2019.11.14晚21：00直播《任务管理系统》。

一、简介

在上一篇Prometheus+Node_exporter+Grafana+Alertmanager 监控部署里，已经有了 promethues 的一套监控环境。但是这里有个问题，就是在 prometheus.yml 里配置需要监听的服务时，是按服务名写死的，如果后面增加了节点或者组件信息，就得手动修改此配置，并重启 promethues；那么能否动态的监听微服务呢？Prometheus 提供了多种动态服务发现的功能，这里以 consul 为例。

二、引入 consul 的好处

在没有使用 consul 服务自动发现的时候，我们需要频繁对 Prometheus 配置文件进行修改，无疑给运维人员带来很大的负担，还有可能直接变成一个"配置达人"，即使是配置达人也会存在人为失误的情况。

三、Prometheus 支持的多种服务发现机制

#Prometheus数据源的配置主要分为静态配置和动态发现, 常用的为以下几类:

static_configs: #静态服务发现

file_sd_configs: #文件服务发现

dns_sd_configs: DNS #服务发现

kubernetes_sd_configs: #Kubernetes 服务发现

consul_sd_configs: Consul #服务发现

...

#在监控kubernetes的应用场景中，频繁更新的pod，svc，等等资源配置应该是最能体现Prometheus监控目标自动发现服务的好处

四、基于 docker 的 consul 集群

这里使用 docker-compose 方式部署 consul 集群

cat > /data0/consul/docker-compose.yaml << \EOF

version: '2'

networks:

byfn:

services:

consul1:

image: consul

container_name: node1

volumes:

- /data0/consul/conf_with_acl:/consul/config

command: agent -server -bootstrap-expect=3 -node=node1 -bind=0.0.0.0 -client=0.0.0.0 -config-dir=/consul/config

networks:

- byfn

consul2:

image: consul

container_name: node2

volumes:

- /data0/consul/conf_with_acl:/consul/config

command: agent -server -retry-join=node1 -node=node2 -bind=0.0.0.0 -client=0.0.0.0 -config-dir=/consul/config

ports:

- 8500:8500

depends_on:

- consul1

networks:

- byfn

consul3:

image: consul

volumes:

- /data0/consul/conf_with_acl:/consul/config

container_name: node3

command: agent -server -retry-join=node1 -node=node3 -bind=0.0.0.0 -client=0.0.0.0 -config-dir=/consul/config

depends_on:

- consul1

networks:

- byfn

consul4:

image: consul

container_name: node4

volumes:

- /data0/consul/conf_with_acl:/consul/config

command: agent -retry-join=node1 -node=ndoe4 -bind=0.0.0.0 -client=0.0.0.0 -ui -config-dir=/consul/config

ports:

- 8501:8500

depends_on:

- consul2

- consul3

networks:

- byfn

consul5:

image: consul

container_name: node5

volumes:

- /data0/consul/conf_without_acl:/consul/config

command: agent -retry-join=node1 -node=ndoe5 -bind=0.0.0.0 -client=0.0.0.0 -ui -config-dir=/consul/config

ports:

- 8502:8500

depends_on:

- consul2

- consul3

networks:

- byfn

EOF

cd /data0/consul/

docker-compose up -d

五、使用接口注册服务

# 注册服务

curl -X PUT -d '{"id": "test1","name": "test1","address": "192.168.56.12","port": 9100,"tags": ["service"],"checks": [{"http": "http://192.168.56.12:9100/","interval": "5s"}]}' http://192.168.56.12:8502/v1/agent/service/register

# 查询指定节点以及指定的服务信息

root># curl http://192.168.56.12:8500/v1/catalog/service/test1

[{"ID":"62c9ea24-a464-ee3f-a7ac-44b608b2a9fc","Node":"ndoe5","Address":"172.18.0.6","Datacenter":"dc1","TaggedAddresses":{"lan":"172.18.0.6","wan":"172.18.0.6"},"NodeMeta":{"consul-network-segment":""},"ServiceKind":"","ServiceID":"test1","ServiceName":"test1","ServiceTags":["service"],"ServiceAddress":"192.168.56.12","ServiceWeights":{"Passing":1,"Warning":1},"ServiceMeta":{},"ServicePort":9100,"ServiceEnableTagOverride":false,"ServiceProxy":{"MeshGateway":{}},"ServiceConnect":{},"CreateIndex":261,"ModifyIndex":261}]

六、修改 prometheus 使用 consul 服务发现

cat > /home/prometheus/prometheus.yml <

scrape_configs:

# The job name is added as a label `job=` to any timeseries scraped from this config.

- job_name: 'prometheus'

static_configs:

- targets: ['localhost:9090']

# 匹配service关键字

- job_name: 'consul-prometheus'

consul_sd_configs:

- server: '192.168.56.12:8502'

services: []

EOF

#重启prometheus

docker restart prometheus

七、结合 grafana 最终展示

参考文档：

https://kuboard.cn/ (强烈推荐一位大佬新品，快速在 Kubernetes 上落地微服务)

作者：Lancger