Skip to content

监控指标

GOST内部通过Prometheus的指标(Metrics)来提供监控数据。

开启监控

通过metrics参数来开启监控指标记录，默认不开启。

gost -L :8080 -metrics :9000

开启认证并设置选项

gost -L :8080 -metrics "user:pass@:9000?path=/metrics"

services:
- name: service-0
  addr: ":8080"
  handler:
    type: auto
  listener:
    type: tcp

metrics:
  addr: :9000
  # unix domain socket
  # addr: unix:///var/run/gost.sock
  path: /metrics
  auth:
    username: user
    password: pass
  auther: auther-0
metrics.addr (string)
监控指标HTTP API服务地址
metrics.path (string, default=/metrics)
API路径

身份认证

身份认证采用HTTP Basic Auth方式。

配置文件中通过auth或auther选项可以设置身份认证信息，如果设置了auther选项，auth选项则会被忽略。

gost -L :8080 -metrics "user:pass@:9000"

metrics:
  addr: :9000
  auth:
    username: user
    password: pass
  auther: auther-0

开启之后可以通过http://localhost:9000/metrics地址查看到指标数据。

指标示例

gost_chain_errors_total{chain="chain-0",host="host-0"} 1

gost_service_handler_errors_total{host="host-0",service="service-0"} 1

gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="0.005"} 0
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="0.01"} 0
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="0.025"} 0
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="0.05"} 0
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="0.1"} 0
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="0.25"} 1
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="0.5"} 1
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="1"} 1
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="2.5"} 1
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="5"} 1
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="10"} 1
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="15"} 1
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="30"} 2
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="60"} 2
gost_service_request_duration_seconds_bucket{host="host-0",service="service-0",le="+Inf"} 2
gost_service_request_duration_seconds_sum{host="host-0",service="service-0"} 15.172895206
gost_service_request_duration_seconds_count{host="host-0",service="service-0"} 2

gost_service_requests_in_flight{host="host-0",service="service-0"} 0

gost_service_requests_total{host="host-0",service="service-0"} 2

gost_service_transfer_input_bytes_total{host="host-0",service="service-0"} 1018

gost_service_transfer_output_bytes_total{host="host-0",service="service-0"} 7327

gost_services{host="host-0"} 1

指标说明

gost_services (type=gauge)
运行的服务数量
gost_service_requests_total (type=counter)
服务处理的请求总数
gost_service_transfer_input_bytes_total (type=counter)
服务接收到的数据字节数
gost_service_transfer_output_bytes_total (type=counter)
服务发送出的数据字节数
gost_service_requests_in_flight (type=gauge)
服务当前正在处理中的请求数
gost_service_request_duration_seconds_* (type=histogram)
服务请求处理的时长分布
gost_service_handler_errors_total (type=counter)
服务处理请求失败数
gost_chain_errors_total (type=counter)
转发链本身建立连接失败数

Prometheus

Prometheus配置文件prometheus.yaml需要在scrape_configs中增加一个Job。

global:
  scrape_interval: 15s 
# A list of scrape configurations.
scrape_configs:
  - job_name: 'gost'
    scrape_interval: 5s
    static_configs:
      - targets: ['127.0.0.1:9000']

Grafana Dashboard

你可以使用以下的Dashboard来呈现监控指标数据

https://grafana.com/grafana/dashboards/16037

GOST Dashboard

Comments