Metrics

How to export the metrics

Service-Center is compatible with the Prometheus standard. By default, the full metrics can be collected by accessing the /metrics API through the 30100 port.

If you want to customize the metrics configuration.

metrics:
  enable: true # enable to start metrics gather
  interval: 30s # the duration of collection
  exporter: prometheus # use the prometheus exporter
  prometheus:
    # optional, listen another ip-port and path if set, e.g. http://127.0.0.1:80/other
    listenURL:

Summary

FamilyName: service_center

Server

|metric|type|description| |:—|:—:|:—| |http_request_total|counter|The total number of received service requests.| |http_success_total|counter|Total number of requests responding to status code 2xx or 3xx.| |http_request_durations_microseconds|summary|The latency of http requests.| |http_query_per_seconds|gauge|TPS of http requests.|

Pub/Sub

|metric|type|description| |:—|:—:|:—| |notify_publish_total|counter|The total number of instance events.| |notify_publish_durations_microseconds|summary|The latency between the event generated in ServiceCenter and received by the client.| |notify_pending_total|counter|The total number of pending instances events.| |notify_pending_durations_microseconds|summary|The latency of pending instances events.| |notify_subscriber_total|counter|The total number of subscriber, e.g. Websocket, gRPC.|

Meta

|metric|type|description| |:—|:—:|:—| |db_heartbeat_total|counter|The total number of received instance heartbeats.| |db_heartbeat_durations_microseconds|summary|The latency of received instance heartbeats.| |db_domain_total|counter|The total number of domains.| |db_service_total|counter|The total number of micro-services.| |db_service_usage|gauge|The usage percentage of service quota.| |db_instance_total|counter|The total number of instances.| |db_instance_usage|gauge|The usage percentage of instances.| |db_schema_total|counter|The total number of schemas.| |db_framework_total|counter|The total number of SDK frameworks.|

Backend

|metric|type|description| |:—|:—:|:—| |db_backend_event_total|counter|The total number of received backend events, e.g. etcd, Mongo.| |db_backend_event_durations_microseconds|summary|The latency between received backend events and finish to build cache.| |db_dispatch_event_total|counter|The total number of dispatch events to resource handlers.| |db_dispatch_event_durations_microseconds|summary|The latency between received backend events and finish to dispatch.|

System

|metric|type|description| |:—|:—:|:—| |db_sc_total|counter|The total number of ServiceCenter instances.| |process_resident_memory_bytes||| |process_cpu_seconds_total||| |process_cpu_usage||| |go_threads||| |go_goroutines|||