This topic describes the metrics that are extracted from the Layer 7 access logs of Application Load Balancer (ALB). The metrics include global metrics and the metrics that are specific to app_lb_id, status, and upstream_status.
The metrics in this topic use the time series data format. You can execute PromQL or SQL statements to query and analyze the metrics. For more information, see Overview of query and analysis on metric data.
Global metrics
The following table describes the global metrics.
Metric | Description |
---|---|
pv | The total number of visits. |
body_bytes_sent_avg | The average number of bytes in the bodies of the responses that are sent to clients. |
body_bytes_sent_sum | The total number of bytes in the bodies of the responses that are sent to clients. |
request_length_avg | The average length of requests. |
request_length_sum | The total length of requests. |
request_time_avg | The average period of time that is consumed to respond to a request. |
request_time_p50 | The period of time that is consumed to respond to a request at the 50th percentile among all periods of time. |
request_time_p90 | The period of time that is consumed to respond to a request at the 90th percentile among all periods of time. |
request_time_p99 | The period of time that is consumed to respond to a request at the 99th percentile among all periods of time. |
upstream_response_time_avg | The average period of time for a request. Note The upstream_response_time metric indicates the interval between the time when an ALB instance is connected to the backend server and the time when the ALB instance is disconnected from the backend server after the required data is received. |
upstream_response_time_p50 | The period of time at the 50th percentile among the periods of time for all requests. |
upstream_response_time_p90 | The period of time at the 90th percentile among the periods of time for all requests. |
upstream_response_time_p99 | The period of time at the 99th percentile among the periods of time for all requests. |
Metrics specific to app_lb_id
The tag of the metrics specific to app_lb_id is app_lb_id. The following table describes the metrics.
Metric | Unit | Description | Tag |
---|---|---|---|
pv:app_lb_id | count | The number of page views (PVs) for the ALB instance. | app_lb_id |
body_bytes_sent_avg:app_lb_id | byte | The average number of bytes in the bodies of the HTTP responses that are sent to the client. | app_lb_id |
body_bytes_sent_sum:app_lb_id | byte | The total number of bytes in the bodies of the HTTP responses that are sent to the client. | app_lb_id |
request_length_avg:app_lb_id | byte | The average length of the requests. | app_lb_id |
request_length_sum:app_lb_id | byte | The total length of the requests. | app_lb_id |
request_time_avg:app_lb_id | second | The average duration of the requests. | app_lb_id |
request_time_p50:app_lb_id | second | The 50th percentile value for the request durations. | app_lb_id |
request_time_p90:app_lb_id | second | The 90th percentile value for the request durations. | app_lb_id |
request_time_p99:app_lb_id | second | The 99th percentile value for the request durations. | app_lb_id |
upstream_response_time_avg:app_lb_id | second | The average duration of the request connections. Note The upstream_response_time field specifies the interval between the time when the ALB instance connects to the backend server and the time when the ALB instance disconnects from the backend server after the required data is received. | app_lb_id |
upstream_response_time_p50:app_lb_id | second | The 50th percentile value for the request connection durations. | app_lb_id |
upstream_response_time_p90:app_lb_id | second | The 90th percentile value for the request connection durations. | app_lb_id |
upstream_response_time_p99:app_lb_id | second | The 99th percentile value for the request connection durations. | app_lb_id |
Metrics specific to status
The tag of the metrics specific to status is app_lb_id+host+status. The following table describes the metrics.
Metric | Unit | Description | Tag |
---|---|---|---|
pv:app_lb_id:host:status | count | The numbers of PVs that are measured in the app_lb_id, host, and status dimensions. | app_lb_id+host+status |
body_bytes_sent_avg:app_lb_id:host:status | byte | The average number of bytes in the bodies of the HTTP responses that are sent to the client. | app_lb_id+host+status |
body_bytes_sent_sum:app_lb_id:host:status | byte | The total number of bytes in the bodies of the HTTP responses that are sent to the client. | app_lb_id+host+status |
request_length_avg:app_lb_id:host:status | byte | The average length of the requests. | app_lb_id+host+status |
request_length_sum:app_lb_id:host:status | byte | The total length of the requests. | app_lb_id+host+status |
request_time_avg:app_lb_id:host:status | second | The average duration of the requests. | app_lb_id+host+status |
request_time_p50:app_lb_id:host:status | second | The 50th percentile value for the request durations. | app_lb_id+host+status |
request_time_p90:app_lb_id:host:status | second | The 90th percentile value for the request durations. | app_lb_id+host+status |
request_time_p99:app_lb_id:host:status | second | The 99th percentile value for the request durations. | app_lb_id+host+status |
upstream_response_time_avg:app_lb_id:host:status | second | The average duration of the request connections. Note The upstream_response_time field specifies the interval between the time when the ALB instance connects to the backend server and the time when the ALB instance disconnects from the backend server after the required data is received. | app_lb_id+host+status |
upstream_response_time_p50:app_lb_id:host:status | second | The 50th percentile value for the request connection durations. | app_lb_id+host+status |
upstream_response_time_p90:app_lb_id:host:status | second | The 90th percentile value for the request connection durations. | app_lb_id+host+status |
upstream_response_time_p99:app_lb_id:host:status | second | The 99th percentile value for the request connection durations. | app_lb_id+host+status |