Once you’ve integrated with AWS CloudWatch, you have access to metrics from AWS ELB, which distributes your incoming traffic over a variety of targets, including IP addresses, EC2 instances, and containers in one or more Availability Zones.

All available AWS integrations

To verify metrics are reporting, search for the metrics on the Metric details page in Project settings.

The following table shows the ELB metrics ingested by Lightstep.

Metric Name Unit Description
aws.elb.backend_connection_errors count The number of connection attempts between the load balancer and the registered instances that failed. This count may be more than the request rate since the load balancer retries on some errors. Note that connection errors for health checks are also included in this count.
aws.elb.desync_mitigation_mode_non_compliant_request_count count The quantity of requests that don't follow RFC 7230 [HTTP listener].
aws.elb.healthy_host_count count The quantity of healthy instances that your load balancer has registered. After passing the initial health check, a freshly registered instance is said to be in good condition. The number of healthy instances for the LoadBalancerName dimension is determined across all Availability Zones if cross-zone load balancing is enabled. If not, it is determined based on Availability Zone.
aws.elb.http_code_backend_2xx count The quantity of 2XX HTTP response codes produced by registered instances [HTTP listener]. Any response codes generated by the load balancer are not included in this total.
aws.elb.http_code_backend_3xx count The quantity of 3XX HTTP response codes produced by registered instances [HTTP listener]. Any response codes generated by the load balancer are not included in this total.
aws.elb.http_code_backend_4xx count The quantity of 4XX HTTP response codes produced by registered instances [HTTP listener]. Any response codes generated by the load balancer are not included in this total.
aws.elb.http_code_backend_5xx count The quantity of 5XX HTTP response codes produced by registered instances [HTTP listener]. Any response codes generated by the load balancer are not included in this total.
aws.elb.http_code_elb_4xx count The quantity of HTTP 4XX client error codes produced by the load balancer [HTTP listener]. Any incomplete or improper requests result in client errors.
aws.elb.http_code_elb_5xx count The quantity of HTTP 5XX server error codes produced by the load balancer [HTTP listener]. Any response codes generated by the registered instances are not included in this total. If there are no healthy instances registered to the load balancer, or if the rate of requests exceeds the load balancer's or the instances' capacity (spillover), the metric is reported.
aws.elb.latency seconds The time in seconds between the load balancer sending a request to a registered instance and the instance beginning to deliver response headers.
aws.elb.request_count count The quantity of connections or successfully completed requests during the specified time period The quantity of requests that were received, routed, and that included HTTP error responses from the registered instances The number of connections to the registered instances made by the TCP listener.
aws.elb.spillover_count count The number of requests which were refused because the surge queue was full.
aws.elb.surge_queue_length count The total number of connections or HTTP requests that are waiting to be forwarded to a healthy instance (TCP listener). The queue can grow to a maximum of 1,024. When the queue is full, new connections or requests are declined. See SpilloverCount for more details.
aws.elb.un_healthy_host_count count The quantity of harmful instances that your load balancer has recorded. After exceeding the unhealthy threshold defined for health checks, an instance is deemed unhealthy. Once an unhealthy instance reaches the healthy threshold set for health checks, it is once more regarded as healthy.
aws.elb.estimated_alb_active_connection_count count The estimated quantity of active TCP connections at once, both from clients and from the load balancer to the targets.
aws.elb.estimated_alb_consumed_lcus count The anticipated amount of load balancer capacity units (LCU) that an application load balancer will use. You are charged based on how many LCUs you consume each hour. See Elastic Load Balancing Pricing for more details.
aws.elb.estimated_alb_new_connection_count count A rough estimate of the quantity of brand-new TCP connections made by clients and load balancers, respectively.
aws.elb.estimated_processed_bytes bytes The estimated amount of data that an application load balancer will process.
aws.elb.availability_zone count Utilizes the specified Availability Zone to filter the metric data.
aws.elb.load_balancer_name count Specifies the load balancer to be used to filter the metric data.