如何在 docker 容器中更改 prometheus 的命令行标志

Question

我按照 Jeff Geerling 的指南（https://www.jeffgeerling.com/blog/2021/monitor-your-internet-raspberry-pi）使用在 docker 容器中运行的 prometheus 和 grafana 安装互联网监控仪表板。

一切正常，但我注意到数据在 15 天后被删除。经过快速搜索，我发现这是 prometheus 中存储保留的默认设置。

我自己尝试了很多，但找不到更改此设置的方法。

尽管我发现本教程（https://mkezz.wordpress.com/2017/11/13/prometheus-command-line-flags-in-docker-service/）据我所知应该完全解决我的问题有但是不起作用。我收到错误：Error response from daemon: This node is not a swarm manager. Use "docker swarm init" or "docker swarm join" to connect this node to swarm and try again.运行提到的第一个命令时。

我还发现了这个问题（增加 Prometheus 存储保留），但我无法使用最佳答案，因为我的 Prometheus 正在 docker 容器中运行。

有没有一种简单的方法可以为普罗米修斯设置类似这样的命令行标志--storage.tsdb.retention.time=30d？

这是我第一次安装时下载的自述文件：

# Internet Monitoring Docker Stack with Prometheus + Grafana

> This repository is a fork from [maxandersen/internet-monitoring](https://github.com/maxandersen/internet-monitoring), tailored for use on a Raspberry Pi. It has only been tested on a Raspberry Pi 4 running Pi OS 64-bit beta.

Stand-up a Docker [Prometheus](http://prometheus.io/) stack containing Prometheus, Grafana with [blackbox-exporter](https://github.com/prometheus/blackbox_exporter), and [speedtest-exporter](https://github.com/MiguelNdeCarvalho/speedtest-exporter) to collect and graph home Internet reliability and throughput.

## Pre-requisites

Make sure Docker and [Docker Compose](https://docs.docker.com/compose/install/) are installed on your Docker host machine.

## Quick Start

`
git clone https://github.com/geerlingguy/internet-monitoring
cd internet-monitoring
docker-compose up -d
`

Go to [http://localhost:3030/d/o9mIe_Aik/internet-connection](http://localhost:3030/d/o9mIe_Aik/internet-connection) (change `localhost` to your docker host ip/name).



## Configuration

To change what hosts you ping you change the `targets` section in [/prometheus/pinghosts.yaml](./prometheus/pinghosts.yaml) file.

For speedtest the only relevant configuration is how often you want the check to happen. It is at 30 minutes by default which might be too much if you have limit on downloads. This is changed by editing `scrape_interval` under `speedtest` in [/prometheus/prometheus.yml](./prometheus/prometheus.yml).

Once configurations are done, run the following command:

    $ docker-compose up -d

That's it. docker-compose builds the entire Grafana and Prometheus stack automagically.

The Grafana Dashboard is now accessible via: `http://<Host IP Address>:3030` for example http://localhost:3030

username - admin
password - wonka (Password is stored in the `config.monitoring` env file)

The DataSource and Dashboard for Grafana are automatically provisioned.

If all works it should be available at http://localhost:3030/d/o9mIe_Aik/internet-connection - if no data shows up try change the timeduration to something smaller.

<center><img src="images/dashboard.png" width="4600" heighth="500"></center>



## Interesting urls

http://localhost:9090/targets shows status of monitored targets as seen from prometheus - in this case which hosts being pinged and speedtest. note: speedtest will take a while before it shows as UP as it takes about 30s to respond.

http://localhost:9090/graph?g0.expr=probe_http_status_code&g0.tab=1 shows prometheus value for `probe_http_status_code` for each host. You can edit/play with additional values. Useful to check everything is okey in prometheus (in case Grafana is not showing the data you expect).

http://localhost:9115 blackbox exporter endpoint. Lets you see what have failed/succeded.

http://localhost:9798/metrics speedtest exporter endpoint. Does take about 30 seconds to show its result as it runs an actual speedtest when requested.



## Thanks and a disclaimer

Thanks to @maxandersen for making the original project this fork is based on.

Thanks to @vegasbrianc work on making a [super easy docker](https://github.com/vegasbrianc/github-monitoring) stack for running prometheus and grafana.

This setup is not secured in any way, so please only use on non-public networks, or find a way to secure it on your own.

经过进一步修改后，我发现了一个 docker-compose.yml 文件，我只是在commandprometheus 的 -section下添加了--storage.tsdb.retention.time=30d如下所示：

version: "3.1"

volumes:
    prometheus_data: {}
    grafana_data: {}

networks:
  front-tier:
  back-tier:

services:
  prometheus:
    image: prom/prometheus:v2.25.2
    restart: always
    volumes:
      - ./prometheus/:/etc/prometheus/
      - prometheus_data:/prometheus
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--web.console.libraries=/usr/share/prometheus/console_libraries'
      - '--web.console.templates=/usr/share/prometheus/consoles'
      - '--storage.tsdb.retention.time=30d'
    ports:
      - 9090:9090
    links:
      - ping:ping
      - speedtest:speedtest
    networks:
      - back-tier

  grafana:
    image: grafana/grafana
    restart: always
    volumes:
      - grafana_data:/var/lib/grafana
      - ./grafana/provisioning/:/etc/grafana/provisioning/
    depends_on:
      - prometheus
    ports:
      - 3030:3000
    env_file:
      - ./grafana/config.monitoring
    networks:
      - back-tier
      - front-tier

  ping:
    tty: true
    stdin_open: true
    expose:
      - 9115
    ports:
      - 9115:9115
    image: prom/blackbox-exporter
    restart: always
    volumes:
      - ./blackbox/config:/config
    command:
      - '--config.file=/config/blackbox.yml'
    networks:
      - back-tier

  speedtest:
    tty: true
    stdin_open: true
    expose:
      - 9798
    ports:
      - 9798:9798
    image: miguelndecarvalho/speedtest-exporter
    restart: always
    networks:
      - back-tier

  nodeexp:
    privileged: true
    image: prom/node-exporter
    volumes:
      - /proc:/host/proc:ro
      - /sys:/host/sys:ro
      - /:/rootfs:ro
    ports:
      - 9100:9100
    restart: always
    command:
      - '--path.procfs=/host/proc'
      - '--path.sysfs=/host/sys'
      - --collector.filesystem.ignored-mount-points
      - "^/(sys|proc|dev|host|etc|rootfs/var/lib/docker/containers|rootfs/var/lib/docker/overlay2|rootfs/run/docker/netns|rootfs/var/lib/docker/aufs)($$|/)"
    networks:
      - back-tier

然后运行，docker-compose create然后我可以在“存储保留 30 天”下docker start internet-monitoring_prometheus_1看到。[Hostname of Server]:9090/status这是应该这样做的吗，因为我认为我找到了解决方案。