id="elastic集群监控配置">elastic集群监控配置

因为业务量不稳定时高时低

data-src="https://i-blog.csdnimg.cn/blog_migrate/7e761dcaa9954f839aa4460a7d6e0591.png#pic_center"

各个版本基本上变动不大

https://github.com/justwatchcom/elasticsearch_exporter

data-src="https://i-blog.csdnimg.cn/blog_migrate/25ad14a3807136164a4d689e64e478e4.png#pic_center"

--es.uri=http://10.128.128.51:9200

--es.all

--web.listen-address=:9114无报错之后查看

curl

http://127.0.0.1:9114/metrics

id="开始配置prometheus">开始配置prometheus

../elasticsearch_exporter-1.1.0.linux-amd64/elasticsearch.rules这个文件夹有监控示例,稍作修改即可使用

calculate

elasticsearch_filesystem_data_used_percent

100

(elasticsearch_filesystem_data_size_bytes

elasticsearch_filesystem_data_free_bytes)

/elasticsearch_filesystem_data_size_bytes

elasticsearch_filesystem_data_free_percent

100

elasticsearch_filesystem_data_used_percent#

alert

ElasticsearchTooFewNodesRunningIF

elasticsearch_cluster_health_number_of_nodes

<

{severity="critical"}ANNOTATIONS

{description="There

elasticsearch_jvm_memory_used_bytes{area="heap"}

elasticsearch_jvm_memory_max_bytes{area="heap"}

>

{severity="critical"}ANNOTATIONS

{description="The

class="language-bash">二进制安装的prometheus目录结构

[root@jmenv

service.json.bakprometheus自己有一个监控规则文档

当服务重启时会加载监控规则

将上文中的监控规则按照rule.txt中的格式稍作修改将修改后的文本复制进去,只要变量和配置格式不变的情况下可以自由发挥:

alert:

elasticsearch_cluster_health_number_of_nodes

<

warningannotations:description:

"There

elasticsearch_jvm_memory_used_bytes{area="heap"}

elasticsearch_jvm_memory_max_bytes{area="heap"}

>

warningannotations:description:

"The

class="language-bash">[root@jmenv

prometheus-2.12.0.linux-amd64]#

total

"/root/prometheus-2.12.0.linux-amd64/rules.txt"#

scrape

'service-node_export'file_sd_configs:-

files:-

/root/prometheus-2.12.0.linux-amd64/service.jsonrefresh_interval:

10s-

'elasticsearch'scrape_interval:

30sstatic_configs:-

是我启动的es的export地址与端口

prometheus-2.12.0.linux-amd64/prometheus.yml

添加job_name

'elasticsearch'scrape_interval:

30sstatic_configs:-