enhanced-infrastructure-metrics.html
#
# @option params [required, Array] :recommendation_preference_names
# The name of the recommendation preference to delete.
#
# @return [Struct] Returns an empty {Seahorse::Client::Response response}.
#
# @example Request syntax with placeholder values
#
# resp = client.delete_recommendation_preferences({
# resource_type: "Ec2Instance", # required, accepts Ec2Instance, AutoScalingGroup, EbsVolume, LambdaFunction, NotApplicable, EcsService, License
# scope: {
# name: "Organization", # accepts Organization, AccountId, ResourceArn
# value: "ScopeValue",
# },
# recommendation_preference_names: ["EnhancedInfrastructureMetrics"], # required, accepts EnhancedInfrastructureMetrics, InferredWorkloadTypes, ExternalMetricsPreference, LookBackPeriodPreference, PreferredResources, UtilizationPreferences
# })
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/DeleteRecommendationPreferences AWS API Documentation
#
# @overload delete_recommendation_preferences(params = {})
# @param [Hash] params ({})
def delete_recommendation_preferences(params = {}, options = {})
req = build_request(:delete_recommendation_preferences, params)
req.send_request(options)
end
# Describes recommendation export jobs created in the last seven days.
#
# Use the ExportAutoScalingGroupRecommendations or
# ExportEC2InstanceRecommendations actions to request an export of your
# recommendations. Then use the DescribeRecommendationExportJobs action
# to view your export jobs.
#
# @option params [Array] :job_ids
# The identification numbers of the export jobs to return.
#
# An export job ID is returned when you create an export using the
# ExportAutoScalingGroupRecommendations or
# ExportEC2InstanceRecommendations actions.
#
# All export jobs created in the last seven days are returned if this
# parameter is omitted.
#
# @option params [Array] :filters
# An array of objects to specify a filter that returns a more specific
# list of export jobs.
#
# @option params [String] :next_token
# The token to advance to the next page of export jobs.
#
# @option params [Integer] :max_results
# The maximum number of export jobs to return with a single request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @return [Types::DescribeRecommendationExportJobsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::DescribeRecommendationExportJobsResponse#recommendation_export_jobs #recommendation_export_jobs} => Array<Types::RecommendationExportJob>
# * {Types::DescribeRecommendationExportJobsResponse#next_token #next_token} => String
#
# The returned {Seahorse::Client::Response response} is a pageable response and is Enumerable. For details on usage see {Aws::PageableResponse PageableResponse}.
#
# @example Request syntax with placeholder values
#
# resp = client.describe_recommendation_export_jobs({
# job_ids: ["JobId"],
# filters: [
# {
# name: "ResourceType", # accepts ResourceType, JobStatus
# values: ["FilterValue"],
# },
# ],
# next_token: "NextToken",
# max_results: 1,
# })
#
# @example Response structure
#
# resp.recommendation_export_jobs #=> Array
# resp.recommendation_export_jobs[0].job_id #=> String
# resp.recommendation_export_jobs[0].destination.s3.bucket #=> String
# resp.recommendation_export_jobs[0].destination.s3.key #=> String
# resp.recommendation_export_jobs[0].destination.s3.metadata_key #=> String
# resp.recommendation_export_jobs[0].resource_type #=> String, one of "Ec2Instance", "AutoScalingGroup", "EbsVolume", "LambdaFunction", "NotApplicable", "EcsService", "License"
# resp.recommendation_export_jobs[0].status #=> String, one of "Queued", "InProgress", "Complete", "Failed"
# resp.recommendation_export_jobs[0].creation_timestamp #=> Time
# resp.recommendation_export_jobs[0].last_updated_timestamp #=> Time
# resp.recommendation_export_jobs[0].failure_reason #=> String
# resp.next_token #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/DescribeRecommendationExportJobs AWS API Documentation
#
# @overload describe_recommendation_export_jobs(params = {})
# @param [Hash] params ({})
def describe_recommendation_export_jobs(params = {}, options = {})
req = build_request(:describe_recommendation_export_jobs, params)
req.send_request(options)
end
# Exports optimization recommendations for Auto Scaling groups.
#
# Recommendations are exported in a comma-separated values (.csv) file,
# and its metadata in a JavaScript Object Notation (JSON) (.json) file,
# to an existing Amazon Simple Storage Service (Amazon S3) bucket that
# you specify. For more information, see [Exporting Recommendations][1]
# in the *Compute Optimizer User Guide*.
#
# You can have only one Auto Scaling group export job in progress per
# Amazon Web Services Region.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html
#
# @option params [Array] :account_ids
# The IDs of the Amazon Web Services accounts for which to export Auto
# Scaling group recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to export
# recommendations.
#
# This parameter cannot be specified together with the include member
# accounts parameter. The parameters are mutually exclusive.
#
# Recommendations for member accounts are not included in the export if
# this parameter, or the include member accounts parameter, is omitted.
#
# You can specify multiple account IDs per request.
#
# @option params [Array] :filters
# An array of objects to specify a filter that exports a more specific
# set of Auto Scaling group recommendations.
#
# @option params [Array] :fields_to_export
# The recommendations data to include in the export file. For more
# information about the fields that can be exported, see [Exported
# files][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html#exported-files
#
# @option params [required, Types::S3DestinationConfig] :s3_destination_config
# An object to specify the destination Amazon Simple Storage Service
# (Amazon S3) bucket name and key prefix for the export job.
#
# You must create the destination Amazon S3 bucket for your
# recommendations export before you create the export job. Compute
# Optimizer does not create the S3 bucket for you. After you create the
# S3 bucket, ensure that it has the required permissions policy to allow
# Compute Optimizer to write the export file to it. If you plan to
# specify an object prefix when you create the export job, you must
# include the object prefix in the policy that you add to the S3 bucket.
# For more information, see [Amazon S3 Bucket Policy for Compute
# Optimizer][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/create-s3-bucket-policy-for-compute-optimizer.html
#
# @option params [String] :file_format
# The format of the export file.
#
# The only export file format currently supported is `Csv`.
#
# @option params [Boolean] :include_member_accounts
# Indicates whether to include recommendations for resources in all
# member accounts of the organization if your account is the management
# account of an organization.
#
# The member accounts must also be opted in to Compute Optimizer, and
# trusted access for Compute Optimizer must be enabled in the
# organization account. For more information, see [Compute Optimizer and
# Amazon Web Services Organizations trusted access][1] in the *Compute
# Optimizer User Guide*.
#
# Recommendations for member accounts of the organization are not
# included in the export file if this parameter is omitted.
#
# This parameter cannot be specified together with the account IDs
# parameter. The parameters are mutually exclusive.
#
# Recommendations for member accounts are not included in the export if
# this parameter, or the account IDs parameter, is omitted.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/security-iam.html#trusted-service-access
#
# @option params [Types::RecommendationPreferences] :recommendation_preferences
# An object to specify the preferences for the Auto Scaling group
# recommendations to export.
#
# @return [Types::ExportAutoScalingGroupRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::ExportAutoScalingGroupRecommendationsResponse#job_id #job_id} => String
# * {Types::ExportAutoScalingGroupRecommendationsResponse#s3_destination #s3_destination} => Types::S3Destination
#
# @example Request syntax with placeholder values
#
# resp = client.export_auto_scaling_group_recommendations({
# account_ids: ["AccountId"],
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCodes, RecommendationSourceType, InferredWorkloadTypes
# values: ["FilterValue"],
# },
# ],
# fields_to_export: ["AccountId"], # accepts AccountId, AutoScalingGroupArn, AutoScalingGroupName, Finding, UtilizationMetricsCpuMaximum, UtilizationMetricsMemoryMaximum, UtilizationMetricsEbsReadOpsPerSecondMaximum, UtilizationMetricsEbsWriteOpsPerSecondMaximum, UtilizationMetricsEbsReadBytesPerSecondMaximum, UtilizationMetricsEbsWriteBytesPerSecondMaximum, UtilizationMetricsDiskReadOpsPerSecondMaximum, UtilizationMetricsDiskWriteOpsPerSecondMaximum, UtilizationMetricsDiskReadBytesPerSecondMaximum, UtilizationMetricsDiskWriteBytesPerSecondMaximum, UtilizationMetricsNetworkInBytesPerSecondMaximum, UtilizationMetricsNetworkOutBytesPerSecondMaximum, UtilizationMetricsNetworkPacketsInPerSecondMaximum, UtilizationMetricsNetworkPacketsOutPerSecondMaximum, LookbackPeriodInDays, CurrentConfigurationInstanceType, CurrentConfigurationDesiredCapacity, CurrentConfigurationMinSize, CurrentConfigurationMaxSize, CurrentOnDemandPrice, CurrentStandardOneYearNoUpfrontReservedPrice, CurrentStandardThreeYearNoUpfrontReservedPrice, CurrentVCpus, CurrentMemory, CurrentStorage, CurrentNetwork, RecommendationOptionsConfigurationInstanceType, RecommendationOptionsConfigurationDesiredCapacity, RecommendationOptionsConfigurationMinSize, RecommendationOptionsConfigurationMaxSize, RecommendationOptionsProjectedUtilizationMetricsCpuMaximum, RecommendationOptionsProjectedUtilizationMetricsMemoryMaximum, RecommendationOptionsPerformanceRisk, RecommendationOptionsOnDemandPrice, RecommendationOptionsStandardOneYearNoUpfrontReservedPrice, RecommendationOptionsStandardThreeYearNoUpfrontReservedPrice, RecommendationOptionsVcpus, RecommendationOptionsMemory, RecommendationOptionsStorage, RecommendationOptionsNetwork, LastRefreshTimestamp, CurrentPerformanceRisk, RecommendationOptionsSavingsOpportunityPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrency, RecommendationOptionsEstimatedMonthlySavingsValue, EffectiveRecommendationPreferencesCpuVendorArchitectures, EffectiveRecommendationPreferencesEnhancedInfrastructureMetrics, EffectiveRecommendationPreferencesInferredWorkloadTypes, InferredWorkloadTypes, RecommendationOptionsMigrationEffort, CurrentInstanceGpuInfo, RecommendationOptionsInstanceGpuInfo, UtilizationMetricsGpuPercentageMaximum, UtilizationMetricsGpuMemoryPercentageMaximum, RecommendationOptionsProjectedUtilizationMetricsGpuPercentageMaximum, RecommendationOptionsProjectedUtilizationMetricsGpuMemoryPercentageMaximum, EffectiveRecommendationPreferencesSavingsEstimationMode, RecommendationOptionsSavingsOpportunityAfterDiscountsPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrencyAfterDiscounts, RecommendationOptionsEstimatedMonthlySavingsValueAfterDiscounts, EffectiveRecommendationPreferencesPreferredResources, EffectiveRecommendationPreferencesLookBackPeriod
# s3_destination_config: { # required
# bucket: "DestinationBucket",
# key_prefix: "DestinationKeyPrefix",
# },
# file_format: "Csv", # accepts Csv
# include_member_accounts: false,
# recommendation_preferences: {
# cpu_vendor_architectures: ["AWS_ARM64"], # accepts AWS_ARM64, CURRENT
# },
# })
#
# @example Response structure
#
# resp.job_id #=> String
# resp.s3_destination.bucket #=> String
# resp.s3_destination.key #=> String
# resp.s3_destination.metadata_key #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/ExportAutoScalingGroupRecommendations AWS API Documentation
#
# @overload export_auto_scaling_group_recommendations(params = {})
# @param [Hash] params ({})
def export_auto_scaling_group_recommendations(params = {}, options = {})
req = build_request(:export_auto_scaling_group_recommendations, params)
req.send_request(options)
end
# Exports optimization recommendations for Amazon EBS volumes.
#
# Recommendations are exported in a comma-separated values (.csv) file,
# and its metadata in a JavaScript Object Notation (JSON) (.json) file,
# to an existing Amazon Simple Storage Service (Amazon S3) bucket that
# you specify. For more information, see [Exporting Recommendations][1]
# in the *Compute Optimizer User Guide*.
#
# You can have only one Amazon EBS volume export job in progress per
# Amazon Web Services Region.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html
#
# @option params [Array] :account_ids
# The IDs of the Amazon Web Services accounts for which to export Amazon
# EBS volume recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to export
# recommendations.
#
# This parameter cannot be specified together with the include member
# accounts parameter. The parameters are mutually exclusive.
#
# Recommendations for member accounts are not included in the export if
# this parameter, or the include member accounts parameter, is omitted.
#
# You can specify multiple account IDs per request.
#
# @option params [Array] :filters
# An array of objects to specify a filter that exports a more specific
# set of Amazon EBS volume recommendations.
#
# @option params [Array] :fields_to_export
# The recommendations data to include in the export file. For more
# information about the fields that can be exported, see [Exported
# files][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html#exported-files
#
# @option params [required, Types::S3DestinationConfig] :s3_destination_config
# Describes the destination Amazon Simple Storage Service (Amazon S3)
# bucket name and key prefix for a recommendations export job.
#
# You must create the destination Amazon S3 bucket for your
# recommendations export before you create the export job. Compute
# Optimizer does not create the S3 bucket for you. After you create the
# S3 bucket, ensure that it has the required permission policy to allow
# Compute Optimizer to write the export file to it. If you plan to
# specify an object prefix when you create the export job, you must
# include the object prefix in the policy that you add to the S3 bucket.
# For more information, see [Amazon S3 Bucket Policy for Compute
# Optimizer][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/create-s3-bucket-policy-for-compute-optimizer.html
#
# @option params [String] :file_format
# The format of the export file.
#
# The only export file format currently supported is `Csv`.
#
# @option params [Boolean] :include_member_accounts
# Indicates whether to include recommendations for resources in all
# member accounts of the organization if your account is the management
# account of an organization.
#
# The member accounts must also be opted in to Compute Optimizer, and
# trusted access for Compute Optimizer must be enabled in the
# organization account. For more information, see [Compute Optimizer and
# Amazon Web Services Organizations trusted access][1] in the *Compute
# Optimizer User Guide*.
#
# Recommendations for member accounts of the organization are not
# included in the export file if this parameter is omitted.
#
# This parameter cannot be specified together with the account IDs
# parameter. The parameters are mutually exclusive.
#
# Recommendations for member accounts are not included in the export if
# this parameter, or the account IDs parameter, is omitted.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/security-iam.html#trusted-service-access
#
# @return [Types::ExportEBSVolumeRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::ExportEBSVolumeRecommendationsResponse#job_id #job_id} => String
# * {Types::ExportEBSVolumeRecommendationsResponse#s3_destination #s3_destination} => Types::S3Destination
#
# @example Request syntax with placeholder values
#
# resp = client.export_ebs_volume_recommendations({
# account_ids: ["AccountId"],
# filters: [
# {
# name: "Finding", # accepts Finding
# values: ["FilterValue"],
# },
# ],
# fields_to_export: ["AccountId"], # accepts AccountId, VolumeArn, Finding, UtilizationMetricsVolumeReadOpsPerSecondMaximum, UtilizationMetricsVolumeWriteOpsPerSecondMaximum, UtilizationMetricsVolumeReadBytesPerSecondMaximum, UtilizationMetricsVolumeWriteBytesPerSecondMaximum, LookbackPeriodInDays, CurrentConfigurationVolumeType, CurrentConfigurationVolumeBaselineIOPS, CurrentConfigurationVolumeBaselineThroughput, CurrentConfigurationVolumeBurstIOPS, CurrentConfigurationVolumeBurstThroughput, CurrentConfigurationVolumeSize, CurrentMonthlyPrice, RecommendationOptionsConfigurationVolumeType, RecommendationOptionsConfigurationVolumeBaselineIOPS, RecommendationOptionsConfigurationVolumeBaselineThroughput, RecommendationOptionsConfigurationVolumeBurstIOPS, RecommendationOptionsConfigurationVolumeBurstThroughput, RecommendationOptionsConfigurationVolumeSize, RecommendationOptionsMonthlyPrice, RecommendationOptionsPerformanceRisk, LastRefreshTimestamp, CurrentPerformanceRisk, RecommendationOptionsSavingsOpportunityPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrency, RecommendationOptionsEstimatedMonthlySavingsValue, RootVolume, Tags, CurrentConfigurationRootVolume, EffectiveRecommendationPreferencesSavingsEstimationMode, RecommendationOptionsSavingsOpportunityAfterDiscountsPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrencyAfterDiscounts, RecommendationOptionsEstimatedMonthlySavingsValueAfterDiscounts
# s3_destination_config: { # required
# bucket: "DestinationBucket",
# key_prefix: "DestinationKeyPrefix",
# },
# file_format: "Csv", # accepts Csv
# include_member_accounts: false,
# })
#
# @example Response structure
#
# resp.job_id #=> String
# resp.s3_destination.bucket #=> String
# resp.s3_destination.key #=> String
# resp.s3_destination.metadata_key #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/ExportEBSVolumeRecommendations AWS API Documentation
#
# @overload export_ebs_volume_recommendations(params = {})
# @param [Hash] params ({})
def export_ebs_volume_recommendations(params = {}, options = {})
req = build_request(:export_ebs_volume_recommendations, params)
req.send_request(options)
end
# Exports optimization recommendations for Amazon EC2 instances.
#
# Recommendations are exported in a comma-separated values (.csv) file,
# and its metadata in a JavaScript Object Notation (JSON) (.json) file,
# to an existing Amazon Simple Storage Service (Amazon S3) bucket that
# you specify. For more information, see [Exporting Recommendations][1]
# in the *Compute Optimizer User Guide*.
#
# You can have only one Amazon EC2 instance export job in progress per
# Amazon Web Services Region.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html
#
# @option params [Array] :account_ids
# The IDs of the Amazon Web Services accounts for which to export
# instance recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to export
# recommendations.
#
# This parameter cannot be specified together with the include member
# accounts parameter. The parameters are mutually exclusive.
#
# Recommendations for member accounts are not included in the export if
# this parameter, or the include member accounts parameter, is omitted.
#
# You can specify multiple account IDs per request.
#
# @option params [Array] :filters
# An array of objects to specify a filter that exports a more specific
# set of instance recommendations.
#
# @option params [Array] :fields_to_export
# The recommendations data to include in the export file. For more
# information about the fields that can be exported, see [Exported
# files][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html#exported-files
#
# @option params [required, Types::S3DestinationConfig] :s3_destination_config
# An object to specify the destination Amazon Simple Storage Service
# (Amazon S3) bucket name and key prefix for the export job.
#
# You must create the destination Amazon S3 bucket for your
# recommendations export before you create the export job. Compute
# Optimizer does not create the S3 bucket for you. After you create the
# S3 bucket, ensure that it has the required permissions policy to allow
# Compute Optimizer to write the export file to it. If you plan to
# specify an object prefix when you create the export job, you must
# include the object prefix in the policy that you add to the S3 bucket.
# For more information, see [Amazon S3 Bucket Policy for Compute
# Optimizer][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/create-s3-bucket-policy-for-compute-optimizer.html
#
# @option params [String] :file_format
# The format of the export file.
#
# The only export file format currently supported is `Csv`.
#
# @option params [Boolean] :include_member_accounts
# Indicates whether to include recommendations for resources in all
# member accounts of the organization if your account is the management
# account of an organization.
#
# The member accounts must also be opted in to Compute Optimizer, and
# trusted access for Compute Optimizer must be enabled in the
# organization account. For more information, see [Compute Optimizer and
# Amazon Web Services Organizations trusted access][1] in the *Compute
# Optimizer User Guide*.
#
# Recommendations for member accounts of the organization are not
# included in the export file if this parameter is omitted.
#
# Recommendations for member accounts are not included in the export if
# this parameter, or the account IDs parameter, is omitted.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/security-iam.html#trusted-service-access
#
# @option params [Types::RecommendationPreferences] :recommendation_preferences
# An object to specify the preferences for the Amazon EC2 instance
# recommendations to export.
#
# @return [Types::ExportEC2InstanceRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::ExportEC2InstanceRecommendationsResponse#job_id #job_id} => String
# * {Types::ExportEC2InstanceRecommendationsResponse#s3_destination #s3_destination} => Types::S3Destination
#
# @example Request syntax with placeholder values
#
# resp = client.export_ec2_instance_recommendations({
# account_ids: ["AccountId"],
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCodes, RecommendationSourceType, InferredWorkloadTypes
# values: ["FilterValue"],
# },
# ],
# fields_to_export: ["AccountId"], # accepts AccountId, InstanceArn, InstanceName, Finding, FindingReasonCodes, LookbackPeriodInDays, CurrentInstanceType, UtilizationMetricsCpuMaximum, UtilizationMetricsMemoryMaximum, UtilizationMetricsEbsReadOpsPerSecondMaximum, UtilizationMetricsEbsWriteOpsPerSecondMaximum, UtilizationMetricsEbsReadBytesPerSecondMaximum, UtilizationMetricsEbsWriteBytesPerSecondMaximum, UtilizationMetricsDiskReadOpsPerSecondMaximum, UtilizationMetricsDiskWriteOpsPerSecondMaximum, UtilizationMetricsDiskReadBytesPerSecondMaximum, UtilizationMetricsDiskWriteBytesPerSecondMaximum, UtilizationMetricsNetworkInBytesPerSecondMaximum, UtilizationMetricsNetworkOutBytesPerSecondMaximum, UtilizationMetricsNetworkPacketsInPerSecondMaximum, UtilizationMetricsNetworkPacketsOutPerSecondMaximum, CurrentOnDemandPrice, CurrentStandardOneYearNoUpfrontReservedPrice, CurrentStandardThreeYearNoUpfrontReservedPrice, CurrentVCpus, CurrentMemory, CurrentStorage, CurrentNetwork, RecommendationOptionsInstanceType, RecommendationOptionsProjectedUtilizationMetricsCpuMaximum, RecommendationOptionsProjectedUtilizationMetricsMemoryMaximum, RecommendationOptionsPlatformDifferences, RecommendationOptionsPerformanceRisk, RecommendationOptionsVcpus, RecommendationOptionsMemory, RecommendationOptionsStorage, RecommendationOptionsNetwork, RecommendationOptionsOnDemandPrice, RecommendationOptionsStandardOneYearNoUpfrontReservedPrice, RecommendationOptionsStandardThreeYearNoUpfrontReservedPrice, RecommendationsSourcesRecommendationSourceArn, RecommendationsSourcesRecommendationSourceType, LastRefreshTimestamp, CurrentPerformanceRisk, RecommendationOptionsSavingsOpportunityPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrency, RecommendationOptionsEstimatedMonthlySavingsValue, EffectiveRecommendationPreferencesCpuVendorArchitectures, EffectiveRecommendationPreferencesEnhancedInfrastructureMetrics, EffectiveRecommendationPreferencesInferredWorkloadTypes, InferredWorkloadTypes, RecommendationOptionsMigrationEffort, EffectiveRecommendationPreferencesExternalMetricsSource, InstanceState, Tags, ExternalMetricStatusCode, ExternalMetricStatusReason, CurrentInstanceGpuInfo, RecommendationOptionsInstanceGpuInfo, UtilizationMetricsGpuPercentageMaximum, UtilizationMetricsGpuMemoryPercentageMaximum, RecommendationOptionsProjectedUtilizationMetricsGpuPercentageMaximum, RecommendationOptionsProjectedUtilizationMetricsGpuMemoryPercentageMaximum, Idle, EffectiveRecommendationPreferencesPreferredResources, EffectiveRecommendationPreferencesLookBackPeriod, EffectiveRecommendationPreferencesUtilizationPreferences, EffectiveRecommendationPreferencesSavingsEstimationMode, RecommendationOptionsSavingsOpportunityAfterDiscountsPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrencyAfterDiscounts, RecommendationOptionsEstimatedMonthlySavingsValueAfterDiscounts
# s3_destination_config: { # required
# bucket: "DestinationBucket",
# key_prefix: "DestinationKeyPrefix",
# },
# file_format: "Csv", # accepts Csv
# include_member_accounts: false,
# recommendation_preferences: {
# cpu_vendor_architectures: ["AWS_ARM64"], # accepts AWS_ARM64, CURRENT
# },
# })
#
# @example Response structure
#
# resp.job_id #=> String
# resp.s3_destination.bucket #=> String
# resp.s3_destination.key #=> String
# resp.s3_destination.metadata_key #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/ExportEC2InstanceRecommendations AWS API Documentation
#
# @overload export_ec2_instance_recommendations(params = {})
# @param [Hash] params ({})
def export_ec2_instance_recommendations(params = {}, options = {})
req = build_request(:export_ec2_instance_recommendations, params)
req.send_request(options)
end
# Exports optimization recommendations for Amazon ECS services on
# Fargate.
#
# Recommendations are exported in a CSV file, and its metadata in a JSON
# file, to an existing Amazon Simple Storage Service (Amazon S3) bucket
# that you specify. For more information, see [Exporting
# Recommendations][1] in the *Compute Optimizer User Guide*.
#
# You can only have one Amazon ECS service export job in progress per
# Amazon Web Services Region.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html
#
# @option params [Array] :account_ids
# The Amazon Web Services account IDs for the export Amazon ECS service
# recommendations.
#
# If your account is the management account or the delegated
# administrator of an organization, use this parameter to specify the
# member account you want to export recommendations to.
#
# This parameter can't be specified together with the include member
# accounts parameter. The parameters are mutually exclusive.
#
# If this parameter or the include member accounts parameter is omitted,
# the recommendations for member accounts aren't included in the
# export.
#
# You can specify multiple account IDs per request.
#
# @option params [Array] :filters
# An array of objects to specify a filter that exports a more specific
# set of Amazon ECS service recommendations.
#
# @option params [Array] :fields_to_export
# The recommendations data to include in the export file. For more
# information about the fields that can be exported, see [Exported
# files][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html#exported-files
#
# @option params [required, Types::S3DestinationConfig] :s3_destination_config
# Describes the destination Amazon Simple Storage Service (Amazon S3)
# bucket name and key prefix for a recommendations export job.
#
# You must create the destination Amazon S3 bucket for your
# recommendations export before you create the export job. Compute
# Optimizer does not create the S3 bucket for you. After you create the
# S3 bucket, ensure that it has the required permission policy to allow
# Compute Optimizer to write the export file to it. If you plan to
# specify an object prefix when you create the export job, you must
# include the object prefix in the policy that you add to the S3 bucket.
# For more information, see [Amazon S3 Bucket Policy for Compute
# Optimizer][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/create-s3-bucket-policy-for-compute-optimizer.html
#
# @option params [String] :file_format
# The format of the export file.
#
# The CSV file is the only export file format currently supported.
#
# @option params [Boolean] :include_member_accounts
# If your account is the management account or the delegated
# administrator of an organization, this parameter indicates whether to
# include recommendations for resources in all member accounts of the
# organization.
#
# The member accounts must also be opted in to Compute Optimizer, and
# trusted access for Compute Optimizer must be enabled in the
# organization account. For more information, see [Compute Optimizer and
# Amazon Web Services Organizations trusted access][1] in the *Compute
# Optimizer User Guide*.
#
# If this parameter is omitted, recommendations for member accounts of
# the organization aren't included in the export file.
#
# If this parameter or the account ID parameter is omitted,
# recommendations for member accounts aren't included in the export.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/security-iam.html#trusted-service-access
#
# @return [Types::ExportECSServiceRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::ExportECSServiceRecommendationsResponse#job_id #job_id} => String
# * {Types::ExportECSServiceRecommendationsResponse#s3_destination #s3_destination} => Types::S3Destination
#
# @example Request syntax with placeholder values
#
# resp = client.export_ecs_service_recommendations({
# account_ids: ["AccountId"],
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCode
# values: ["FilterValue"],
# },
# ],
# fields_to_export: ["AccountId"], # accepts AccountId, ServiceArn, LookbackPeriodInDays, LastRefreshTimestamp, LaunchType, CurrentPerformanceRisk, CurrentServiceConfigurationMemory, CurrentServiceConfigurationCpu, CurrentServiceConfigurationTaskDefinitionArn, CurrentServiceConfigurationAutoScalingConfiguration, CurrentServiceContainerConfigurations, UtilizationMetricsCpuMaximum, UtilizationMetricsMemoryMaximum, Finding, FindingReasonCodes, RecommendationOptionsMemory, RecommendationOptionsCpu, RecommendationOptionsSavingsOpportunityPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrency, RecommendationOptionsEstimatedMonthlySavingsValue, RecommendationOptionsContainerRecommendations, RecommendationOptionsProjectedUtilizationMetricsCpuMaximum, RecommendationOptionsProjectedUtilizationMetricsMemoryMaximum, Tags, EffectiveRecommendationPreferencesSavingsEstimationMode, RecommendationOptionsSavingsOpportunityAfterDiscountsPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrencyAfterDiscounts, RecommendationOptionsEstimatedMonthlySavingsValueAfterDiscounts
# s3_destination_config: { # required
# bucket: "DestinationBucket",
# key_prefix: "DestinationKeyPrefix",
# },
# file_format: "Csv", # accepts Csv
# include_member_accounts: false,
# })
#
# @example Response structure
#
# resp.job_id #=> String
# resp.s3_destination.bucket #=> String
# resp.s3_destination.key #=> String
# resp.s3_destination.metadata_key #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/ExportECSServiceRecommendations AWS API Documentation
#
# @overload export_ecs_service_recommendations(params = {})
# @param [Hash] params ({})
def export_ecs_service_recommendations(params = {}, options = {})
req = build_request(:export_ecs_service_recommendations, params)
req.send_request(options)
end
# Exports optimization recommendations for Lambda functions.
#
# Recommendations are exported in a comma-separated values (.csv) file,
# and its metadata in a JavaScript Object Notation (JSON) (.json) file,
# to an existing Amazon Simple Storage Service (Amazon S3) bucket that
# you specify. For more information, see [Exporting Recommendations][1]
# in the *Compute Optimizer User Guide*.
#
# You can have only one Lambda function export job in progress per
# Amazon Web Services Region.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html
#
# @option params [Array] :account_ids
# The IDs of the Amazon Web Services accounts for which to export Lambda
# function recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to export
# recommendations.
#
# This parameter cannot be specified together with the include member
# accounts parameter. The parameters are mutually exclusive.
#
# Recommendations for member accounts are not included in the export if
# this parameter, or the include member accounts parameter, is omitted.
#
# You can specify multiple account IDs per request.
#
# @option params [Array] :filters
# An array of objects to specify a filter that exports a more specific
# set of Lambda function recommendations.
#
# @option params [Array] :fields_to_export
# The recommendations data to include in the export file. For more
# information about the fields that can be exported, see [Exported
# files][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html#exported-files
#
# @option params [required, Types::S3DestinationConfig] :s3_destination_config
# Describes the destination Amazon Simple Storage Service (Amazon S3)
# bucket name and key prefix for a recommendations export job.
#
# You must create the destination Amazon S3 bucket for your
# recommendations export before you create the export job. Compute
# Optimizer does not create the S3 bucket for you. After you create the
# S3 bucket, ensure that it has the required permission policy to allow
# Compute Optimizer to write the export file to it. If you plan to
# specify an object prefix when you create the export job, you must
# include the object prefix in the policy that you add to the S3 bucket.
# For more information, see [Amazon S3 Bucket Policy for Compute
# Optimizer][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/create-s3-bucket-policy-for-compute-optimizer.html
#
# @option params [String] :file_format
# The format of the export file.
#
# The only export file format currently supported is `Csv`.
#
# @option params [Boolean] :include_member_accounts
# Indicates whether to include recommendations for resources in all
# member accounts of the organization if your account is the management
# account of an organization.
#
# The member accounts must also be opted in to Compute Optimizer, and
# trusted access for Compute Optimizer must be enabled in the
# organization account. For more information, see [Compute Optimizer and
# Amazon Web Services Organizations trusted access][1] in the *Compute
# Optimizer User Guide*.
#
# Recommendations for member accounts of the organization are not
# included in the export file if this parameter is omitted.
#
# This parameter cannot be specified together with the account IDs
# parameter. The parameters are mutually exclusive.
#
# Recommendations for member accounts are not included in the export if
# this parameter, or the account IDs parameter, is omitted.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/security-iam.html#trusted-service-access
#
# @return [Types::ExportLambdaFunctionRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::ExportLambdaFunctionRecommendationsResponse#job_id #job_id} => String
# * {Types::ExportLambdaFunctionRecommendationsResponse#s3_destination #s3_destination} => Types::S3Destination
#
# @example Request syntax with placeholder values
#
# resp = client.export_lambda_function_recommendations({
# account_ids: ["AccountId"],
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCode
# values: ["FilterValue"],
# },
# ],
# fields_to_export: ["AccountId"], # accepts AccountId, FunctionArn, FunctionVersion, Finding, FindingReasonCodes, NumberOfInvocations, UtilizationMetricsDurationMaximum, UtilizationMetricsDurationAverage, UtilizationMetricsMemoryMaximum, UtilizationMetricsMemoryAverage, LookbackPeriodInDays, CurrentConfigurationMemorySize, CurrentConfigurationTimeout, CurrentCostTotal, CurrentCostAverage, RecommendationOptionsConfigurationMemorySize, RecommendationOptionsCostLow, RecommendationOptionsCostHigh, RecommendationOptionsProjectedUtilizationMetricsDurationLowerBound, RecommendationOptionsProjectedUtilizationMetricsDurationUpperBound, RecommendationOptionsProjectedUtilizationMetricsDurationExpected, LastRefreshTimestamp, CurrentPerformanceRisk, RecommendationOptionsSavingsOpportunityPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrency, RecommendationOptionsEstimatedMonthlySavingsValue, Tags, EffectiveRecommendationPreferencesSavingsEstimationMode, RecommendationOptionsSavingsOpportunityAfterDiscountsPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrencyAfterDiscounts, RecommendationOptionsEstimatedMonthlySavingsValueAfterDiscounts
# s3_destination_config: { # required
# bucket: "DestinationBucket",
# key_prefix: "DestinationKeyPrefix",
# },
# file_format: "Csv", # accepts Csv
# include_member_accounts: false,
# })
#
# @example Response structure
#
# resp.job_id #=> String
# resp.s3_destination.bucket #=> String
# resp.s3_destination.key #=> String
# resp.s3_destination.metadata_key #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/ExportLambdaFunctionRecommendations AWS API Documentation
#
# @overload export_lambda_function_recommendations(params = {})
# @param [Hash] params ({})
def export_lambda_function_recommendations(params = {}, options = {})
req = build_request(:export_lambda_function_recommendations, params)
req.send_request(options)
end
# Export optimization recommendations for your licenses.
#
# Recommendations are exported in a comma-separated values (CSV) file,
# and its metadata in a JavaScript Object Notation (JSON) file, to an
# existing Amazon Simple Storage Service (Amazon S3) bucket that you
# specify. For more information, see [Exporting Recommendations][1] in
# the *Compute Optimizer User Guide*.
#
# You can have only one license export job in progress per Amazon Web
# Services Region.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html
#
# @option params [Array] :account_ids
# The IDs of the Amazon Web Services accounts for which to export
# license recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to export
# recommendations.
#
# This parameter can't be specified together with the include member
# accounts parameter. The parameters are mutually exclusive.
#
# If this parameter is omitted, recommendations for member accounts
# aren't included in the export.
#
# You can specify multiple account IDs per request.
#
# @option params [Array] :filters
# An array of objects to specify a filter that exports a more specific
# set of license recommendations.
#
# @option params [Array] :fields_to_export
# The recommendations data to include in the export file. For more
# information about the fields that can be exported, see [Exported
# files][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/exporting-recommendations.html#exported-files
#
# @option params [required, Types::S3DestinationConfig] :s3_destination_config
# Describes the destination Amazon Simple Storage Service (Amazon S3)
# bucket name and key prefix for a recommendations export job.
#
# You must create the destination Amazon S3 bucket for your
# recommendations export before you create the export job. Compute
# Optimizer does not create the S3 bucket for you. After you create the
# S3 bucket, ensure that it has the required permission policy to allow
# Compute Optimizer to write the export file to it. If you plan to
# specify an object prefix when you create the export job, you must
# include the object prefix in the policy that you add to the S3 bucket.
# For more information, see [Amazon S3 Bucket Policy for Compute
# Optimizer][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/create-s3-bucket-policy-for-compute-optimizer.html
#
# @option params [String] :file_format
# The format of the export file.
#
# A CSV file is the only export format currently supported.
#
# @option params [Boolean] :include_member_accounts
# Indicates whether to include recommendations for resources in all
# member accounts of the organization if your account is the management
# account of an organization.
#
# The member accounts must also be opted in to Compute Optimizer, and
# trusted access for Compute Optimizer must be enabled in the
# organization account. For more information, see [Compute Optimizer and
# Amazon Web Services Organizations trusted access][1] in the *Compute
# Optimizer User Guide*.
#
# If this parameter is omitted, recommendations for member accounts of
# the organization aren't included in the export file .
#
# This parameter cannot be specified together with the account IDs
# parameter. The parameters are mutually exclusive.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/security-iam.html#trusted-service-access
#
# @return [Types::ExportLicenseRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::ExportLicenseRecommendationsResponse#job_id #job_id} => String
# * {Types::ExportLicenseRecommendationsResponse#s3_destination #s3_destination} => Types::S3Destination
#
# @example Request syntax with placeholder values
#
# resp = client.export_license_recommendations({
# account_ids: ["AccountId"],
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCode, LicenseName
# values: ["FilterValue"],
# },
# ],
# fields_to_export: ["AccountId"], # accepts AccountId, ResourceArn, LookbackPeriodInDays, LastRefreshTimestamp, Finding, FindingReasonCodes, CurrentLicenseConfigurationNumberOfCores, CurrentLicenseConfigurationInstanceType, CurrentLicenseConfigurationOperatingSystem, CurrentLicenseConfigurationLicenseName, CurrentLicenseConfigurationLicenseEdition, CurrentLicenseConfigurationLicenseModel, CurrentLicenseConfigurationLicenseVersion, CurrentLicenseConfigurationMetricsSource, RecommendationOptionsOperatingSystem, RecommendationOptionsLicenseEdition, RecommendationOptionsLicenseModel, RecommendationOptionsSavingsOpportunityPercentage, RecommendationOptionsEstimatedMonthlySavingsCurrency, RecommendationOptionsEstimatedMonthlySavingsValue, Tags
# s3_destination_config: { # required
# bucket: "DestinationBucket",
# key_prefix: "DestinationKeyPrefix",
# },
# file_format: "Csv", # accepts Csv
# include_member_accounts: false,
# })
#
# @example Response structure
#
# resp.job_id #=> String
# resp.s3_destination.bucket #=> String
# resp.s3_destination.key #=> String
# resp.s3_destination.metadata_key #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/ExportLicenseRecommendations AWS API Documentation
#
# @overload export_license_recommendations(params = {})
# @param [Hash] params ({})
def export_license_recommendations(params = {}, options = {})
req = build_request(:export_license_recommendations, params)
req.send_request(options)
end
# Returns Auto Scaling group recommendations.
#
# Compute Optimizer generates recommendations for Amazon EC2 Auto
# Scaling groups that meet a specific set of requirements. For more
# information, see the [Supported resources and requirements][1] in the
# *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/requirements.html
#
# @option params [Array] :account_ids
# The ID of the Amazon Web Services account for which to return Auto
# Scaling group recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to return
# Auto Scaling group recommendations.
#
# Only one account ID can be specified per request.
#
# @option params [Array] :auto_scaling_group_arns
# The Amazon Resource Name (ARN) of the Auto Scaling groups for which to
# return recommendations.
#
# @option params [String] :next_token
# The token to advance to the next page of Auto Scaling group
# recommendations.
#
# @option params [Integer] :max_results
# The maximum number of Auto Scaling group recommendations to return
# with a single request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @option params [Array] :filters
# An array of objects to specify a filter that returns a more specific
# list of Auto Scaling group recommendations.
#
# @option params [Types::RecommendationPreferences] :recommendation_preferences
# An object to specify the preferences for the Auto Scaling group
# recommendations to return in the response.
#
# @return [Types::GetAutoScalingGroupRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetAutoScalingGroupRecommendationsResponse#next_token #next_token} => String
# * {Types::GetAutoScalingGroupRecommendationsResponse#auto_scaling_group_recommendations #auto_scaling_group_recommendations} => Array<Types::AutoScalingGroupRecommendation>
# * {Types::GetAutoScalingGroupRecommendationsResponse#errors #errors} => Array<Types::GetRecommendationError>
#
# @example Request syntax with placeholder values
#
# resp = client.get_auto_scaling_group_recommendations({
# account_ids: ["AccountId"],
# auto_scaling_group_arns: ["AutoScalingGroupArn"],
# next_token: "NextToken",
# max_results: 1,
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCodes, RecommendationSourceType, InferredWorkloadTypes
# values: ["FilterValue"],
# },
# ],
# recommendation_preferences: {
# cpu_vendor_architectures: ["AWS_ARM64"], # accepts AWS_ARM64, CURRENT
# },
# })
#
# @example Response structure
#
# resp.next_token #=> String
# resp.auto_scaling_group_recommendations #=> Array
# resp.auto_scaling_group_recommendations[0].account_id #=> String
# resp.auto_scaling_group_recommendations[0].auto_scaling_group_arn #=> String
# resp.auto_scaling_group_recommendations[0].auto_scaling_group_name #=> String
# resp.auto_scaling_group_recommendations[0].finding #=> String, one of "Underprovisioned", "Overprovisioned", "Optimized", "NotOptimized"
# resp.auto_scaling_group_recommendations[0].utilization_metrics #=> Array
# resp.auto_scaling_group_recommendations[0].utilization_metrics[0].name #=> String, one of "Cpu", "Memory", "EBS_READ_OPS_PER_SECOND", "EBS_WRITE_OPS_PER_SECOND", "EBS_READ_BYTES_PER_SECOND", "EBS_WRITE_BYTES_PER_SECOND", "DISK_READ_OPS_PER_SECOND", "DISK_WRITE_OPS_PER_SECOND", "DISK_READ_BYTES_PER_SECOND", "DISK_WRITE_BYTES_PER_SECOND", "NETWORK_IN_BYTES_PER_SECOND", "NETWORK_OUT_BYTES_PER_SECOND", "NETWORK_PACKETS_IN_PER_SECOND", "NETWORK_PACKETS_OUT_PER_SECOND", "GPU_PERCENTAGE", "GPU_MEMORY_PERCENTAGE"
# resp.auto_scaling_group_recommendations[0].utilization_metrics[0].statistic #=> String, one of "Maximum", "Average"
# resp.auto_scaling_group_recommendations[0].utilization_metrics[0].value #=> Float
# resp.auto_scaling_group_recommendations[0].look_back_period_in_days #=> Float
# resp.auto_scaling_group_recommendations[0].current_configuration.desired_capacity #=> Integer
# resp.auto_scaling_group_recommendations[0].current_configuration.min_size #=> Integer
# resp.auto_scaling_group_recommendations[0].current_configuration.max_size #=> Integer
# resp.auto_scaling_group_recommendations[0].current_configuration.instance_type #=> String
# resp.auto_scaling_group_recommendations[0].recommendation_options #=> Array
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].configuration.desired_capacity #=> Integer
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].configuration.min_size #=> Integer
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].configuration.max_size #=> Integer
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].configuration.instance_type #=> String
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].projected_utilization_metrics #=> Array
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].projected_utilization_metrics[0].name #=> String, one of "Cpu", "Memory", "EBS_READ_OPS_PER_SECOND", "EBS_WRITE_OPS_PER_SECOND", "EBS_READ_BYTES_PER_SECOND", "EBS_WRITE_BYTES_PER_SECOND", "DISK_READ_OPS_PER_SECOND", "DISK_WRITE_OPS_PER_SECOND", "DISK_READ_BYTES_PER_SECOND", "DISK_WRITE_BYTES_PER_SECOND", "NETWORK_IN_BYTES_PER_SECOND", "NETWORK_OUT_BYTES_PER_SECOND", "NETWORK_PACKETS_IN_PER_SECOND", "NETWORK_PACKETS_OUT_PER_SECOND", "GPU_PERCENTAGE", "GPU_MEMORY_PERCENTAGE"
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].projected_utilization_metrics[0].statistic #=> String, one of "Maximum", "Average"
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].projected_utilization_metrics[0].value #=> Float
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].performance_risk #=> Float
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].rank #=> Integer
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].savings_opportunity.savings_opportunity_percentage #=> Float
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].savings_opportunity.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].savings_opportunity.estimated_monthly_savings.value #=> Float
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].migration_effort #=> String, one of "VeryLow", "Low", "Medium", "High"
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].instance_gpu_info.gpus #=> Array
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].instance_gpu_info.gpus[0].gpu_count #=> Integer
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].instance_gpu_info.gpus[0].gpu_memory_size_in_mi_b #=> Integer
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].savings_opportunity_after_discounts.savings_opportunity_percentage #=> Float
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.auto_scaling_group_recommendations[0].recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.value #=> Float
# resp.auto_scaling_group_recommendations[0].last_refresh_timestamp #=> Time
# resp.auto_scaling_group_recommendations[0].current_performance_risk #=> String, one of "VeryLow", "Low", "Medium", "High"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.cpu_vendor_architectures #=> Array
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.cpu_vendor_architectures[0] #=> String, one of "AWS_ARM64", "CURRENT"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.enhanced_infrastructure_metrics #=> String, one of "Active", "Inactive"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.inferred_workload_types #=> String, one of "Active", "Inactive"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.external_metrics_preference.source #=> String, one of "Datadog", "Dynatrace", "NewRelic", "Instana"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.look_back_period #=> String, one of "DAYS_14", "DAYS_32", "DAYS_93"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.utilization_preferences #=> Array
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.utilization_preferences[0].metric_name #=> String, one of "CpuUtilization", "MemoryUtilization"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.utilization_preferences[0].metric_parameters.threshold #=> String, one of "P90", "P95", "P99_5"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.utilization_preferences[0].metric_parameters.headroom #=> String, one of "PERCENT_30", "PERCENT_20", "PERCENT_10", "PERCENT_0"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.preferred_resources #=> Array
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.preferred_resources[0].name #=> String, one of "Ec2InstanceTypes"
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.preferred_resources[0].include_list #=> Array
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.preferred_resources[0].include_list[0] #=> String
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.preferred_resources[0].effective_include_list #=> Array
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.preferred_resources[0].effective_include_list[0] #=> String
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.preferred_resources[0].exclude_list #=> Array
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.preferred_resources[0].exclude_list[0] #=> String
# resp.auto_scaling_group_recommendations[0].effective_recommendation_preferences.savings_estimation_mode.source #=> String, one of "PublicPricing", "CostExplorerRightsizing", "CostOptimizationHub"
# resp.auto_scaling_group_recommendations[0].inferred_workload_types #=> Array
# resp.auto_scaling_group_recommendations[0].inferred_workload_types[0] #=> String, one of "AmazonEmr", "ApacheCassandra", "ApacheHadoop", "Memcached", "Nginx", "PostgreSql", "Redis", "Kafka", "SQLServer"
# resp.auto_scaling_group_recommendations[0].current_instance_gpu_info.gpus #=> Array
# resp.auto_scaling_group_recommendations[0].current_instance_gpu_info.gpus[0].gpu_count #=> Integer
# resp.auto_scaling_group_recommendations[0].current_instance_gpu_info.gpus[0].gpu_memory_size_in_mi_b #=> Integer
# resp.errors #=> Array
# resp.errors[0].identifier #=> String
# resp.errors[0].code #=> String
# resp.errors[0].message #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetAutoScalingGroupRecommendations AWS API Documentation
#
# @overload get_auto_scaling_group_recommendations(params = {})
# @param [Hash] params ({})
def get_auto_scaling_group_recommendations(params = {}, options = {})
req = build_request(:get_auto_scaling_group_recommendations, params)
req.send_request(options)
end
# Returns Amazon Elastic Block Store (Amazon EBS) volume
# recommendations.
#
# Compute Optimizer generates recommendations for Amazon EBS volumes
# that meet a specific set of requirements. For more information, see
# the [Supported resources and requirements][1] in the *Compute
# Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/requirements.html
#
# @option params [Array] :volume_arns
# The Amazon Resource Name (ARN) of the volumes for which to return
# recommendations.
#
# @option params [String] :next_token
# The token to advance to the next page of volume recommendations.
#
# @option params [Integer] :max_results
# The maximum number of volume recommendations to return with a single
# request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @option params [Array] :filters
# An array of objects to specify a filter that returns a more specific
# list of volume recommendations.
#
# @option params [Array] :account_ids
# The ID of the Amazon Web Services account for which to return volume
# recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to return
# volume recommendations.
#
# Only one account ID can be specified per request.
#
# @return [Types::GetEBSVolumeRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetEBSVolumeRecommendationsResponse#next_token #next_token} => String
# * {Types::GetEBSVolumeRecommendationsResponse#volume_recommendations #volume_recommendations} => Array<Types::VolumeRecommendation>
# * {Types::GetEBSVolumeRecommendationsResponse#errors #errors} => Array<Types::GetRecommendationError>
#
# @example Request syntax with placeholder values
#
# resp = client.get_ebs_volume_recommendations({
# volume_arns: ["VolumeArn"],
# next_token: "NextToken",
# max_results: 1,
# filters: [
# {
# name: "Finding", # accepts Finding
# values: ["FilterValue"],
# },
# ],
# account_ids: ["AccountId"],
# })
#
# @example Response structure
#
# resp.next_token #=> String
# resp.volume_recommendations #=> Array
# resp.volume_recommendations[0].volume_arn #=> String
# resp.volume_recommendations[0].account_id #=> String
# resp.volume_recommendations[0].current_configuration.volume_type #=> String
# resp.volume_recommendations[0].current_configuration.volume_size #=> Integer
# resp.volume_recommendations[0].current_configuration.volume_baseline_iops #=> Integer
# resp.volume_recommendations[0].current_configuration.volume_burst_iops #=> Integer
# resp.volume_recommendations[0].current_configuration.volume_baseline_throughput #=> Integer
# resp.volume_recommendations[0].current_configuration.volume_burst_throughput #=> Integer
# resp.volume_recommendations[0].current_configuration.root_volume #=> Boolean
# resp.volume_recommendations[0].finding #=> String, one of "Optimized", "NotOptimized"
# resp.volume_recommendations[0].utilization_metrics #=> Array
# resp.volume_recommendations[0].utilization_metrics[0].name #=> String, one of "VolumeReadOpsPerSecond", "VolumeWriteOpsPerSecond", "VolumeReadBytesPerSecond", "VolumeWriteBytesPerSecond"
# resp.volume_recommendations[0].utilization_metrics[0].statistic #=> String, one of "Maximum", "Average"
# resp.volume_recommendations[0].utilization_metrics[0].value #=> Float
# resp.volume_recommendations[0].look_back_period_in_days #=> Float
# resp.volume_recommendations[0].volume_recommendation_options #=> Array
# resp.volume_recommendations[0].volume_recommendation_options[0].configuration.volume_type #=> String
# resp.volume_recommendations[0].volume_recommendation_options[0].configuration.volume_size #=> Integer
# resp.volume_recommendations[0].volume_recommendation_options[0].configuration.volume_baseline_iops #=> Integer
# resp.volume_recommendations[0].volume_recommendation_options[0].configuration.volume_burst_iops #=> Integer
# resp.volume_recommendations[0].volume_recommendation_options[0].configuration.volume_baseline_throughput #=> Integer
# resp.volume_recommendations[0].volume_recommendation_options[0].configuration.volume_burst_throughput #=> Integer
# resp.volume_recommendations[0].volume_recommendation_options[0].configuration.root_volume #=> Boolean
# resp.volume_recommendations[0].volume_recommendation_options[0].performance_risk #=> Float
# resp.volume_recommendations[0].volume_recommendation_options[0].rank #=> Integer
# resp.volume_recommendations[0].volume_recommendation_options[0].savings_opportunity.savings_opportunity_percentage #=> Float
# resp.volume_recommendations[0].volume_recommendation_options[0].savings_opportunity.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.volume_recommendations[0].volume_recommendation_options[0].savings_opportunity.estimated_monthly_savings.value #=> Float
# resp.volume_recommendations[0].volume_recommendation_options[0].savings_opportunity_after_discounts.savings_opportunity_percentage #=> Float
# resp.volume_recommendations[0].volume_recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.volume_recommendations[0].volume_recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.value #=> Float
# resp.volume_recommendations[0].last_refresh_timestamp #=> Time
# resp.volume_recommendations[0].current_performance_risk #=> String, one of "VeryLow", "Low", "Medium", "High"
# resp.volume_recommendations[0].tags #=> Array
# resp.volume_recommendations[0].tags[0].key #=> String
# resp.volume_recommendations[0].tags[0].value #=> String
# resp.volume_recommendations[0].effective_recommendation_preferences.savings_estimation_mode.source #=> String, one of "PublicPricing", "CostExplorerRightsizing", "CostOptimizationHub"
# resp.errors #=> Array
# resp.errors[0].identifier #=> String
# resp.errors[0].code #=> String
# resp.errors[0].message #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetEBSVolumeRecommendations AWS API Documentation
#
# @overload get_ebs_volume_recommendations(params = {})
# @param [Hash] params ({})
def get_ebs_volume_recommendations(params = {}, options = {})
req = build_request(:get_ebs_volume_recommendations, params)
req.send_request(options)
end
# Returns Amazon EC2 instance recommendations.
#
# Compute Optimizer generates recommendations for Amazon Elastic Compute
# Cloud (Amazon EC2) instances that meet a specific set of requirements.
# For more information, see the [Supported resources and
# requirements][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/requirements.html
#
# @option params [Array] :instance_arns
# The Amazon Resource Name (ARN) of the instances for which to return
# recommendations.
#
# @option params [String] :next_token
# The token to advance to the next page of instance recommendations.
#
# @option params [Integer] :max_results
# The maximum number of instance recommendations to return with a single
# request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @option params [Array] :filters
# An array of objects to specify a filter that returns a more specific
# list of instance recommendations.
#
# @option params [Array] :account_ids
# The ID of the Amazon Web Services account for which to return instance
# recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to return
# instance recommendations.
#
# Only one account ID can be specified per request.
#
# @option params [Types::RecommendationPreferences] :recommendation_preferences
# An object to specify the preferences for the Amazon EC2 instance
# recommendations to return in the response.
#
# @return [Types::GetEC2InstanceRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetEC2InstanceRecommendationsResponse#next_token #next_token} => String
# * {Types::GetEC2InstanceRecommendationsResponse#instance_recommendations #instance_recommendations} => Array<Types::InstanceRecommendation>
# * {Types::GetEC2InstanceRecommendationsResponse#errors #errors} => Array<Types::GetRecommendationError>
#
# @example Request syntax with placeholder values
#
# resp = client.get_ec2_instance_recommendations({
# instance_arns: ["InstanceArn"],
# next_token: "NextToken",
# max_results: 1,
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCodes, RecommendationSourceType, InferredWorkloadTypes
# values: ["FilterValue"],
# },
# ],
# account_ids: ["AccountId"],
# recommendation_preferences: {
# cpu_vendor_architectures: ["AWS_ARM64"], # accepts AWS_ARM64, CURRENT
# },
# })
#
# @example Response structure
#
# resp.next_token #=> String
# resp.instance_recommendations #=> Array
# resp.instance_recommendations[0].instance_arn #=> String
# resp.instance_recommendations[0].account_id #=> String
# resp.instance_recommendations[0].instance_name #=> String
# resp.instance_recommendations[0].current_instance_type #=> String
# resp.instance_recommendations[0].finding #=> String, one of "Underprovisioned", "Overprovisioned", "Optimized", "NotOptimized"
# resp.instance_recommendations[0].finding_reason_codes #=> Array
# resp.instance_recommendations[0].finding_reason_codes[0] #=> String, one of "CPUOverprovisioned", "CPUUnderprovisioned", "MemoryOverprovisioned", "MemoryUnderprovisioned", "EBSThroughputOverprovisioned", "EBSThroughputUnderprovisioned", "EBSIOPSOverprovisioned", "EBSIOPSUnderprovisioned", "NetworkBandwidthOverprovisioned", "NetworkBandwidthUnderprovisioned", "NetworkPPSOverprovisioned", "NetworkPPSUnderprovisioned", "DiskIOPSOverprovisioned", "DiskIOPSUnderprovisioned", "DiskThroughputOverprovisioned", "DiskThroughputUnderprovisioned", "GPUUnderprovisioned", "GPUOverprovisioned", "GPUMemoryUnderprovisioned", "GPUMemoryOverprovisioned"
# resp.instance_recommendations[0].utilization_metrics #=> Array
# resp.instance_recommendations[0].utilization_metrics[0].name #=> String, one of "Cpu", "Memory", "EBS_READ_OPS_PER_SECOND", "EBS_WRITE_OPS_PER_SECOND", "EBS_READ_BYTES_PER_SECOND", "EBS_WRITE_BYTES_PER_SECOND", "DISK_READ_OPS_PER_SECOND", "DISK_WRITE_OPS_PER_SECOND", "DISK_READ_BYTES_PER_SECOND", "DISK_WRITE_BYTES_PER_SECOND", "NETWORK_IN_BYTES_PER_SECOND", "NETWORK_OUT_BYTES_PER_SECOND", "NETWORK_PACKETS_IN_PER_SECOND", "NETWORK_PACKETS_OUT_PER_SECOND", "GPU_PERCENTAGE", "GPU_MEMORY_PERCENTAGE"
# resp.instance_recommendations[0].utilization_metrics[0].statistic #=> String, one of "Maximum", "Average"
# resp.instance_recommendations[0].utilization_metrics[0].value #=> Float
# resp.instance_recommendations[0].look_back_period_in_days #=> Float
# resp.instance_recommendations[0].recommendation_options #=> Array
# resp.instance_recommendations[0].recommendation_options[0].instance_type #=> String
# resp.instance_recommendations[0].recommendation_options[0].projected_utilization_metrics #=> Array
# resp.instance_recommendations[0].recommendation_options[0].projected_utilization_metrics[0].name #=> String, one of "Cpu", "Memory", "EBS_READ_OPS_PER_SECOND", "EBS_WRITE_OPS_PER_SECOND", "EBS_READ_BYTES_PER_SECOND", "EBS_WRITE_BYTES_PER_SECOND", "DISK_READ_OPS_PER_SECOND", "DISK_WRITE_OPS_PER_SECOND", "DISK_READ_BYTES_PER_SECOND", "DISK_WRITE_BYTES_PER_SECOND", "NETWORK_IN_BYTES_PER_SECOND", "NETWORK_OUT_BYTES_PER_SECOND", "NETWORK_PACKETS_IN_PER_SECOND", "NETWORK_PACKETS_OUT_PER_SECOND", "GPU_PERCENTAGE", "GPU_MEMORY_PERCENTAGE"
# resp.instance_recommendations[0].recommendation_options[0].projected_utilization_metrics[0].statistic #=> String, one of "Maximum", "Average"
# resp.instance_recommendations[0].recommendation_options[0].projected_utilization_metrics[0].value #=> Float
# resp.instance_recommendations[0].recommendation_options[0].platform_differences #=> Array
# resp.instance_recommendations[0].recommendation_options[0].platform_differences[0] #=> String, one of "Hypervisor", "NetworkInterface", "StorageInterface", "InstanceStoreAvailability", "VirtualizationType", "Architecture"
# resp.instance_recommendations[0].recommendation_options[0].performance_risk #=> Float
# resp.instance_recommendations[0].recommendation_options[0].rank #=> Integer
# resp.instance_recommendations[0].recommendation_options[0].savings_opportunity.savings_opportunity_percentage #=> Float
# resp.instance_recommendations[0].recommendation_options[0].savings_opportunity.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.instance_recommendations[0].recommendation_options[0].savings_opportunity.estimated_monthly_savings.value #=> Float
# resp.instance_recommendations[0].recommendation_options[0].migration_effort #=> String, one of "VeryLow", "Low", "Medium", "High"
# resp.instance_recommendations[0].recommendation_options[0].instance_gpu_info.gpus #=> Array
# resp.instance_recommendations[0].recommendation_options[0].instance_gpu_info.gpus[0].gpu_count #=> Integer
# resp.instance_recommendations[0].recommendation_options[0].instance_gpu_info.gpus[0].gpu_memory_size_in_mi_b #=> Integer
# resp.instance_recommendations[0].recommendation_options[0].savings_opportunity_after_discounts.savings_opportunity_percentage #=> Float
# resp.instance_recommendations[0].recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.instance_recommendations[0].recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.value #=> Float
# resp.instance_recommendations[0].recommendation_sources #=> Array
# resp.instance_recommendations[0].recommendation_sources[0].recommendation_source_arn #=> String
# resp.instance_recommendations[0].recommendation_sources[0].recommendation_source_type #=> String, one of "Ec2Instance", "AutoScalingGroup", "EbsVolume", "LambdaFunction", "EcsService", "License"
# resp.instance_recommendations[0].last_refresh_timestamp #=> Time
# resp.instance_recommendations[0].current_performance_risk #=> String, one of "VeryLow", "Low", "Medium", "High"
# resp.instance_recommendations[0].effective_recommendation_preferences.cpu_vendor_architectures #=> Array
# resp.instance_recommendations[0].effective_recommendation_preferences.cpu_vendor_architectures[0] #=> String, one of "AWS_ARM64", "CURRENT"
# resp.instance_recommendations[0].effective_recommendation_preferences.enhanced_infrastructure_metrics #=> String, one of "Active", "Inactive"
# resp.instance_recommendations[0].effective_recommendation_preferences.inferred_workload_types #=> String, one of "Active", "Inactive"
# resp.instance_recommendations[0].effective_recommendation_preferences.external_metrics_preference.source #=> String, one of "Datadog", "Dynatrace", "NewRelic", "Instana"
# resp.instance_recommendations[0].effective_recommendation_preferences.look_back_period #=> String, one of "DAYS_14", "DAYS_32", "DAYS_93"
# resp.instance_recommendations[0].effective_recommendation_preferences.utilization_preferences #=> Array
# resp.instance_recommendations[0].effective_recommendation_preferences.utilization_preferences[0].metric_name #=> String, one of "CpuUtilization", "MemoryUtilization"
# resp.instance_recommendations[0].effective_recommendation_preferences.utilization_preferences[0].metric_parameters.threshold #=> String, one of "P90", "P95", "P99_5"
# resp.instance_recommendations[0].effective_recommendation_preferences.utilization_preferences[0].metric_parameters.headroom #=> String, one of "PERCENT_30", "PERCENT_20", "PERCENT_10", "PERCENT_0"
# resp.instance_recommendations[0].effective_recommendation_preferences.preferred_resources #=> Array
# resp.instance_recommendations[0].effective_recommendation_preferences.preferred_resources[0].name #=> String, one of "Ec2InstanceTypes"
# resp.instance_recommendations[0].effective_recommendation_preferences.preferred_resources[0].include_list #=> Array
# resp.instance_recommendations[0].effective_recommendation_preferences.preferred_resources[0].include_list[0] #=> String
# resp.instance_recommendations[0].effective_recommendation_preferences.preferred_resources[0].effective_include_list #=> Array
# resp.instance_recommendations[0].effective_recommendation_preferences.preferred_resources[0].effective_include_list[0] #=> String
# resp.instance_recommendations[0].effective_recommendation_preferences.preferred_resources[0].exclude_list #=> Array
# resp.instance_recommendations[0].effective_recommendation_preferences.preferred_resources[0].exclude_list[0] #=> String
# resp.instance_recommendations[0].effective_recommendation_preferences.savings_estimation_mode.source #=> String, one of "PublicPricing", "CostExplorerRightsizing", "CostOptimizationHub"
# resp.instance_recommendations[0].inferred_workload_types #=> Array
# resp.instance_recommendations[0].inferred_workload_types[0] #=> String, one of "AmazonEmr", "ApacheCassandra", "ApacheHadoop", "Memcached", "Nginx", "PostgreSql", "Redis", "Kafka", "SQLServer"
# resp.instance_recommendations[0].instance_state #=> String, one of "pending", "running", "shutting-down", "terminated", "stopping", "stopped"
# resp.instance_recommendations[0].tags #=> Array
# resp.instance_recommendations[0].tags[0].key #=> String
# resp.instance_recommendations[0].tags[0].value #=> String
# resp.instance_recommendations[0].external_metric_status.status_code #=> String, one of "NO_EXTERNAL_METRIC_SET", "INTEGRATION_SUCCESS", "DATADOG_INTEGRATION_ERROR", "DYNATRACE_INTEGRATION_ERROR", "NEWRELIC_INTEGRATION_ERROR", "INSTANA_INTEGRATION_ERROR", "INSUFFICIENT_DATADOG_METRICS", "INSUFFICIENT_DYNATRACE_METRICS", "INSUFFICIENT_NEWRELIC_METRICS", "INSUFFICIENT_INSTANA_METRICS"
# resp.instance_recommendations[0].external_metric_status.status_reason #=> String
# resp.instance_recommendations[0].current_instance_gpu_info.gpus #=> Array
# resp.instance_recommendations[0].current_instance_gpu_info.gpus[0].gpu_count #=> Integer
# resp.instance_recommendations[0].current_instance_gpu_info.gpus[0].gpu_memory_size_in_mi_b #=> Integer
# resp.instance_recommendations[0].idle #=> String, one of "True", "False"
# resp.errors #=> Array
# resp.errors[0].identifier #=> String
# resp.errors[0].code #=> String
# resp.errors[0].message #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetEC2InstanceRecommendations AWS API Documentation
#
# @overload get_ec2_instance_recommendations(params = {})
# @param [Hash] params ({})
def get_ec2_instance_recommendations(params = {}, options = {})
req = build_request(:get_ec2_instance_recommendations, params)
req.send_request(options)
end
# Returns the projected utilization metrics of Amazon EC2 instance
# recommendations.
#
# The `Cpu` and `Memory` metrics are the only projected utilization
# metrics returned when you run this action. Additionally, the `Memory`
# metric is returned only for resources that have the unified CloudWatch
# agent installed on them. For more information, see [Enabling Memory
# Utilization with the CloudWatch Agent][1].
#
#
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/metrics.html#cw-agent
#
# @option params [required, String] :instance_arn
# The Amazon Resource Name (ARN) of the instances for which to return
# recommendation projected metrics.
#
# @option params [required, String] :stat
# The statistic of the projected metrics.
#
# @option params [required, Integer] :period
# The granularity, in seconds, of the projected metrics data points.
#
# @option params [required, Time,DateTime,Date,Integer,String] :start_time
# The timestamp of the first projected metrics data point to return.
#
# @option params [required, Time,DateTime,Date,Integer,String] :end_time
# The timestamp of the last projected metrics data point to return.
#
# @option params [Types::RecommendationPreferences] :recommendation_preferences
# An object to specify the preferences for the Amazon EC2 recommendation
# projected metrics to return in the response.
#
# @return [Types::GetEC2RecommendationProjectedMetricsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetEC2RecommendationProjectedMetricsResponse#recommended_option_projected_metrics #recommended_option_projected_metrics} => Array<Types::RecommendedOptionProjectedMetric>
#
# @example Request syntax with placeholder values
#
# resp = client.get_ec2_recommendation_projected_metrics({
# instance_arn: "InstanceArn", # required
# stat: "Maximum", # required, accepts Maximum, Average
# period: 1, # required
# start_time: Time.now, # required
# end_time: Time.now, # required
# recommendation_preferences: {
# cpu_vendor_architectures: ["AWS_ARM64"], # accepts AWS_ARM64, CURRENT
# },
# })
#
# @example Response structure
#
# resp.recommended_option_projected_metrics #=> Array
# resp.recommended_option_projected_metrics[0].recommended_instance_type #=> String
# resp.recommended_option_projected_metrics[0].rank #=> Integer
# resp.recommended_option_projected_metrics[0].projected_metrics #=> Array
# resp.recommended_option_projected_metrics[0].projected_metrics[0].name #=> String, one of "Cpu", "Memory", "EBS_READ_OPS_PER_SECOND", "EBS_WRITE_OPS_PER_SECOND", "EBS_READ_BYTES_PER_SECOND", "EBS_WRITE_BYTES_PER_SECOND", "DISK_READ_OPS_PER_SECOND", "DISK_WRITE_OPS_PER_SECOND", "DISK_READ_BYTES_PER_SECOND", "DISK_WRITE_BYTES_PER_SECOND", "NETWORK_IN_BYTES_PER_SECOND", "NETWORK_OUT_BYTES_PER_SECOND", "NETWORK_PACKETS_IN_PER_SECOND", "NETWORK_PACKETS_OUT_PER_SECOND", "GPU_PERCENTAGE", "GPU_MEMORY_PERCENTAGE"
# resp.recommended_option_projected_metrics[0].projected_metrics[0].timestamps #=> Array
# resp.recommended_option_projected_metrics[0].projected_metrics[0].timestamps[0] #=> Time
# resp.recommended_option_projected_metrics[0].projected_metrics[0].values #=> Array
# resp.recommended_option_projected_metrics[0].projected_metrics[0].values[0] #=> Float
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetEC2RecommendationProjectedMetrics AWS API Documentation
#
# @overload get_ec2_recommendation_projected_metrics(params = {})
# @param [Hash] params ({})
def get_ec2_recommendation_projected_metrics(params = {}, options = {})
req = build_request(:get_ec2_recommendation_projected_metrics, params)
req.send_request(options)
end
# Returns the projected metrics of Amazon ECS service recommendations.
#
# @option params [required, String] :service_arn
# The ARN that identifies the Amazon ECS service.
#
# The following is the format of the ARN:
#
# `arn:aws:ecs:region:aws_account_id:service/cluster-name/service-name`
#
# @option params [required, String] :stat
# The statistic of the projected metrics.
#
# @option params [required, Integer] :period
# The granularity, in seconds, of the projected metrics data points.
#
# @option params [required, Time,DateTime,Date,Integer,String] :start_time
# The timestamp of the first projected metrics data point to return.
#
# @option params [required, Time,DateTime,Date,Integer,String] :end_time
# The timestamp of the last projected metrics data point to return.
#
# @return [Types::GetECSServiceRecommendationProjectedMetricsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetECSServiceRecommendationProjectedMetricsResponse#recommended_option_projected_metrics #recommended_option_projected_metrics} => Array<Types::ECSServiceRecommendedOptionProjectedMetric>
#
# @example Request syntax with placeholder values
#
# resp = client.get_ecs_service_recommendation_projected_metrics({
# service_arn: "ServiceArn", # required
# stat: "Maximum", # required, accepts Maximum, Average
# period: 1, # required
# start_time: Time.now, # required
# end_time: Time.now, # required
# })
#
# @example Response structure
#
# resp.recommended_option_projected_metrics #=> Array
# resp.recommended_option_projected_metrics[0].recommended_cpu_units #=> Integer
# resp.recommended_option_projected_metrics[0].recommended_memory_size #=> Integer
# resp.recommended_option_projected_metrics[0].projected_metrics #=> Array
# resp.recommended_option_projected_metrics[0].projected_metrics[0].name #=> String, one of "Cpu", "Memory"
# resp.recommended_option_projected_metrics[0].projected_metrics[0].timestamps #=> Array
# resp.recommended_option_projected_metrics[0].projected_metrics[0].timestamps[0] #=> Time
# resp.recommended_option_projected_metrics[0].projected_metrics[0].upper_bound_values #=> Array
# resp.recommended_option_projected_metrics[0].projected_metrics[0].upper_bound_values[0] #=> Float
# resp.recommended_option_projected_metrics[0].projected_metrics[0].lower_bound_values #=> Array
# resp.recommended_option_projected_metrics[0].projected_metrics[0].lower_bound_values[0] #=> Float
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetECSServiceRecommendationProjectedMetrics AWS API Documentation
#
# @overload get_ecs_service_recommendation_projected_metrics(params = {})
# @param [Hash] params ({})
def get_ecs_service_recommendation_projected_metrics(params = {}, options = {})
req = build_request(:get_ecs_service_recommendation_projected_metrics, params)
req.send_request(options)
end
# Returns Amazon ECS service recommendations.
#
# Compute Optimizer generates recommendations for Amazon ECS services on
# Fargate that meet a specific set of requirements. For more
# information, see the [Supported resources and requirements][1] in the
# *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/requirements.html
#
# @option params [Array] :service_arns
# The ARN that identifies the Amazon ECS service.
#
# The following is the format of the ARN:
#
# `arn:aws:ecs:region:aws_account_id:service/cluster-name/service-name`
#
# @option params [String] :next_token
# The token to advance to the next page of Amazon ECS service
# recommendations.
#
# @option params [Integer] :max_results
# The maximum number of Amazon ECS service recommendations to return
# with a single request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @option params [Array] :filters
# An array of objects to specify a filter that returns a more specific
# list of Amazon ECS service recommendations.
#
# @option params [Array] :account_ids
# Return the Amazon ECS service recommendations to the specified Amazon
# Web Services account IDs.
#
# If your account is the management account or the delegated
# administrator of an organization, use this parameter to return the
# Amazon ECS service recommendations to specific member accounts.
#
# You can only specify one account ID per request.
#
# @return [Types::GetECSServiceRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetECSServiceRecommendationsResponse#next_token #next_token} => String
# * {Types::GetECSServiceRecommendationsResponse#ecs_service_recommendations #ecs_service_recommendations} => Array<Types::ECSServiceRecommendation>
# * {Types::GetECSServiceRecommendationsResponse#errors #errors} => Array<Types::GetRecommendationError>
#
# @example Request syntax with placeholder values
#
# resp = client.get_ecs_service_recommendations({
# service_arns: ["ServiceArn"],
# next_token: "NextToken",
# max_results: 1,
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCode
# values: ["FilterValue"],
# },
# ],
# account_ids: ["AccountId"],
# })
#
# @example Response structure
#
# resp.next_token #=> String
# resp.ecs_service_recommendations #=> Array
# resp.ecs_service_recommendations[0].service_arn #=> String
# resp.ecs_service_recommendations[0].account_id #=> String
# resp.ecs_service_recommendations[0].current_service_configuration.memory #=> Integer
# resp.ecs_service_recommendations[0].current_service_configuration.cpu #=> Integer
# resp.ecs_service_recommendations[0].current_service_configuration.container_configurations #=> Array
# resp.ecs_service_recommendations[0].current_service_configuration.container_configurations[0].container_name #=> String
# resp.ecs_service_recommendations[0].current_service_configuration.container_configurations[0].memory_size_configuration.memory #=> Integer
# resp.ecs_service_recommendations[0].current_service_configuration.container_configurations[0].memory_size_configuration.memory_reservation #=> Integer
# resp.ecs_service_recommendations[0].current_service_configuration.container_configurations[0].cpu #=> Integer
# resp.ecs_service_recommendations[0].current_service_configuration.auto_scaling_configuration #=> String, one of "TargetTrackingScalingCpu", "TargetTrackingScalingMemory"
# resp.ecs_service_recommendations[0].current_service_configuration.task_definition_arn #=> String
# resp.ecs_service_recommendations[0].utilization_metrics #=> Array
# resp.ecs_service_recommendations[0].utilization_metrics[0].name #=> String, one of "Cpu", "Memory"
# resp.ecs_service_recommendations[0].utilization_metrics[0].statistic #=> String, one of "Maximum", "Average"
# resp.ecs_service_recommendations[0].utilization_metrics[0].value #=> Float
# resp.ecs_service_recommendations[0].lookback_period_in_days #=> Float
# resp.ecs_service_recommendations[0].launch_type #=> String, one of "EC2", "Fargate"
# resp.ecs_service_recommendations[0].last_refresh_timestamp #=> Time
# resp.ecs_service_recommendations[0].finding #=> String, one of "Optimized", "Underprovisioned", "Overprovisioned"
# resp.ecs_service_recommendations[0].finding_reason_codes #=> Array
# resp.ecs_service_recommendations[0].finding_reason_codes[0] #=> String, one of "MemoryOverprovisioned", "MemoryUnderprovisioned", "CPUOverprovisioned", "CPUUnderprovisioned"
# resp.ecs_service_recommendations[0].service_recommendation_options #=> Array
# resp.ecs_service_recommendations[0].service_recommendation_options[0].memory #=> Integer
# resp.ecs_service_recommendations[0].service_recommendation_options[0].cpu #=> Integer
# resp.ecs_service_recommendations[0].service_recommendation_options[0].savings_opportunity.savings_opportunity_percentage #=> Float
# resp.ecs_service_recommendations[0].service_recommendation_options[0].savings_opportunity.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.ecs_service_recommendations[0].service_recommendation_options[0].savings_opportunity.estimated_monthly_savings.value #=> Float
# resp.ecs_service_recommendations[0].service_recommendation_options[0].projected_utilization_metrics #=> Array
# resp.ecs_service_recommendations[0].service_recommendation_options[0].projected_utilization_metrics[0].name #=> String, one of "Cpu", "Memory"
# resp.ecs_service_recommendations[0].service_recommendation_options[0].projected_utilization_metrics[0].statistic #=> String, one of "Maximum", "Average"
# resp.ecs_service_recommendations[0].service_recommendation_options[0].projected_utilization_metrics[0].lower_bound_value #=> Float
# resp.ecs_service_recommendations[0].service_recommendation_options[0].projected_utilization_metrics[0].upper_bound_value #=> Float
# resp.ecs_service_recommendations[0].service_recommendation_options[0].container_recommendations #=> Array
# resp.ecs_service_recommendations[0].service_recommendation_options[0].container_recommendations[0].container_name #=> String
# resp.ecs_service_recommendations[0].service_recommendation_options[0].container_recommendations[0].memory_size_configuration.memory #=> Integer
# resp.ecs_service_recommendations[0].service_recommendation_options[0].container_recommendations[0].memory_size_configuration.memory_reservation #=> Integer
# resp.ecs_service_recommendations[0].service_recommendation_options[0].container_recommendations[0].cpu #=> Integer
# resp.ecs_service_recommendations[0].service_recommendation_options[0].savings_opportunity_after_discounts.savings_opportunity_percentage #=> Float
# resp.ecs_service_recommendations[0].service_recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.ecs_service_recommendations[0].service_recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.value #=> Float
# resp.ecs_service_recommendations[0].current_performance_risk #=> String, one of "VeryLow", "Low", "Medium", "High"
# resp.ecs_service_recommendations[0].tags #=> Array
# resp.ecs_service_recommendations[0].tags[0].key #=> String
# resp.ecs_service_recommendations[0].tags[0].value #=> String
# resp.ecs_service_recommendations[0].effective_recommendation_preferences.savings_estimation_mode.source #=> String, one of "PublicPricing", "CostExplorerRightsizing", "CostOptimizationHub"
# resp.errors #=> Array
# resp.errors[0].identifier #=> String
# resp.errors[0].code #=> String
# resp.errors[0].message #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetECSServiceRecommendations AWS API Documentation
#
# @overload get_ecs_service_recommendations(params = {})
# @param [Hash] params ({})
def get_ecs_service_recommendations(params = {}, options = {})
req = build_request(:get_ecs_service_recommendations, params)
req.send_request(options)
end
# Returns the recommendation preferences that are in effect for a given
# resource, such as enhanced infrastructure metrics. Considers all
# applicable preferences that you might have set at the resource,
# account, and organization level.
#
# When you create a recommendation preference, you can set its status to
# `Active` or `Inactive`. Use this action to view the recommendation
# preferences that are in effect, or `Active`.
#
# @option params [required, String] :resource_arn
# The Amazon Resource Name (ARN) of the resource for which to confirm
# effective recommendation preferences. Only EC2 instance and Auto
# Scaling group ARNs are currently supported.
#
# @return [Types::GetEffectiveRecommendationPreferencesResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetEffectiveRecommendationPreferencesResponse#enhanced_infrastructure_metrics #enhanced_infrastructure_metrics} => String
# * {Types::GetEffectiveRecommendationPreferencesResponse#external_metrics_preference #external_metrics_preference} => Types::ExternalMetricsPreference
# * {Types::GetEffectiveRecommendationPreferencesResponse#look_back_period #look_back_period} => String
# * {Types::GetEffectiveRecommendationPreferencesResponse#utilization_preferences #utilization_preferences} => Array<Types::UtilizationPreference>
# * {Types::GetEffectiveRecommendationPreferencesResponse#preferred_resources #preferred_resources} => Array<Types::EffectivePreferredResource>
#
# @example Request syntax with placeholder values
#
# resp = client.get_effective_recommendation_preferences({
# resource_arn: "ResourceArn", # required
# })
#
# @example Response structure
#
# resp.enhanced_infrastructure_metrics #=> String, one of "Active", "Inactive"
# resp.external_metrics_preference.source #=> String, one of "Datadog", "Dynatrace", "NewRelic", "Instana"
# resp.look_back_period #=> String, one of "DAYS_14", "DAYS_32", "DAYS_93"
# resp.utilization_preferences #=> Array
# resp.utilization_preferences[0].metric_name #=> String, one of "CpuUtilization", "MemoryUtilization"
# resp.utilization_preferences[0].metric_parameters.threshold #=> String, one of "P90", "P95", "P99_5"
# resp.utilization_preferences[0].metric_parameters.headroom #=> String, one of "PERCENT_30", "PERCENT_20", "PERCENT_10", "PERCENT_0"
# resp.preferred_resources #=> Array
# resp.preferred_resources[0].name #=> String, one of "Ec2InstanceTypes"
# resp.preferred_resources[0].include_list #=> Array
# resp.preferred_resources[0].include_list[0] #=> String
# resp.preferred_resources[0].effective_include_list #=> Array
# resp.preferred_resources[0].effective_include_list[0] #=> String
# resp.preferred_resources[0].exclude_list #=> Array
# resp.preferred_resources[0].exclude_list[0] #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetEffectiveRecommendationPreferences AWS API Documentation
#
# @overload get_effective_recommendation_preferences(params = {})
# @param [Hash] params ({})
def get_effective_recommendation_preferences(params = {}, options = {})
req = build_request(:get_effective_recommendation_preferences, params)
req.send_request(options)
end
# Returns the enrollment (opt in) status of an account to the Compute
# Optimizer service.
#
# If the account is the management account of an organization, this
# action also confirms the enrollment status of member accounts of the
# organization. Use the GetEnrollmentStatusesForOrganization action to
# get detailed information about the enrollment status of member
# accounts of an organization.
#
# @return [Types::GetEnrollmentStatusResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetEnrollmentStatusResponse#status #status} => String
# * {Types::GetEnrollmentStatusResponse#status_reason #status_reason} => String
# * {Types::GetEnrollmentStatusResponse#member_accounts_enrolled #member_accounts_enrolled} => Boolean
# * {Types::GetEnrollmentStatusResponse#last_updated_timestamp #last_updated_timestamp} => Time
# * {Types::GetEnrollmentStatusResponse#number_of_member_accounts_opted_in #number_of_member_accounts_opted_in} => Integer
#
# @example Response structure
#
# resp.status #=> String, one of "Active", "Inactive", "Pending", "Failed"
# resp.status_reason #=> String
# resp.member_accounts_enrolled #=> Boolean
# resp.last_updated_timestamp #=> Time
# resp.number_of_member_accounts_opted_in #=> Integer
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetEnrollmentStatus AWS API Documentation
#
# @overload get_enrollment_status(params = {})
# @param [Hash] params ({})
def get_enrollment_status(params = {}, options = {})
req = build_request(:get_enrollment_status, params)
req.send_request(options)
end
# Returns the Compute Optimizer enrollment (opt-in) status of
# organization member accounts, if your account is an organization
# management account.
#
# To get the enrollment status of standalone accounts, use the
# GetEnrollmentStatus action.
#
# @option params [Array] :filters
# An array of objects to specify a filter that returns a more specific
# list of account enrollment statuses.
#
# @option params [String] :next_token
# The token to advance to the next page of account enrollment statuses.
#
# @option params [Integer] :max_results
# The maximum number of account enrollment statuses to return with a
# single request. You can specify up to 100 statuses to return with each
# request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @return [Types::GetEnrollmentStatusesForOrganizationResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetEnrollmentStatusesForOrganizationResponse#account_enrollment_statuses #account_enrollment_statuses} => Array<Types::AccountEnrollmentStatus>
# * {Types::GetEnrollmentStatusesForOrganizationResponse#next_token #next_token} => String
#
# The returned {Seahorse::Client::Response response} is a pageable response and is Enumerable. For details on usage see {Aws::PageableResponse PageableResponse}.
#
# @example Request syntax with placeholder values
#
# resp = client.get_enrollment_statuses_for_organization({
# filters: [
# {
# name: "Status", # accepts Status
# values: ["FilterValue"],
# },
# ],
# next_token: "NextToken",
# max_results: 1,
# })
#
# @example Response structure
#
# resp.account_enrollment_statuses #=> Array
# resp.account_enrollment_statuses[0].account_id #=> String
# resp.account_enrollment_statuses[0].status #=> String, one of "Active", "Inactive", "Pending", "Failed"
# resp.account_enrollment_statuses[0].status_reason #=> String
# resp.account_enrollment_statuses[0].last_updated_timestamp #=> Time
# resp.next_token #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetEnrollmentStatusesForOrganization AWS API Documentation
#
# @overload get_enrollment_statuses_for_organization(params = {})
# @param [Hash] params ({})
def get_enrollment_statuses_for_organization(params = {}, options = {})
req = build_request(:get_enrollment_statuses_for_organization, params)
req.send_request(options)
end
# Returns Lambda function recommendations.
#
# Compute Optimizer generates recommendations for functions that meet a
# specific set of requirements. For more information, see the [Supported
# resources and requirements][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/requirements.html
#
# @option params [Array] :function_arns
# The Amazon Resource Name (ARN) of the functions for which to return
# recommendations.
#
# You can specify a qualified or unqualified ARN. If you specify an
# unqualified ARN without a function version suffix, Compute Optimizer
# will return recommendations for the latest (`$LATEST`) version of the
# function. If you specify a qualified ARN with a version suffix,
# Compute Optimizer will return recommendations for the specified
# function version. For more information about using function versions,
# see [Using versions][1] in the *Lambda Developer Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/lambda/latest/dg/configuration-versions.html#versioning-versions-using
#
# @option params [Array] :account_ids
# The ID of the Amazon Web Services account for which to return function
# recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to return
# function recommendations.
#
# Only one account ID can be specified per request.
#
# @option params [Array] :filters
# An array of objects to specify a filter that returns a more specific
# list of function recommendations.
#
# @option params [String] :next_token
# The token to advance to the next page of function recommendations.
#
# @option params [Integer] :max_results
# The maximum number of function recommendations to return with a single
# request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @return [Types::GetLambdaFunctionRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetLambdaFunctionRecommendationsResponse#next_token #next_token} => String
# * {Types::GetLambdaFunctionRecommendationsResponse#lambda_function_recommendations #lambda_function_recommendations} => Array<Types::LambdaFunctionRecommendation>
#
# The returned {Seahorse::Client::Response response} is a pageable response and is Enumerable. For details on usage see {Aws::PageableResponse PageableResponse}.
#
# @example Request syntax with placeholder values
#
# resp = client.get_lambda_function_recommendations({
# function_arns: ["FunctionArn"],
# account_ids: ["AccountId"],
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCode
# values: ["FilterValue"],
# },
# ],
# next_token: "NextToken",
# max_results: 1,
# })
#
# @example Response structure
#
# resp.next_token #=> String
# resp.lambda_function_recommendations #=> Array
# resp.lambda_function_recommendations[0].function_arn #=> String
# resp.lambda_function_recommendations[0].function_version #=> String
# resp.lambda_function_recommendations[0].account_id #=> String
# resp.lambda_function_recommendations[0].current_memory_size #=> Integer
# resp.lambda_function_recommendations[0].number_of_invocations #=> Integer
# resp.lambda_function_recommendations[0].utilization_metrics #=> Array
# resp.lambda_function_recommendations[0].utilization_metrics[0].name #=> String, one of "Duration", "Memory"
# resp.lambda_function_recommendations[0].utilization_metrics[0].statistic #=> String, one of "Maximum", "Average"
# resp.lambda_function_recommendations[0].utilization_metrics[0].value #=> Float
# resp.lambda_function_recommendations[0].lookback_period_in_days #=> Float
# resp.lambda_function_recommendations[0].last_refresh_timestamp #=> Time
# resp.lambda_function_recommendations[0].finding #=> String, one of "Optimized", "NotOptimized", "Unavailable"
# resp.lambda_function_recommendations[0].finding_reason_codes #=> Array
# resp.lambda_function_recommendations[0].finding_reason_codes[0] #=> String, one of "MemoryOverprovisioned", "MemoryUnderprovisioned", "InsufficientData", "Inconclusive"
# resp.lambda_function_recommendations[0].memory_size_recommendation_options #=> Array
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].rank #=> Integer
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].memory_size #=> Integer
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].projected_utilization_metrics #=> Array
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].projected_utilization_metrics[0].name #=> String, one of "Duration"
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].projected_utilization_metrics[0].statistic #=> String, one of "LowerBound", "UpperBound", "Expected"
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].projected_utilization_metrics[0].value #=> Float
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].savings_opportunity.savings_opportunity_percentage #=> Float
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].savings_opportunity.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].savings_opportunity.estimated_monthly_savings.value #=> Float
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].savings_opportunity_after_discounts.savings_opportunity_percentage #=> Float
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.lambda_function_recommendations[0].memory_size_recommendation_options[0].savings_opportunity_after_discounts.estimated_monthly_savings.value #=> Float
# resp.lambda_function_recommendations[0].current_performance_risk #=> String, one of "VeryLow", "Low", "Medium", "High"
# resp.lambda_function_recommendations[0].tags #=> Array
# resp.lambda_function_recommendations[0].tags[0].key #=> String
# resp.lambda_function_recommendations[0].tags[0].value #=> String
# resp.lambda_function_recommendations[0].effective_recommendation_preferences.savings_estimation_mode.source #=> String, one of "PublicPricing", "CostExplorerRightsizing", "CostOptimizationHub"
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetLambdaFunctionRecommendations AWS API Documentation
#
# @overload get_lambda_function_recommendations(params = {})
# @param [Hash] params ({})
def get_lambda_function_recommendations(params = {}, options = {})
req = build_request(:get_lambda_function_recommendations, params)
req.send_request(options)
end
# Returns license recommendations for Amazon EC2 instances that run on a
# specific license.
#
# Compute Optimizer generates recommendations for licenses that meet a
# specific set of requirements. For more information, see the [Supported
# resources and requirements][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/requirements.html
#
# @option params [Array] :resource_arns
# The ARN that identifies the Amazon EC2 instance.
#
# The following is the format of the ARN:
#
# `arn:aws:ec2:region:aws_account_id:instance/instance-id`
#
# @option params [String] :next_token
# The token to advance to the next page of license recommendations.
#
# @option params [Integer] :max_results
# The maximum number of license recommendations to return with a single
# request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @option params [Array] :filters
# An array of objects to specify a filter that returns a more specific
# list of license recommendations.
#
# @option params [Array] :account_ids
# The ID of the Amazon Web Services account for which to return license
# recommendations.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to return
# license recommendations.
#
# Only one account ID can be specified per request.
#
# @return [Types::GetLicenseRecommendationsResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetLicenseRecommendationsResponse#next_token #next_token} => String
# * {Types::GetLicenseRecommendationsResponse#license_recommendations #license_recommendations} => Array<Types::LicenseRecommendation>
# * {Types::GetLicenseRecommendationsResponse#errors #errors} => Array<Types::GetRecommendationError>
#
# @example Request syntax with placeholder values
#
# resp = client.get_license_recommendations({
# resource_arns: ["ResourceArn"],
# next_token: "NextToken",
# max_results: 1,
# filters: [
# {
# name: "Finding", # accepts Finding, FindingReasonCode, LicenseName
# values: ["FilterValue"],
# },
# ],
# account_ids: ["AccountId"],
# })
#
# @example Response structure
#
# resp.next_token #=> String
# resp.license_recommendations #=> Array
# resp.license_recommendations[0].resource_arn #=> String
# resp.license_recommendations[0].account_id #=> String
# resp.license_recommendations[0].current_license_configuration.number_of_cores #=> Integer
# resp.license_recommendations[0].current_license_configuration.instance_type #=> String
# resp.license_recommendations[0].current_license_configuration.operating_system #=> String
# resp.license_recommendations[0].current_license_configuration.license_edition #=> String, one of "Enterprise", "Standard", "Free", "NoLicenseEditionFound"
# resp.license_recommendations[0].current_license_configuration.license_name #=> String, one of "SQLServer"
# resp.license_recommendations[0].current_license_configuration.license_model #=> String, one of "LicenseIncluded", "BringYourOwnLicense"
# resp.license_recommendations[0].current_license_configuration.license_version #=> String
# resp.license_recommendations[0].current_license_configuration.metrics_source #=> Array
# resp.license_recommendations[0].current_license_configuration.metrics_source[0].provider #=> String, one of "CloudWatchApplicationInsights"
# resp.license_recommendations[0].current_license_configuration.metrics_source[0].provider_arn #=> String
# resp.license_recommendations[0].lookback_period_in_days #=> Float
# resp.license_recommendations[0].last_refresh_timestamp #=> Time
# resp.license_recommendations[0].finding #=> String, one of "InsufficientMetrics", "Optimized", "NotOptimized"
# resp.license_recommendations[0].finding_reason_codes #=> Array
# resp.license_recommendations[0].finding_reason_codes[0] #=> String, one of "InvalidCloudWatchApplicationInsightsSetup", "CloudWatchApplicationInsightsError", "LicenseOverprovisioned", "Optimized"
# resp.license_recommendations[0].license_recommendation_options #=> Array
# resp.license_recommendations[0].license_recommendation_options[0].rank #=> Integer
# resp.license_recommendations[0].license_recommendation_options[0].operating_system #=> String
# resp.license_recommendations[0].license_recommendation_options[0].license_edition #=> String, one of "Enterprise", "Standard", "Free", "NoLicenseEditionFound"
# resp.license_recommendations[0].license_recommendation_options[0].license_model #=> String, one of "LicenseIncluded", "BringYourOwnLicense"
# resp.license_recommendations[0].license_recommendation_options[0].savings_opportunity.savings_opportunity_percentage #=> Float
# resp.license_recommendations[0].license_recommendation_options[0].savings_opportunity.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.license_recommendations[0].license_recommendation_options[0].savings_opportunity.estimated_monthly_savings.value #=> Float
# resp.license_recommendations[0].tags #=> Array
# resp.license_recommendations[0].tags[0].key #=> String
# resp.license_recommendations[0].tags[0].value #=> String
# resp.errors #=> Array
# resp.errors[0].identifier #=> String
# resp.errors[0].code #=> String
# resp.errors[0].message #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetLicenseRecommendations AWS API Documentation
#
# @overload get_license_recommendations(params = {})
# @param [Hash] params ({})
def get_license_recommendations(params = {}, options = {})
req = build_request(:get_license_recommendations, params)
req.send_request(options)
end
# Returns existing recommendation preferences, such as enhanced
# infrastructure metrics.
#
# Use the `scope` parameter to specify which preferences to return. You
# can specify to return preferences for an organization, a specific
# account ID, or a specific EC2 instance or Auto Scaling group Amazon
# Resource Name (ARN).
#
# For more information, see [Activating enhanced infrastructure
# metrics][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/enhanced-infrastructure-metrics.html
#
# @option params [required, String] :resource_type
# The target resource type of the recommendation preference for which to
# return preferences.
#
# The `Ec2Instance` option encompasses standalone instances and
# instances that are part of Auto Scaling groups. The `AutoScalingGroup`
# option encompasses only instances that are part of an Auto Scaling
# group.
#
# The valid values for this parameter are `Ec2Instance` and
# `AutoScalingGroup`.
#
#
#
# @option params [Types::Scope] :scope
# An object that describes the scope of the recommendation preference to
# return.
#
# You can return recommendation preferences that are created at the
# organization level (for management accounts of an organization only),
# account level, and resource level. For more information, see
# [Activating enhanced infrastructure metrics][1] in the *Compute
# Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/enhanced-infrastructure-metrics.html
#
# @option params [String] :next_token
# The token to advance to the next page of recommendation preferences.
#
# @option params [Integer] :max_results
# The maximum number of recommendation preferences to return with a
# single request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @return [Types::GetRecommendationPreferencesResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetRecommendationPreferencesResponse#next_token #next_token} => String
# * {Types::GetRecommendationPreferencesResponse#recommendation_preferences_details #recommendation_preferences_details} => Array<Types::RecommendationPreferencesDetail>
#
# The returned {Seahorse::Client::Response response} is a pageable response and is Enumerable. For details on usage see {Aws::PageableResponse PageableResponse}.
#
# @example Request syntax with placeholder values
#
# resp = client.get_recommendation_preferences({
# resource_type: "Ec2Instance", # required, accepts Ec2Instance, AutoScalingGroup, EbsVolume, LambdaFunction, NotApplicable, EcsService, License
# scope: {
# name: "Organization", # accepts Organization, AccountId, ResourceArn
# value: "ScopeValue",
# },
# next_token: "NextToken",
# max_results: 1,
# })
#
# @example Response structure
#
# resp.next_token #=> String
# resp.recommendation_preferences_details #=> Array
# resp.recommendation_preferences_details[0].scope.name #=> String, one of "Organization", "AccountId", "ResourceArn"
# resp.recommendation_preferences_details[0].scope.value #=> String
# resp.recommendation_preferences_details[0].resource_type #=> String, one of "Ec2Instance", "AutoScalingGroup", "EbsVolume", "LambdaFunction", "NotApplicable", "EcsService", "License"
# resp.recommendation_preferences_details[0].enhanced_infrastructure_metrics #=> String, one of "Active", "Inactive"
# resp.recommendation_preferences_details[0].inferred_workload_types #=> String, one of "Active", "Inactive"
# resp.recommendation_preferences_details[0].external_metrics_preference.source #=> String, one of "Datadog", "Dynatrace", "NewRelic", "Instana"
# resp.recommendation_preferences_details[0].look_back_period #=> String, one of "DAYS_14", "DAYS_32", "DAYS_93"
# resp.recommendation_preferences_details[0].utilization_preferences #=> Array
# resp.recommendation_preferences_details[0].utilization_preferences[0].metric_name #=> String, one of "CpuUtilization", "MemoryUtilization"
# resp.recommendation_preferences_details[0].utilization_preferences[0].metric_parameters.threshold #=> String, one of "P90", "P95", "P99_5"
# resp.recommendation_preferences_details[0].utilization_preferences[0].metric_parameters.headroom #=> String, one of "PERCENT_30", "PERCENT_20", "PERCENT_10", "PERCENT_0"
# resp.recommendation_preferences_details[0].preferred_resources #=> Array
# resp.recommendation_preferences_details[0].preferred_resources[0].name #=> String, one of "Ec2InstanceTypes"
# resp.recommendation_preferences_details[0].preferred_resources[0].include_list #=> Array
# resp.recommendation_preferences_details[0].preferred_resources[0].include_list[0] #=> String
# resp.recommendation_preferences_details[0].preferred_resources[0].effective_include_list #=> Array
# resp.recommendation_preferences_details[0].preferred_resources[0].effective_include_list[0] #=> String
# resp.recommendation_preferences_details[0].preferred_resources[0].exclude_list #=> Array
# resp.recommendation_preferences_details[0].preferred_resources[0].exclude_list[0] #=> String
# resp.recommendation_preferences_details[0].savings_estimation_mode #=> String, one of "AfterDiscounts", "BeforeDiscounts"
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetRecommendationPreferences AWS API Documentation
#
# @overload get_recommendation_preferences(params = {})
# @param [Hash] params ({})
def get_recommendation_preferences(params = {}, options = {})
req = build_request(:get_recommendation_preferences, params)
req.send_request(options)
end
# Returns the optimization findings for an account.
#
# It returns the number of:
#
# * Amazon EC2 instances in an account that are `Underprovisioned`,
# `Overprovisioned`, or `Optimized`.
#
# * Auto Scaling groups in an account that are `NotOptimized`, or
# `Optimized`.
#
# * Amazon EBS volumes in an account that are `NotOptimized`, or
# `Optimized`.
#
# * Lambda functions in an account that are `NotOptimized`, or
# `Optimized`.
#
# * Amazon ECS services in an account that are `Underprovisioned`,
# `Overprovisioned`, or `Optimized`.
#
# @option params [Array] :account_ids
# The ID of the Amazon Web Services account for which to return
# recommendation summaries.
#
# If your account is the management account of an organization, use this
# parameter to specify the member account for which you want to return
# recommendation summaries.
#
# Only one account ID can be specified per request.
#
# @option params [String] :next_token
# The token to advance to the next page of recommendation summaries.
#
# @option params [Integer] :max_results
# The maximum number of recommendation summaries to return with a single
# request.
#
# To retrieve the remaining results, make another request with the
# returned `nextToken` value.
#
# @return [Types::GetRecommendationSummariesResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::GetRecommendationSummariesResponse#next_token #next_token} => String
# * {Types::GetRecommendationSummariesResponse#recommendation_summaries #recommendation_summaries} => Array<Types::RecommendationSummary>
#
# The returned {Seahorse::Client::Response response} is a pageable response and is Enumerable. For details on usage see {Aws::PageableResponse PageableResponse}.
#
# @example Request syntax with placeholder values
#
# resp = client.get_recommendation_summaries({
# account_ids: ["AccountId"],
# next_token: "NextToken",
# max_results: 1,
# })
#
# @example Response structure
#
# resp.next_token #=> String
# resp.recommendation_summaries #=> Array
# resp.recommendation_summaries[0].summaries #=> Array
# resp.recommendation_summaries[0].summaries[0].name #=> String, one of "Underprovisioned", "Overprovisioned", "Optimized", "NotOptimized"
# resp.recommendation_summaries[0].summaries[0].value #=> Float
# resp.recommendation_summaries[0].summaries[0].reason_code_summaries #=> Array
# resp.recommendation_summaries[0].summaries[0].reason_code_summaries[0].name #=> String, one of "MemoryOverprovisioned", "MemoryUnderprovisioned"
# resp.recommendation_summaries[0].summaries[0].reason_code_summaries[0].value #=> Float
# resp.recommendation_summaries[0].recommendation_resource_type #=> String, one of "Ec2Instance", "AutoScalingGroup", "EbsVolume", "LambdaFunction", "EcsService", "License"
# resp.recommendation_summaries[0].account_id #=> String
# resp.recommendation_summaries[0].savings_opportunity.savings_opportunity_percentage #=> Float
# resp.recommendation_summaries[0].savings_opportunity.estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.recommendation_summaries[0].savings_opportunity.estimated_monthly_savings.value #=> Float
# resp.recommendation_summaries[0].current_performance_risk_ratings.high #=> Integer
# resp.recommendation_summaries[0].current_performance_risk_ratings.medium #=> Integer
# resp.recommendation_summaries[0].current_performance_risk_ratings.low #=> Integer
# resp.recommendation_summaries[0].current_performance_risk_ratings.very_low #=> Integer
# resp.recommendation_summaries[0].inferred_workload_savings #=> Array
# resp.recommendation_summaries[0].inferred_workload_savings[0].inferred_workload_types #=> Array
# resp.recommendation_summaries[0].inferred_workload_savings[0].inferred_workload_types[0] #=> String, one of "AmazonEmr", "ApacheCassandra", "ApacheHadoop", "Memcached", "Nginx", "PostgreSql", "Redis", "Kafka", "SQLServer"
# resp.recommendation_summaries[0].inferred_workload_savings[0].estimated_monthly_savings.currency #=> String, one of "USD", "CNY"
# resp.recommendation_summaries[0].inferred_workload_savings[0].estimated_monthly_savings.value #=> Float
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/GetRecommendationSummaries AWS API Documentation
#
# @overload get_recommendation_summaries(params = {})
# @param [Hash] params ({})
def get_recommendation_summaries(params = {}, options = {})
req = build_request(:get_recommendation_summaries, params)
req.send_request(options)
end
# Creates a new recommendation preference or updates an existing
# recommendation preference, such as enhanced infrastructure metrics.
#
# For more information, see [Activating enhanced infrastructure
# metrics][1] in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/enhanced-infrastructure-metrics.html
#
# @option params [required, String] :resource_type
# The target resource type of the recommendation preference to create.
#
# The `Ec2Instance` option encompasses standalone instances and
# instances that are part of Auto Scaling groups. The `AutoScalingGroup`
# option encompasses only instances that are part of an Auto Scaling
# group.
#
# The valid values for this parameter are `Ec2Instance` and
# `AutoScalingGroup`.
#
#
#
# @option params [Types::Scope] :scope
# An object that describes the scope of the recommendation preference to
# create.
#
# You can create recommendation preferences at the organization level
# (for management accounts of an organization only), account level, and
# resource level. For more information, see [Activating enhanced
# infrastructure metrics][1] in the *Compute Optimizer User Guide*.
#
# You cannot create recommendation preferences for Auto Scaling groups
# at the organization and account levels. You can create recommendation
# preferences for Auto Scaling groups only at the resource level by
# specifying a scope name of `ResourceArn` and a scope value of the Auto
# Scaling group Amazon Resource Name (ARN). This will configure the
# preference for all instances that are part of the specified Auto
# Scaling group. You also cannot create recommendation preferences at
# the resource level for instances that are part of an Auto Scaling
# group. You can create recommendation preferences at the resource level
# only for standalone instances.
#
#
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/enhanced-infrastructure-metrics.html
#
# @option params [String] :enhanced_infrastructure_metrics
# The status of the enhanced infrastructure metrics recommendation
# preference to create or update.
#
# Specify the `Active` status to activate the preference, or specify
# `Inactive` to deactivate the preference.
#
# For more information, see [Enhanced infrastructure metrics][1] in the
# *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/enhanced-infrastructure-metrics.html
#
# @option params [String] :inferred_workload_types
# The status of the inferred workload types recommendation preference to
# create or update.
#
# The inferred workload type feature is active by default. To deactivate
# it, create a recommendation preference.
#
#
#
# Specify the `Inactive` status to deactivate the feature, or specify
# `Active` to activate it.
#
# For more information, see [Inferred workload types][1] in the *Compute
# Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/inferred-workload-types.html
#
# @option params [Types::ExternalMetricsPreference] :external_metrics_preference
# The provider of the external metrics recommendation preference to
# create or update.
#
# Specify a valid provider in the `source` field to activate the
# preference. To delete this preference, see the
# DeleteRecommendationPreferences action.
#
# This preference can only be set for the `Ec2Instance` resource type.
#
# For more information, see [External metrics ingestion][1] in the
# *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/external-metrics-ingestion.html
#
# @option params [String] :look_back_period
# The preference to control the number of days the utilization metrics
# of the Amazon Web Services resource are analyzed. When this preference
# isn't specified, we use the default value `DAYS_14`.
#
# You can only set this preference for the Amazon EC2 instance and Auto
# Scaling group resource types.
#
#
#
# @option params [Array] :utilization_preferences
# The preference to control the resource’s CPU utilization threshold,
# CPU utilization headroom, and memory utilization headroom. When this
# preference isn't specified, we use the following default values.
#
# CPU utilization:
#
# * `P99_5` for threshold
#
# * `PERCENT_20` for headroom
#
# Memory utilization:
#
# * `PERCENT_20` for headroom
#
# ^
#
# * You can only set CPU and memory utilization preferences for the
# Amazon EC2 instance resource type.
#
# * The threshold setting isn’t available for memory utilization.
#
#
#
# @option params [Array] :preferred_resources
# The preference to control which resource type values are considered
# when generating rightsizing recommendations. You can specify this
# preference as a combination of include and exclude lists. You must
# specify either an `includeList` or `excludeList`. If the preference is
# an empty set of resource type values, an error occurs.
#
# You can only set this preference for the Amazon EC2 instance and Auto
# Scaling group resource types.
#
#
#
# @option params [String] :savings_estimation_mode
# The status of the savings estimation mode preference to create or
# update.
#
# Specify the `AfterDiscounts` status to activate the preference, or
# specify `BeforeDiscounts` to deactivate the preference.
#
# Only the account manager or delegated administrator of your
# organization can activate this preference.
#
# For more information, see [ Savings estimation mode][1] in the
# *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/savings-estimation-mode.html
#
# @return [Struct] Returns an empty {Seahorse::Client::Response response}.
#
# @example Request syntax with placeholder values
#
# resp = client.put_recommendation_preferences({
# resource_type: "Ec2Instance", # required, accepts Ec2Instance, AutoScalingGroup, EbsVolume, LambdaFunction, NotApplicable, EcsService, License
# scope: {
# name: "Organization", # accepts Organization, AccountId, ResourceArn
# value: "ScopeValue",
# },
# enhanced_infrastructure_metrics: "Active", # accepts Active, Inactive
# inferred_workload_types: "Active", # accepts Active, Inactive
# external_metrics_preference: {
# source: "Datadog", # accepts Datadog, Dynatrace, NewRelic, Instana
# },
# look_back_period: "DAYS_14", # accepts DAYS_14, DAYS_32, DAYS_93
# utilization_preferences: [
# {
# metric_name: "CpuUtilization", # accepts CpuUtilization, MemoryUtilization
# metric_parameters: {
# threshold: "P90", # accepts P90, P95, P99_5
# headroom: "PERCENT_30", # accepts PERCENT_30, PERCENT_20, PERCENT_10, PERCENT_0
# },
# },
# ],
# preferred_resources: [
# {
# name: "Ec2InstanceTypes", # accepts Ec2InstanceTypes
# include_list: ["PreferredResourceValue"],
# exclude_list: ["PreferredResourceValue"],
# },
# ],
# savings_estimation_mode: "AfterDiscounts", # accepts AfterDiscounts, BeforeDiscounts
# })
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/PutRecommendationPreferences AWS API Documentation
#
# @overload put_recommendation_preferences(params = {})
# @param [Hash] params ({})
def put_recommendation_preferences(params = {}, options = {})
req = build_request(:put_recommendation_preferences, params)
req.send_request(options)
end
# Updates the enrollment (opt in and opt out) status of an account to
# the Compute Optimizer service.
#
# If the account is a management account of an organization, this action
# can also be used to enroll member accounts of the organization.
#
# You must have the appropriate permissions to opt in to Compute
# Optimizer, to view its recommendations, and to opt out. For more
# information, see [Controlling access with Amazon Web Services Identity
# and Access Management][1] in the *Compute Optimizer User Guide*.
#
# When you opt in, Compute Optimizer automatically creates a
# service-linked role in your account to access its data. For more
# information, see [Using Service-Linked Roles for Compute Optimizer][2]
# in the *Compute Optimizer User Guide*.
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/security-iam.html
# [2]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/using-service-linked-roles.html
#
# @option params [required, String] :status
# The new enrollment status of the account.
#
# The following status options are available:
#
# * `Active` - Opts in your account to the Compute Optimizer service.
# Compute Optimizer begins analyzing the configuration and utilization
# metrics of your Amazon Web Services resources after you opt in. For
# more information, see [Metrics analyzed by Compute Optimizer][1] in
# the *Compute Optimizer User Guide*.
#
# * `Inactive` - Opts out your account from the Compute Optimizer
# service. Your account's recommendations and related metrics data
# will be deleted from Compute Optimizer after you opt out.
#
# The `Pending` and `Failed` options cannot be used to update the
# enrollment status of an account. They are returned in the response of
# a request to update the enrollment status of an account.
#
#
#
#
#
# [1]: https://docs.aws.amazon.com/compute-optimizer/latest/ug/metrics.html
#
# @option params [Boolean] :include_member_accounts
# Indicates whether to enroll member accounts of the organization if the
# account is the management account of an organization.
#
# @return [Types::UpdateEnrollmentStatusResponse] Returns a {Seahorse::Client::Response response} object which responds to the following methods:
#
# * {Types::UpdateEnrollmentStatusResponse#status #status} => String
# * {Types::UpdateEnrollmentStatusResponse#status_reason #status_reason} => String
#
# @example Request syntax with placeholder values
#
# resp = client.update_enrollment_status({
# status: "Active", # required, accepts Active, Inactive, Pending, Failed
# include_member_accounts: false,
# })
#
# @example Response structure
#
# resp.status #=> String, one of "Active", "Inactive", "Pending", "Failed"
# resp.status_reason #=> String
#
# @see http://docs.aws.amazon.com/goto/WebAPI/compute-optimizer-2019-11-01/UpdateEnrollmentStatus AWS API Documentation
#
# @overload update_enrollment_status(params = {})
# @param [Hash] params ({})
def update_enrollment_status(params = {}, options = {})
req = build_request(:update_enrollment_status, params)
req.send_request(options)
end
# @!endgroup
# @param params ({})
# @api private
def build_request(operation_name, params = {})
handlers = @handlers.for(operation_name)
context = Seahorse::Client::RequestContext.new(
operation_name: operation_name,
operation: config.api.operation(operation_name),
client: self,
params: params,
config: config)
context[:gem_name] = 'aws-sdk-computeoptimizer'
context[:gem_version] = '1.57.0'
Seahorse::Client::Request.new(handlers, context)
end
# @api private
# @deprecated
def waiter_names
[]
end
class << self
# @api private
attr_reader :identifier
# @api private
def errors_module
Errors
end
end
end
end