Amazon Elastic Load Balancing

ELB デフォルトダッシュボード

概要

Elastic Load Balancing (ELB) は、アプリケーションから異なるアベイラビリティーゾーンにある複数の Amazon EC2 バックエンドインスタンスに受信 Web トラフィックをディスパッチするために使用される AWS サービスです。ELB は、スムーズなユーザーエクスペリエンスを保証しつつ、フォールトトレランスを向上させ、トラフィックのピークや機能停止した EC2 インスタンスを中断なく処理します。

Datadog は、AWS が提供する 3 つのバリエーションの Elastic Load Balancer (Application、Classic、Network Load Balancer) からメトリクスとメタデータを収集します。

セットアップ

インストール

Amazon Web Services インテグレーションをまだセットアップしていない場合は、最初にセットアップします。

メトリクスの収集

  1. AWS インテグレーションページの Metric Collection タブで、ELB が有効になっていることを確認します。Application ELB と Network ELB のメトリクスをそれぞれ ApplicationELB と NetworkELB に設定します。

  2. Amazon ELB のメトリクスを収集するには、次のアクセス許可を Datadog IAM ポリシーに追加します。詳細については、AWS ウェブサイト上の ELB ポリシーを参照してください。

    AWS アクセス許可説明
    elasticloadbalancing:DescribeLoadBalancersELB を一覧表示し、タグとメトリクスを追加します。
    elasticloadbalancing:DescribeTagsELB メトリクスにカスタム ELB タグを追加します。
    elasticloadbalancing:DescribeInstanceHealthインスタンスの状態を追加します。
    elasticloadbalancing:DescribeTargetGroups指定したターゲットグループまたはすべてのターゲットグループを説明します。

  3. Datadog - Amazon Elastic Load Balancing インテグレーションをインストールします。

ログの収集

AWS ELB および ALB ログの有効化

ログを収集するには、まず ELB または ALB でログを有効にします。ELB または ALB ログを AWS S3 バケットに書き込み、Lambda 関数で使用することができます。詳細については、Classic Load Balancer のアクセスログを有効にするを参照してください。

AWS ELB ログ有効

間隔を 5 分に設定し、S3 バケットとプレフィックスを定義します。S3 イベント通知設定をあいまいに定義するのを回避するには、他のロードバランサーのログの場所と重複しない一意の場所を使用してください。複数のロードバランサーが同じバケットにログを記録している場合は、必ず my-bucket-for-elb-logs/my-elb-name などの一意のプレフィックスを使用して、ログを別々の場所に保存してください。

AWS ELB ログ構成

ログを Datadog に送信する方法

  1. AWS アカウントで Datadog Forwarder Lambda 関数をまだセットアップしていない場合は、セットアップします。
  2. 設定が完了したら、Datadog Forwarder Lambda 関数に移動します。ELB ログを含む S3 バケットで、自動または 手動のトリガーをセットアップします。手動セットアップでは、All object create events のイベントタイプを使用します。
  3. ログエクスプローラーを使って、ログを確認します。

AWS Services のログを収集する方法については、Datadog Lambda 関数で AWS Services のログを送信するを参照してください。

収集データ

メトリクスは次のネームスペースの下に収集されます。

Datadog ネームスペースAWS サービス
aws.applicationelbApplication Load Balancer
aws.elbClassic Elastic Load Balancing
aws.networkelbNetwork Load Balancer

メトリクス

aws.applicationelb.active_connection_count
(count)		The total number of concurrent TCP connections active from clients to the load balancer and from the load balancer to targets.
Shown as connection
aws.applicationelb.client_tlsnegotiation_error_count
(count)		Number of TLS negotiation errors
Shown as error
aws.applicationelb.consumed_lcus
(gauge)		The number of load balancer capacity units (LCU) used by your load balancer.
Shown as unit
aws.applicationelb.desync_mitigation_mode_non_compliant_request
(count)		The number of requests that do not comply with RFC 7230.
Shown as request
aws.applicationelb.elb_auth_error
(count)		The number of user authentications that could not be completed because an authenticate action was misconfigured, the load balancer couldn't establish a connection with the IdP, or the load balancer couldn't complete the authentication flow due to an internal error.
Shown as error
aws.applicationelb.elb_auth_failure
(count)		The number of user authentications that could not be completed because the IdP denied access to the user or an authorization code was used more than once.
Shown as error
aws.applicationelb.elb_auth_latency
(gauge)		The time elapsed, in milliseconds, to query the IdP for the ID token and user info. If one or more of these operations fail, this is the time to failure.
Shown as millisecond
aws.applicationelb.elb_auth_refresh_token_success
(count)		The number of times the load balancer successfully refreshed user claims using a refresh token provided by the IdP.
Shown as success
aws.applicationelb.elb_auth_success
(count)		The number of authenticate actions that were successful.
Shown as success
aws.applicationelb.elb_authuser_claims_size_exceeded
(count)		The number of times that a configured IdP returned user claims that exceeded 11K bytes in size.
aws.applicationelb.grpc_request_count
(count)		The number of gRPC requests processed over IPv4 and IPv6.
Shown as request
aws.applicationelb.healthy_host_count
(gauge)		Average number of healthy instances in each Availability Zone.
Shown as host
aws.applicationelb.healthy_host_count.maximum
(gauge)		Maximum number of healthy instances in each Availability Zone.
Shown as host
aws.applicationelb.healthy_host_count.minimum
(gauge)		Minimum number of healthy instances in each Availability Zone.
Shown as host
aws.applicationelb.healthy_host_count_deduped
(count)		The number of healthy instances per Availability Zone, regardless of if the Cross-Zone Load Balancing option is enabled or not.
Shown as host
aws.applicationelb.httpcode_elb_3xx
(count)		The number of HTTP 3XX redirection codes that originate from the load balancer.
Shown as response
aws.applicationelb.httpcode_elb_4xx
(count)		Number of HTTP 4XX client error codes generated by the load balancer.
Shown as response
aws.applicationelb.httpcode_elb_5_0_0
(count)		The number of HTTP 500 error codes that originate from the load balancer.
Shown as response
aws.applicationelb.httpcode_elb_5_0_2
(count)		The number of HTTP 502 error codes that originate from the load balancer.
Shown as response
aws.applicationelb.httpcode_elb_5_0_3
(count)		The number of HTTP 503 error codes that originate from the load balancer.
Shown as response
aws.applicationelb.httpcode_elb_5_0_4
(count)		The number of HTTP 504 error codes that originate from the load balancer.
Shown as response
aws.applicationelb.httpcode_elb_5xx
(count)		Number of HTTP 5XX client error codes generated by the load balancer.
Shown as response
aws.applicationelb.httpcode_redirect
(count)		The number of redirect actions that were successful.
Shown as response
aws.applicationelb.httpcode_target_2xx
(count)		Number of HTTP 2XX response codes generated by registered instances.
Shown as response
aws.applicationelb.httpcode_target_3xx
(count)		Number of HTTP 3XX response codes generated by registered instances.
Shown as response
aws.applicationelb.httpcode_target_4xx
(count)		Number of HTTP 4XX response codes generated by registered instances.
Shown as response
aws.applicationelb.httpcode_target_5xx
(count)		Number of HTTP 5XX response codes generated by registered instances.
Shown as response
aws.applicationelb.httpfixed_response
(count)		The number of fixed-response actions that were successful.
Shown as response
aws.applicationelb.httpredirect
(count)		The number of redirect actions that were successful.
aws.applicationelb.httpredirect_url_limit_exceeded
(count)		The number of redirect actions that couldn't be completed because the URL in the response location header is larger than 8K.
aws.applicationelb.ipv_6processed_bytes
(count)		The total number of bytes processed by the load balancer over IPv6.
Shown as byte
aws.applicationelb.ipv_6request_count
(count)		The number of IPv6 requests received by the load balancer.
Shown as request
aws.applicationelb.lambda_internal_error
(count)		The number of requests to a Lambda function that failed because of an issue internal to the load balancer or AWS Lambda.
Shown as request
aws.applicationelb.lambda_target_processed_bytes
(gauge)		The total number of bytes processed by the load balancer for requests to and responses from a Lambda function.
Shown as byte
aws.applicationelb.lambda_user_error
(count)		The number of requests to a Lambda function that failed because of an issue with the Lambda function.
Shown as request
aws.applicationelb.new_connection_count
(count)		The total number of new TCP connections established from clients to the load balancer and from the load balancer to targets.
Shown as connection
aws.applicationelb.non_sticky_request_count
(count)		The number of requests where the load balancer chose a new target because it couldn't use an existing sticky session.
Shown as request
aws.applicationelb.processed_bytes
(count)		The total number of bytes processed by the load balancer over IPv4 and IPv6.
Shown as byte
aws.applicationelb.rejected_connection_count
(count)		The number of connections that were rejected because the load balancer had reached its maximum number of connections.
Shown as connection
aws.applicationelb.request_count
(count)		Total number of completed requests that were received and routed to the registered instances. Not inclusive of HTTP 460, HTTP 400, and some kinds of HTTP 503 and 500.
Shown as request
aws.applicationelb.request_count_per_target
(count)		The average number of requests received by each target in a target group.
Shown as request
aws.applicationelb.rule_evaluations
(count)		The number of rules processed by the load balancer given a request rate averaged over an hour.
aws.applicationelb.target_connection_error_count
(count)		Number of connections that were not successfully established between the load balancer and the registered instances.
Shown as error
aws.applicationelb.target_response_time.average
(gauge)		Average time elapsed after the request leaves the load balancer until a response is received.
Shown as second
aws.applicationelb.target_response_time.maximum
(gauge)		Maximum time elapsed after the request leaves the load balancer until a response is received.
Shown as second
aws.applicationelb.target_response_time.p50
(gauge)		50th percentile of the time elapsed after the request leaves the load balancer until a response is received.
Shown as second
aws.applicationelb.target_response_time.p90
(gauge)		90th percentile of the time elapsed after the request leaves the load balancer until a response is received.
Shown as second
aws.applicationelb.target_response_time.p95
(gauge)		95th percentile of the time elapsed after the request leaves the load balancer until a response is received.
Shown as second
aws.applicationelb.target_response_time.p99
(gauge)		99th percentile of the time elapsed after the request leaves the load balancer until a response is received.
Shown as second
aws.applicationelb.target_tlsnegotiation_error_count
(count)		The number of TLS connections initiated by the load balancer that did not establish a session with the target.
Shown as connection
aws.applicationelb.un_healthy_host_count
(gauge)		Average number of unhealthy instances in each Availability Zone.
Shown as host
aws.applicationelb.un_healthy_host_count.maximum
(gauge)		Maximum number of unhealthy instances in each Availability Zone.
Shown as host
aws.applicationelb.un_healthy_host_count.minimum
(gauge)		Minimium number of unhealthy instances in each Availability Zone.
Shown as host
aws.applicationelb.un_healthy_host_count_deduped
(count)		The number of unhealthy instances per Availability Zone, regardless of if the Cross-Zone Load Balancing option is enabled or not.
Shown as host
aws.elb.active_connection_count
(count)		The total number of concurrent TCP connections active from clients to the load balancer and from the load balancer to targets.
Shown as connection
aws.elb.backend_connection_errors
(rate)		Number of connections that were not successfully established between the load balancer and the registered instances.
Shown as error
aws.elb.client_tlsnegotiation_error_count
(count)		Number of TLS negotiation errors
Shown as error
aws.elb.consumed_lbcapacity_units
(gauge)		Number of ELB capacity units consumed
Shown as unit
aws.elb.consumed_lcus
(gauge)		The number of load balancer capacity units (LCU) used by your load balancer.
Shown as unit
aws.elb.estimated_albactive_connection_count
(count)		The estimated total number of concurrent TCP connections active from clients to the load balancer and from the load balancer to targets.
Shown as connection
aws.elb.estimated_albconsumed_lcus
(gauge)		The estimated total number of load balancer capacity units (LCU) used by an Application Load Balancer.
Shown as unit
aws.elb.estimated_albnew_connection_count
(count)		The estimated total number of new TCP connections established from clients to the load balancer and from the load balancer to targets
Shown as connection
aws.elb.estimated_processed_bytes
(count)		The estimated total number of bytes processed by an Application Load Balancer.
Shown as byte
aws.elb.healthy_host_count
(gauge)		Average number of healthy instances in each Availability Zone.
Shown as host
aws.elb.healthy_host_count.maximum
(gauge)		Maximum number of healthy instances in each Availability Zone.
Shown as host
aws.elb.healthy_host_count.minimum
(gauge)		Minimum number of healthy instances in each Availability Zone.
Shown as host
aws.elb.healthy_host_count_deduped
(gauge)		The number of healthy instances per Availability Zone, regardless of if the Cross-Zone Load Balancing option is enabled or not.
Shown as host
aws.elb.httpcode_backend_2xx
(rate)		Number of HTTP 2XX response codes generated by registered instances.
Shown as response
aws.elb.httpcode_backend_3xx
(rate)		Number of HTTP 3XX response codes generated by registered instances.
Shown as response
aws.elb.httpcode_backend_4xx
(rate)		Number of HTTP 4XX response codes generated by registered instances.
Shown as response
aws.elb.httpcode_backend_5xx
(rate)		Number of HTTP 5XX response codes generated by registered instances.
Shown as response
aws.elb.httpcode_elb_4xx
(rate)		Number of HTTP 4XX client error codes generated by the load balancer.
Shown as response
aws.elb.httpcode_elb_5_0_0
(count)		The number of HTTP 500 error codes that originate from the load balancer.
Shown as response
aws.elb.httpcode_elb_5_0_2
(count)		The number of HTTP 502 error codes that originate from the load balancer.
Shown as response
aws.elb.httpcode_elb_5_0_3
(count)		The number of HTTP 503 error codes that originate from the load balancer.
Shown as response
aws.elb.httpcode_elb_5_0_4
(count)		The number of HTTP 504 error codes that originate from the load balancer.
Shown as response
aws.elb.httpcode_elb_5xx
(rate)		Number of HTTP 5XX client error codes generated by the load balancer.
Shown as response
aws.elb.httpcode_redirect
(count)		The number of redirect actions that were successful.
Shown as response
aws.elb.httpcode_target_2xx
(count)		Number of HTTP 2XX response codes generated by the targets.
Shown as response
aws.elb.httpcode_target_3xx
(count)		Number of HTTP 3XX response codes generated by the targets.
Shown as response
aws.elb.httpcode_target_4xx
(count)		Number of HTTP 4XX response codes generated by the targets.
Shown as response
aws.elb.httpcode_target_5xx
(count)		Number of HTTP 5XX response codes generated by the targets.
Shown as response
aws.elb.ipv_6processed_bytes
(count)		The total number of bytes processed by the load balancer over IPv6.
Shown as byte
aws.elb.ipv_6request_count
(count)		The number of IPv6 requests received by the load balancer.
Shown as request
aws.elb.latency
(gauge)		Average time elapsed after the request leaves the load balancer until a response is received. (ELB v1)
Shown as second
aws.elb.latency.maximum
(gauge)		Maximum time elapsed after the request leaves the load balancer until a response is received. (ELB v1)
Shown as second
aws.elb.latency.minimum
(gauge)		Minimum time elapsed after the request leaves the load balancer until a response is received. (ELB v1)
Shown as second
aws.elb.latency.p95
(gauge)		95th percentile of the time elapsed after the request leaves the load balancer until a response is received. (ELB v1)
Shown as second
aws.elb.latency.p99
(gauge)		99th percentile of the time elapsed after the request leaves the load balancer until a response is received. (ELB v1)
Shown as second
aws.elb.new_connection_count
(count)		The total number of new TCP connections established from clients to the load balancer and from the load balancer to targets.
Shown as connection
aws.elb.processed_bytes
(count)		The total number of bytes processed by the load balancer over IPv4 and IPv6.
Shown as byte
aws.elb.request_count
(rate)		Total number of completed requests that were received and routed to the registered instances.
Shown as request
aws.elb.request_count_per_target
(count)		The average number of requests received by each target in a target group.
Shown as request
aws.elb.rule_evaluations
(count)		The number of rules processed by the load balancer given a request rate averaged over an hour.
aws.elb.spillover_count
(rate)		Total number of requests that were rejected because the queue was full.
Shown as request
aws.elb.spillover_count.maximum
(rate)		Maximum number of requests that were rejected because the queue was full per load balancer node.
Shown as request
aws.elb.surge_queue_length
(gauge)		Maximum number of requests that are pending submission to a registered instance.
Shown as request
aws.elb.target_connection_error_count
(count)		Number of connections that were not successfully established between the load balancer and the registered instances.
Shown as error
aws.elb.target_response_time.average
(gauge)		Average time elapsed after the request leaves the load balancer until a response is received. Identical to aws.applicationelb.target_response_time.average.
Shown as second
aws.elb.target_response_time.maximum
(gauge)		Maximum time elapsed after the request leaves the load balancer until a response is received. Identical to aws.applicationelb.target_response_time.maximum.
Shown as second
aws.elb.target_response_time.p95
(gauge)		95th percentile of the time elapsed after the request leaves the load balancer until a response is received. Identical to aws.applicationelb.target_response_time.p95.
Shown as second
aws.elb.target_response_time.p99
(gauge)		99th percentile of the time elapsed after the request leaves the load balancer until a response is received. Identical to aws.applicationelb.target_response_time.p99.
Shown as second
aws.elb.un_healthy_host_count
(gauge)		Average number of unhealthy instances in each Availability Zone.
Shown as host
aws.elb.un_healthy_host_count.maximum
(gauge)		Maximum number of unhealthy instances in each Availability Zone.
Shown as host
aws.elb.un_healthy_host_count.minimum
(gauge)		Minimium number of unhealthy instances in each Availability Zone.
Shown as host
aws.elb.un_healthy_host_count_deduped
(gauge)		The number of unhealthy instances per Availability Zone, regardless of if the Cross-Zone Load Balancing option is enabled or not.
Shown as host
aws.networkelb.active_flow_count
(gauge)		The average number of active established connections from clients to targets
Shown as connection
aws.networkelb.active_flow_count.maximum
(gauge)		The maximum number of active established connections from clients to targets
Shown as connection
aws.networkelb.active_flow_count.minimum
(gauge)		The minimum number of active established connections from clients to targets
Shown as connection
aws.networkelb.active_flow_count_tcp
(count)		The average number of concurrent TCP flows (or connections) from clients to targets.
Shown as connection
aws.networkelb.active_flow_count_tls
(count)		The average number of concurrent TLS flows (or connections) from clients to targets.
Shown as connection
aws.networkelb.active_flow_count_udp
(count)		The average number of concurrent UDP flows (or connections) from clients to targets.
Shown as connection
aws.networkelb.client_tlsnegotiation_error_count
(count)		The total number of TLS handshakes that failed during negotiation between a client and a TLS listener.
Shown as error
aws.networkelb.consumed_lcus
(count)		The number of LCUs used by the load balancer.
Shown as unit
aws.networkelb.consumed_lcus_tcp
(count)		The number of LCUs used by the load balancer for TCP.
Shown as unit
aws.networkelb.consumed_lcus_tls
(count)		The number of LCUs used by the load balancer for TLS.
Shown as unit
aws.networkelb.consumed_lcus_udp
(count)		The number of LCUs used by the load balancer for UDP.
Shown as unit
aws.networkelb.healthy_host_count
(gauge)		Average number of healthy targets
Shown as host
aws.networkelb.healthy_host_count.maximum
(gauge)		Maximum number of healthy targets
Shown as host
aws.networkelb.healthy_host_count.minimum
(gauge)		Minimum number of healthy targets
Shown as host
aws.networkelb.new_flow_count
(count)		The number of new TCP connections from clients to targets
Shown as connection
aws.networkelb.new_flow_count_tcp
(count)		The total number of new TCP flows (or connections) established from clients to targets in the time period.
Shown as connection
aws.networkelb.new_flow_count_tls
(count)		The total number of new TLS flows (or connections) established from clients to targets in the time period.
Shown as connection
aws.networkelb.new_flow_count_udp
(count)		The total number of new UDP flows (or connections) established from clients to targets in the time period.
Shown as connection
aws.networkelb.peak_packets_per_second
(gauge)		The highest average packet rate from the six 10-second measurement windows in the time period.
Shown as packet
aws.networkelb.port_allocation_error_count
(count)		The total number of ephemeral port allocation errors during a client IP translation operation.
Shown as error
aws.networkelb.processed_bytes
(count)		The total number of bytes processed by the load balancer, including TCP/IP headers.
Shown as byte
aws.networkelb.processed_bytes_tcp
(count)		The total number of bytes processed by TCP listeners.
Shown as byte
aws.networkelb.processed_bytes_tls
(count)		The total number of bytes processed by TLS listeners.
Shown as byte
aws.networkelb.processed_bytes_udp
(count)		The total number of bytes processed by UDP listeners.
Shown as byte
aws.networkelb.processed_packets
(count)		The total number of packets processed by the load balancer.
Shown as packet
aws.networkelb.target_tlsnegotiation_error_count
(count)		The total number of TLS handshakes that failed during negotiation between a TLS listener and a target.
Shown as error
aws.networkelb.tcpclient_reset_count
(count)		The number of reset (RST) packets created by a client and sent to a target
Shown as packet
aws.networkelb.tcpelbreset_count
(count)		The number of reset (RST) packets created by a load balancer
Shown as packet
aws.networkelb.tcptarget_reset_count
(count)		The number of reset (RST) packets created by a target and sent to a client
Shown as packet
aws.networkelb.un_healthy_host_count
(gauge)		Average number of unhealthy targets
Shown as host
aws.networkelb.un_healthy_host_count.maximum
(gauge)		Maximum number of unhealthy targets
Shown as host
aws.networkelb.un_healthy_host_count.minimum
(gauge)		Minimum number of unhealthy targets
Shown as host

AWS から取得される各メトリクスには、ホスト名やセキュリティ グループなど、AWS コンソールに表示されるのと同じタグが割り当てられます。

イベント

Amazon Elastic Load Balancing インテグレーションには、イベントは含まれません。

サービスのチェック

Amazon Elastic Load Balancing インテグレーションには、サービスのチェック機能は含まれません。

トラブルシューティング

ご不明な点は、Datadog のサポートチームまでお問合せください。

その他の参考資料

お役に立つドキュメント、リンクや記事:

上位の ELB ヘルスメトリクスとパフォーマンスメトリクスブログ more more