单主机运行3个Nginx，Telegraf多实例监控配置求助

以下是适配3个Nginx实例的完整Telegraf配置，已添加标签区分不同实例，避免数据混淆：

# nginx-metrics and logs
# 第一个Nginx实例监控
[[inputs.nginx]]
   urls = ["https://example.com/nginx_status"]
   response_timeout = "5s"
   tags = ["instance=example1"]

[[inputs.tail]]
   name_override = "nginxlog"
   files = ["/var/www/example/logs/http-access.log"]
   from_beginning = false
   pipe = false
   data_format = "grok"
   grok_patterns = ["%{COMBINED_LOG_FORMAT}"]
   tags = ["instance=example1"]

[[inputs.exec]]
   commands = [
    "sh /etc/telegraf/nginx_stats.sh"
   ]
   timeout = "5s"
   name_override = "nginx_response_code"
   data_format = "json"
   tag_keys = ["resp_code"]
   tags = ["instance=example1"]

# 第二个Nginx实例监控
[[inputs.nginx]]
   urls = ["https://example2.com/nginx_status"] # 替换为第二个Nginx的status端点
   response_timeout = "5s"
   tags = ["instance=example2"]

[[inputs.tail]]
   name_override = "nginxlog"
   files = ["/var/www/example2/logs/http-access.log"]
   from_beginning = false
   pipe = false
   data_format = "grok"
   grok_patterns = ["%{COMBINED_LOG_FORMAT}"]
   tags = ["instance=example2"]

[[inputs.exec]]
   commands = [
     "sh /etc/telegraf/nginx_stats2.sh"
  ]
  timeout = "5s"
  name_override = "nginx_response_code"
  data_format = "json"
  tag_keys = ["resp_code"]
  tags = ["instance=example2"]

# 第三个Nginx实例监控
[[inputs.nginx]]
   urls = ["https://example3.com/nginx_status"] # 替换为第三个Nginx的status端点
   response_timeout = "5s"
   tags = ["instance=example3"]

[[inputs.tail]]
   name_override = "nginxlog"
   files = ["/var/www/example3/logs/http-access.log"]
   from_beginning = false
   pipe = false
   data_format = "grok"
   grok_patterns = ["%{COMBINED_LOG_FORMAT}"]
   tags = ["instance=example3"]

[[inputs.exec]]
   commands = [
     "sh /etc/telegraf/nginx_stats3.sh"
  ]
  timeout = "5s"
  name_override = "nginx_response_code"
  data_format = "json"
  tag_keys = ["resp_code"]
  tags = ["instance=example3"]

[[outputs.prometheus_client]]
    listen = "0.0.0.0:9125"

关键配置说明 #

• 实例区分：每个输入块添加tags = ["instance=xxx"]字段，确保Prometheus中能识别不同Nginx实例的监控数据
• 端点验证：确认每个Nginx的nginx_status端点可正常访问，若为HTTP协议则替换对应URL
• 脚本匹配：保证nginx_stats*.sh脚本分别统计对应Nginx的响应码数据，输出格式需为标准JSON
• 配置生效：修改配置后重启Telegraf服务：sudo systemctl restart telegraf

内容的提问来源于stack exchange，提问作者user3796291