GCP Cloud Load Balancing Metrics¶

This page lists the Load Balancing metrics collected by the GCP app.

In the table below, the metric type prefixes are omitted. For example, the full name of the backend_latencies metric, of type https, is loadbalancing.googleapis.com/https/backend_latencies.

HTTPS¶

Metric type prefix: loadbalancing.googleapis.com/https/

Metric	Description
`backend_latencies`	A distribution of the latency calculated from when the request was sent by the proxy to the backend until the proxy received from the backend the last byte of the response.
`backend_request_bytes_count`	The number of bytes sent as requests from external HTTP(S) load balancer to backends.
`backend_request_count`	The number of requests served by the backends of the external HTTP(S) load balancer.
`backend_response_bytes_count`	The number of bytes sent as responses from backends (or cache) to external HTTP(S) load balancer.
`frontend_tcp_rtt`	A distribution of the RTT measured for each connection between client and proxy.
`internal/backend_latencies`	A distribution of the latency calculated from when the request was sent by the internal HTTP(S) load balancer proxy to the backend until the proxy received from the backend the last byte of the response. Sampled every 60 seconds. After sampling, data is not visible for up to 90 seconds.
`internal/request_bytes_count`	The number of bytes sent as requests from clients to an internal HTTP(S) load balancer. Sampled every 60 seconds. After sampling, data is not visible for up to 90 seconds.
`internal/request_count`	The number of requests served by an internal HTTP(S) load balancer. Sampled every 60 seconds. After sampling, data is not visible for up to 90 seconds.
`internal/response_bytes_count`	The number of bytes sent as responses from an internal HTTP(S) load balancer to clients. Sampled every 60 seconds. After sampling, data is not visible for up to 90 seconds.
`internal/total_latencies`	A distribution of the latency calculated from when the request was received by the internal HTTP(S) load balancer proxy until the proxy got an acknowledgment from the client on the last response byte. Sampled every 60 seconds. After sampling, data is not visible for up to 90 seconds.
`request_bytes_count`	The number of bytes sent as requests from clients to an external HTTP(S) load balancer.
`request_count`	The number of requests served by external HTTP(S) load balancer.
`response_bytes_count`	The number of bytes sent as responses from external HTTP(S) load balancer to clients.
`total_latencies`	A distribution of the latency calculated from when the request was received by the external HTTP(S) load balancer proxy until the proxy got acknowledgment from the client on the last response byte.

No prefix¶

These metrics do not have a metric type prefix.

Metric	Description
`backend_latencies`	A distribution of the latency calculated from when the request was sent by the proxy to the backend until the proxy received from the backend the last byte of the response. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
`request_bytes_count`	The number of bytes sent as requests from clients to HTTP/S load balancer. Sampled every 60 seconds. After sampling, data is not visible for up to 90 seconds.
`request_count`	The number of requests served by HTTP/S load balancer. Sampled every 60 seconds. After sampling, data is not visible for up to 90 seconds.
`response_bytes_count`	The number of bytes sent as responses from an HTTP/S load balancer to clients. Sampled every 60 seconds. After sampling, data is not visible for up to 90 seconds.
`total_latencies`	A distribution of the latency calculated from when the request was received by the proxy until the proxy received acknowledgment from the client on the last response byte. Sampled every 60 seconds. After sampling, data is not visible for up to 90 seconds.

L3¶

Metric type prefix: loadbalancing.googleapis.com/l3/

Metric	Description
`external/egress_bytes_count`	The number of bytes sent from the backend of the external network load balancer to the client. For TCP flows, this metric counts the bytes on the application stream only.
`external/egress_packets_count`	The number of packets sent from the backend of the external network load balancer to the client.
`external/ingress_bytes_count`	The number of bytes sent from the client to the backend of the external network load balancer. For TCP flows, this metric counts the bytes on the application stream only.
`external/ingress_packets_count`	The number of packets sent from the client to the backend of the external network load balancer.
`external/rtt_latencies`	A distribution of the round trip time latency, measured over TCP connections for the external network load balancer.
`internal/egress_bytes_count`	The number of bytes sent from an internal TCP/UDP load balancer backend to the client (for TCP flows it’s counting bytes on application stream only).
`internal/egress_packets_count`	The number of packets sent from an internal TCP/UDP load balancer backend to the client of the flow.
`internal/ingress_bytes_count`	The number of bytes sent from client to internal TCP/UDP load balancer backend (for TCP flows it’s counting bytes on application stream only).
`internal/ingress_packets_count`	The number of packets sent from a client to an internal TCP/UDP load balancer backend.
`internal/rtt_latencies`	A distribution of RTT measured over TCP connections for internal TCP/UDP load balancer flows.

L4_PROXY¶

Metric type prefix: loadbalancing.googleapis.com/l4_proxy/

Metric	Description
`egress_bytes_count`	Number of bytes sent from a VM to a client using a proxy. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
`ingress_bytes_count`	Number of bytes sent from a client to a virtual machine (VM) using a proxy. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.
`tcp/closed_connections_count`	Number of connections terminated over TCP proxy or SSL proxy load balancer. Sampled every 60 seconds. After sampling, data is not visible for up to 210 seconds.

TCP_SSL_PROXY¶

Metric type prefix: loadbalancing.googleapis.com/tcp_ssl_proxy

Metric	Description
`closed_connections`	Number of connections terminated over TCP proxy or SSL proxy load balancer.
`egress_bytes_count`	Number of bytes sent from a VM to a client using a proxy.
`frontend_tcp_rtt`	A distribution of the smoothed RTT (in ms) measured by the proxy TCP stack, each minute application layer bytes pass from proxy to client.
`ingress_bytes_count`	Number of bytes sent from a client to a VM using a proxy.
`new_connections`	Number of connections created over TCP proxy or SSL proxy load balancer.
`open_connections`	Current number of outstanding connections through the TCP proxy or SSL proxy load balancer.