New Relic 检查
newrelic 模块有 6 项检查。New Relic APM 应用程序和综合监视器的黄金信号,由 New Relic 数据源根据它已读取的实体摘要(NerdGraph entitySearch)提供。主机黄金指标则用于 Windows 和 Linux 主机检查。
本页在构建文档时根据产品的检查目录生成,因此规则、间隔和默认严重性始终与你正在阅读的版本一致。阈值是默认值;实体和资源可以覆盖这些阈值。
| 检查 | 间隔 | 默认严重性 | 规则 |
|---|---|---|---|
newrelic.apm_apdexApplication Apdex | 300s | warning | warning < 0.85 · 10m / critical < 0.7 · 5m |
newrelic.apm_error_rateApplication error rate | 300s | warning | warning > 2 percent · 10m / critical > 5 percent · 5m |
newrelic.apm_response_timeApplication response time | 300s | warning | warning > 1000 milliseconds · 10m / critical > 3000 milliseconds · 5m |
newrelic.apm_throughputApplication throughput | 300s | info | ±3σ over 7d (learning only) |
newrelic.monitor_failing_locationsSynthetic monitor failing locations | 300s | critical | warning >= 1 count / critical >= 2 count |
newrelic.monitor_success_rateSynthetic monitor success rate | 300s | warning | warning < 98 percent / critical < 90 percent |
检查
newrelic.apm_apdex
Application Apdex
- 度量
apdex· 每300s·newrelic- 规则
warning < 0.85 · 10m / critical < 0.7 · 5m· 默认严重性: warning- 措施
- —
- 基线
- Apdex 0.85 and above is good; below 0.7 users are frustrated. Each application's own Apdex T applies. · 来源
newrelic.apm_error_rate
Application error rate
- 度量
error_rate(percent) · 每300s·newrelic- 规则
warning > 2 percent · 10m / critical > 5 percent · 5m· 默认严重性: warning- 措施
- —
- 基线
- Share of transactions with errors (New Relic reports a ratio; stored as a percentage).
newrelic.apm_response_time
Application response time
- 度量
response_ms(milliseconds) · 每300s·newrelic- 规则
warning > 1000 milliseconds · 10m / critical > 3000 milliseconds · 5m· 默认严重性: warning- 措施
- —
- 基线
- Average across all transactions; tune per application (entity override).
newrelic.apm_throughput
Application throughput
- 度量
throughput_rpm(per_minute) · 每300s·newrelic- 规则
±3σ over 7d (learning only)· 默认严重性: info- 措施
- —
- 基线
- Requests per minute; a sudden drop usually means traffic is not reaching the application.
newrelic.monitor_failing_locations
Synthetic monitor failing locations
- 度量
failing_locations(count) · 每300s·newrelic- 规则
warning >= 1 count / critical >= 2 count· 默认严重性: critical- 措施
- —
- 基线
- One failing location can be the probe's network; two or more mean the endpoint is failing. · 来源
newrelic.monitor_success_rate
Synthetic monitor success rate
- 度量
success_rate(percent) · 每300s·newrelic- 规则
warning < 98 percent / critical < 90 percent· 默认严重性: warning- 措施
- —
- 基线
- Share of successful checks over New Relic's summary window.