Prometheus 查询指标集
编辑Prometheus 查询指标集编辑
这是 query 指标集，用于从 Prometheus 的查询 API 查询。
配置编辑
瞬时查询编辑
以下配置在单个时间点对 up 指标执行瞬时查询
- module: prometheus
period: 10s
hosts: ["localhost:9090"]
metricsets: ["query"]
queries:
- name: 'up'
path: '/api/v1/query'
params:
query: "up"
也可以使用更复杂的 PromQL 表达式，例如以下表达式，它计算过去 5 分钟内测量的每秒 HTTP 请求率。
- module: prometheus
period: 10s
hosts: ["localhost:9090"]
metricsets: ["query"]
queries:
- name: "rate_http_requests_total"
path: "/api/v1/query"
params:
query: "rate(prometheus_http_requests_total[5m])"
范围查询编辑
以下示例在 30 秒范围内以 15 秒的查询分辨率评估表达式 up
- module: prometheus
period: 10s
metricsets: ["query"]
hosts: ["node:9100"]
queries:
- name: "up_master"
path: "/api/v1/query_range"
params:
query: "up{node='master01'}"
start: "2019-12-20T23:30:30.000Z"
end: "2019-12-21T23:31:00.000Z"
step: 15s
有关指标集中每个字段的描述，请参阅 导出字段 部分。
以下是由此指标集生成的示例文档
{
"@timestamp": "2017-10-12T08:05:34.853Z",
"event": {
"dataset": "prometheus.query",
"duration": 115000,
"module": "prometheus"
},
"metricset": {
"name": "query",
"period": 10000
},
"prometheus": {
"labels": {
"__name__": "go_threads",
"instance": "localhost:9090",
"job": "prometheus"
},
"query": {
"go_threads": 18
}
},
"service": {
"address": "localhost:32769",
"type": "prometheus"
}
}