咨询：可通过HTTP传输Linux服务器状态的免费工具及相关API

Great question—there are plenty of free, open-source tools that let you expose Linux server metrics (CPU/memory usage, open ports, running services, etc.) via a REST API, which you can easily query with curl or Postman to get JSON-formatted responses. Here are the most reliable options, broken down by ease of use and functionality:

1. Netdata #

• Why it’s great: My go-to for quick, no-fuss monitoring APIs. It’s super lightweight (uses barely any system resources) and exposes a full REST API right after installation—no extra config needed for basic access. It tracks every metric you mentioned: CPU/memory utilization, running services, open ports, disk I/O, network stats, and even per-process resource usage.
• Quick example calls:
  • Fetch CPU usage data: curl http://your-server-ip:19999/api/v1/data?chart=system.cpu
  • Get running services: curl http://your-server-ip:19999/api/v1/data?chart=services
• Bonus: It comes with a beautiful real-time dashboard, but you can ignore that entirely and just use the API.

2. Glances #

• Why it’s great: A cross-platform monitoring tool built with Python, which makes it dead simple to set up. It has a dedicated API mode that returns clean, concise JSON for all core system metrics.
• How to get started:
  1. Install Glances (via pip install glances or your package manager)
  2. Launch it in API mode: glances -w
  3. Query the API:
     • Get all metrics at once: curl http://your-server-ip:61208/api/3/all
     • Fetch just memory stats: curl http://your-server-ip:61208/api/3/mem
• Note: It includes open port data under the network endpoint, and running processes under processes.

3. Prometheus + Node Exporter #

• Why it’s great: The industry standard for scalable monitoring. Node Exporter scrapes all your Linux server metrics and exposes them via an HTTP endpoint, while Prometheus lets you query that data with its powerful PromQL language and return JSON results.
• How to get started:
  1. Install Node Exporter on your server—it exposes raw metrics on port 9100 by default (query with curl http://your-server-ip:9100/metrics for text-based output).
  2. Set up a Prometheus server to scrape Node Exporter, then use Prometheus’s API to run queries:
     • Example: Get CPU usage over the last 5 minutes: curl -G 'http://prometheus-ip:9090/api/v1/query_range' --data-urlencode 'query=node_cpu_seconds_total' --data-urlencode 'start=now-5m' --data-urlencode 'end=now'
• Note: This setup is a bit more involved, but it’s perfect if you need to aggregate, filter, or alert on metrics long-term.

4. ServerStatus #

• Why it’s great: A minimal, purpose-built tool designed exactly for exposing server status via a REST API. It’s lightweight, configurable, and returns only the metrics you specify in JSON.
• How to get started:
  1. Clone the open-source repo, configure config.json to pick which metrics to expose (CPU, memory, disk, etc.)
  2. Run the server, then hit the endpoint (usually http://your-server-ip:8080/status) to get your JSON response.

Critical Security Tips #

Since these APIs expose sensitive server data, don’t skip these steps:

• Restrict API access to trusted IPs using your firewall (e.g., ufw allow from your-client-ip to any port 19999 for Netdata)
• Enable basic authentication if the tool supports it (Netdata, Glances, and Prometheus all have built-in options)
• Use HTTPS for internet-facing APIs—put a reverse proxy like Nginx in front with a Let’s Encrypt certificate.

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