Will Anderson
24daf99f9c
Alloy's otelcol.receiver.otlp only has traces→Tempo and logs→Loki pipelines. No metrics output is configured, so /v1/metrics returns 404, flooding MCP server logs every minute. Disable Micrometer OTLP push; Prometheus scrapes metrics from the actuator endpoint instead.
19 lines
678 B
YAML
19 lines
678 B
YAML
apiVersion: v1
|
|
kind: ConfigMap
|
|
metadata:
|
|
name: neuron-prod-config
|
|
namespace: neuron-prod
|
|
data:
|
|
SPRING_PROFILES_ACTIVE: "prod"
|
|
NEURON_DATA_DIR: "/data"
|
|
NEURON_DB_PATH: "/data/neuron.db"
|
|
NEURON_STORAGE_PATH: "/data"
|
|
SERVER_TOMCAT_ACCESSLOG_ENABLED: "true"
|
|
# Prod: minimal surface — never expose internals
|
|
MANAGEMENT_ENDPOINTS_WEB_EXPOSURE_INCLUDE: "health,info"
|
|
MANAGEMENT_ENDPOINT_HEALTH_SHOW_DETAILS: "never"
|
|
# Disable OTLP metrics push — Alloy has no metrics pipeline configured.
|
|
# Logs (Loki) and traces (Tempo) work fine via otelcol.receiver.otlp.
|
|
# Using Prometheus scrape for metrics instead.
|
|
MANAGEMENT_OTLP_METRICS_EXPORT_ENABLED: "false"
|