跳到主要内容
版本:0.4 · 当前版

综合检查模块的检查

参考审阅日期:2026年10月11日
适用于:Excubius 0.4角色: 操作员, 编辑者, 实体管理员
计划Free · Pro · Enterprise部署自托管

synthetics 模块有 9 项检查。主动检查,用于演练此前出过故障的路径:从 VNet 内的采集器(私有位置)发起,也可以选择使用 New Relic Synthetics。

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
synthetics.http_endpoint
HTTP endpoint availability and response time
60scriticalwarning > 2000 milliseconds · 5m / critical > 10000 milliseconds · 2m
synthetics.tls_certificate
TLS certificate validity
3600swarningwarning < 30 days / critical < 7 days
synthetics.dns_resolve
DNS resolution of the endpoint host
300scriticalwarning > 500 milliseconds / critical > 2000 milliseconds
synthetics.ag_listener_resolve
Listener name resolves from the collector
120scriticalwarning > 500 milliseconds / critical > 2000 milliseconds
synthetics.domain_auth
Domain authentication round trip (AADDS)
300scriticalok = true
synthetics.app_to_sql
Application-to-SQL round trip through the listener
60scriticalwarning > 500 milliseconds / critical > 5000 milliseconds
synthetics.ssrs_report_render
SSRS report render test
900scriticalwarning > 15000 milliseconds / critical > 60000 milliseconds
synthetics.ssrs_execution_log
SSRS execution log failures
600swarningwarning >= 1 count / critical >= 5 count
synthetics.platform_heartbeat
Platform dead-man heartbeat
60scriticalcritical > 5

检查​

synthetics.http_endpoint

HTTP endpoint availability and response time

度量
response_ms (milliseconds) · 每 60s · http
规则
warning > 2000 milliseconds · 5m / critical > 10000 milliseconds · 2m · 默认严重性: critical
措施
—
基线
Status or content mismatch is critical regardless of latency.

synthetics.tls_certificate

TLS certificate validity

度量
days_to_expiry (days) · 每 3600s · http
规则
warning < 30 days / critical < 7 days · 默认严重性: warning
措施
—
基线
Chain validity, hostname match and minimum TLS version are hard failures.

synthetics.dns_resolve

DNS resolution of the endpoint host

度量
resolve_ms (milliseconds) · 每 300s · dns
规则
warning > 500 milliseconds / critical > 2000 milliseconds · 默认严重性: critical
措施
—
基线
Resolution failure (or an unexpected address) is critical; slow resolvers are a warning.

synthetics.ag_listener_resolve

Listener name resolves from the collector

度量
resolve_ms (milliseconds) · 每 120s · dns
规则
warning > 500 milliseconds / critical > 2000 milliseconds · 默认严重性: critical
措施
Check the listener's DNS record (RegisterAllProvidersIP, TTL) and the cluster network name resource.
基线
Applications reach the group through this name; a listener that does not resolve is an outage even when every replica is healthy.

synthetics.domain_auth

Domain authentication round trip (AADDS)

度量
latency_ms (milliseconds) · 每 300s · winrm
规则
ok = true · 默认严重性: critical
措施
—
基线
This is the check that would have caught the authentication outage. Critical on first failure.

synthetics.app_to_sql

Application-to-SQL round trip through the listener

度量
latency_ms (milliseconds) · 每 60s · tsql
规则
warning > 500 milliseconds / critical > 5000 milliseconds · 默认严重性: critical
措施
—
基线
Connects to the listener name, not a node, so DNS, cluster IP and primary role are all exercised.

synthetics.ssrs_report_render

SSRS report render test

度量
response_ms (milliseconds) · 每 900s · http
规则
warning > 15000 milliseconds / critical > 60000 milliseconds · 默认严重性: critical
措施
—
基线
HTTP status other than the expected one is critical.

synthetics.ssrs_execution_log

SSRS execution log failures

度量
failed_executions (count) · 每 600s · tsql
规则
warning >= 1 count / critical >= 5 count · 默认严重性: warning
措施
—
基线
Subscriptions that fail silently show up here; group by report path.

synthetics.platform_heartbeat

Platform dead-man heartbeat

度量
heartbeat · 每 60s · http
规则
critical > 5 · 默认严重性: critical
措施
—
基线
An external watcher (healthchecks-style, e.g. Uptime Kuma or Azure Monitor availability test) must alert if the platform stops pinging.