Skip to content

启用 Prometheus 指标(Enable Prometheus metrics)#

为了收集和展示指标，n8n 使用 prom-client 库。

🌐 To collect and expose metrics, n8n uses the prom-client library.

/metrics 端点默认是禁用的，但可以通过 N8N_METRICS 环境变量启用它。

🌐 The /metrics endpoint is disabled by default, but it's possible to enable it using the N8N_METRICS environment variable.

1
export N8N_METRICS=true

请参阅相应的环境变量（N8N_METRICS_INCLUDE_*）以配置应公开哪些指标和标签。

🌐 Refer to the respective Environment Variables (N8N_METRICS_INCLUDE_*) for configuring which metrics and labels should get exposed.

main 和 worker 实例都能够暴露指标。

🌐 Both main and worker instances are able to expose metrics.

队列指标(Queue metrics)#

要启用队列指标，请将 N8N_METRICS_INCLUDE_QUEUE_METRICS 环境变量设置为 true。你可以使用 N8N_METRICS_QUEUE_METRICS_INTERVAL 调整刷新频率。

🌐 To enable queue metrics, set the N8N_METRICS_INCLUDE_QUEUE_METRICS env var to true. You can adjust the refresh rate with N8N_METRICS_QUEUE_METRICS_INTERVAL.

n8n 从 Bull 收集这些指标并在主实例上显示。在多主节点设置中，当聚合查询时，您可以使用 instance_role_leader 指标来识别主节点，主节点设置为 1，其他节点设置为 0。

🌐 n8n gathers these metrics from Bull and exposes them on the main instances. On multi-main setups, when aggregating queries, you can identify the leader using the instance_role_leader gauge, set to 1 for the leader main and 0 otherwise.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
# HELP n8n_scaling_mode_queue_jobs_active Current number of jobs being processed across all workers in scaling mode.
# TYPE n8n_scaling_mode_queue_jobs_active gauge
n8n_scaling_mode_queue_jobs_active 0

# HELP n8n_scaling_mode_queue_jobs_completed Total number of jobs completed across all workers in scaling mode since instance start.
# TYPE n8n_scaling_mode_queue_jobs_completed counter
n8n_scaling_mode_queue_jobs_completed 0

# HELP n8n_scaling_mode_queue_jobs_failed Total number of jobs failed across all workers in scaling mode since instance start.
# TYPE n8n_scaling_mode_queue_jobs_failed counter
n8n_scaling_mode_queue_jobs_failed 0

# HELP n8n_scaling_mode_queue_jobs_waiting Current number of enqueued jobs waiting for pickup in scaling mode.
# TYPE n8n_scaling_mode_queue_jobs_waiting gauge
n8n_scaling_mode_queue_jobs_waiting 0