Nacos monitor guide

Nacos 0.8.0 improves the monitoring system, supporting Nacos operation status monitoring through exposing metrics data access to third-party monitoring system. Currently, prometheus, elastic search and influxdb are supported. The docs introduce how prometheus and grafana monitor Nacos. Here is Nacos grafana monitoring page. You can find out for yourself how to use elastic search and influxdb.

Deploy Nacos cluster to expose metrics data

Deploy the Nacos cluster according to the deploy document

Configure the application. properties file to expose metrics data

management.endpoints.web.exposure.include=*

Access {ip}:8848/nacos/actuator/prometheus to see if metrics data can be accessed

Deploy prometheus to collect Nacos metrics data

Download the Prometheus version you want to install at the address of download prometheus

linux & mac

Decompress prometheus compression package

tar xvfz prometheus-*.tar.gz
cd prometheus-*

Modify configuration file prometheus.yml to collect Nacos metrics data

    metrics_path: '/nacos/actuator/prometheus'
    static_configs:
      - targets: ['{ip1}:8848','{ip2}:8848','{ip3}:8848']

Start prometheus service

./prometheus --config.file="prometheus.yml"

windows

Download the corresponding version of Windows and decompress it

Modify configuration file prometheus.yml to collect Nacos metrics data

    metrics_path: '/nacos/actuator/prometheus'
    static_configs:
      - targets:['{ip1}:8848','{ip2}:8848','{ip3}:8848']

Start prometheus service

prometheus.exe --config.file=prometheus.yml

By accessing http://{ip}:9090/graph, we can see the data collected by prometheus. By searching nacos_monitor in the search bar, we can find Nacos data to show the success of the data collection.

Deploy grafana to graphically display metrics data

Install grafana on the same machine as prometheus, and use yum to install grafana

mac

brew install grafana
brew services start grafana

linux

sudo yum install https://s3-us-west-2.amazonaws.com/grafana-releases/release/grafana-5.2.4-1.x86_64.rpm
sudo service grafana-server start

windows

Reference document：http://docs.grafana.org/installation/windows/

Access grafana: http://{ip}:3000

Configuring prometheus data source

Import Nacos grafana monitoring template

Nacos monitoring is divided into three modules:

nacos monitor shows core monitoring items
nacos detail shows the change curve of index
nacos alert is alerts about nacos

configure grafana alert

When Nacos runs out of order, Grafana can alert the person in charge. Grafana supports a variety of police alert. Mail, DingTalk and webhook are commonly used.

DingTalk alert

Configure DingTalk robots

Configure DingTalk robots url

Test alert

mail alert

Modify defaults.ini configuration file to add mail alerts

#################################### SMTP / Emailing ##########################
[smtp]
enabled = true
host = smtp.126.com:25
user = xxxxxx
password = xxxxx
;cert_file =
;key_file =
skip_verify = true
from_address = xxxxxx@126.com

[emails]
;welcome_email_on_sign_up = false

Configuration notification mailbox

meaning of Nacos metrics

jvm metrics

item	meaning
system_cpu_usage	cpu usage
system_load_average_1m	load
jvm_memory_used_bytes	jvm memory used(bytes)
jvm_memory_max_bytes	jvm memory max(bytes)
jvm_gc_pause_seconds_count	gc count
jvm_gc_pause_seconds_sum	gc time
jvm_threads_daemon	jvm threads count

Nacos metrics

item	meaning
http_server_requests_seconds_count	http requests count
http_server_requests_seconds_sum	http requests time
nacos_timer_seconds_sum	Nacos config notify time
nacos_timer_seconds_count	Nacos config notify count
nacos_monitor{name=‘longPolling’}	Nacos config connection count
nacos_monitor{name=‘configCount’}	Nacos configuration file count
nacos_monitor{name=‘dumpTask’}	Nacos config dump task count
nacos_monitor{name=‘notifyTask’}	Nacos config notify task count
nacos_monitor{name=‘getConfig’}	Nacos config read configuration count
nacos_monitor{name=‘publish’}	Nacos config update configuration count
nacos_monitor{name=‘ipCount’}	Nacos naming ip count
nacos_monitor{name=‘domCount’}	Nacos naming domain count(1.x version)
nacos_monitor{name=‘serviceCount’}	Nacos naming domain count(2.x version)
nacos_monitor{name=‘failedPush’}	Nacos naming push fail count
nacos_monitor{name=‘avgPushCost’}	Nacos naming push cost time(average)(ms)
nacos_monitor{name=‘leaderStatus’}	Nacos naming if node is leader
nacos_monitor{name=‘maxPushCost’}	Nacos naming push cost time(max)(ms)
nacos_monitor{name=‘mysqlhealthCheck’}	Nacos naming mysql health check count
nacos_monitor{name=‘httpHealthCheck’}	Nacos naming http health check count
nacos_monitor{name=‘tcpHealthCheck’}	Nacos naming tcp health check count

nacos exception

item	meaning
nacos_exception_total{name=‘db’}	database exception
nacos_exception_total{name=‘configNotify’}	Nacos config notify exception
nacos_exception_total{name=‘unhealth’}	Nacos config server health check exception
nacos_exception_total{name=‘disk’}	Nacos naming write disk exception
nacos_exception_total{name=‘leaderSendBeatFailed’}	Nacos naming leader send heart beat fail count
nacos_exception_total{name=‘illegalArgument’}	request argument illegal count
nacos_exception_total{name=‘nacos’}	Nacos inner exception

client metrics

item	meaning
nacos_monitor{name=‘subServiceCount’}	subscribed services count
nacos_monitor{name=‘pubServiceCount’}	published services count
nacos_monitor{name=‘configListenSize’}	listened configuration file count
nacos_client_request_seconds_count	request count
nacos_client_request_seconds_sum	request time

Nacos-Sync monitor

With the release of Nacos 0.9, Nacos-Sync 0.3 supports metrics monitoring. It can observe the running status of Nacos-Sync service through metrics data, and improve the monitoring capability of Nacos-Sync in production environment. Reference for the Construction of the Overall Monitoring System Nacos Monitoring Manual

grafana monitor Nacos-Sync

The same as Nacos monitoring, Nacos-Sync also provides monitoring templates to import monitoring Nacos-Sync templates

Nacos-Sync monitoring is also divided into three modules:

nacos-sync monitor shows core monitoring items
nacos-sync detail and alert shows monitoring curves and alarms.

Nacos-Sync metrics meaning

Nacos-Sync metrics is divided into JVM layer and application layer

jvm metrics

item	meaning
system_cpu_usage	cpu usage
system_load_average_1m	load
jvm_memory_used_bytes	jvm memory used(bytes)
jvm_memory_max_bytes	jvm memory max(bytes)
jvm_gc_pause_seconds_count	gc count
jvm_gc_pause_seconds_sum	gc time
jvm_threads_daemon	jvm threads count

application metrics

item	meaning
nacosSync_task_size	sync task count
nacosSync_cluster_size	cluster count
nacosSync_add_task_rt	add task time
nacosSync_delete_task_rt	delete task time
nacosSync_dispatcher_task	dispatcher task time
nacosSync_sync_task_error	sync task error count