Skip to content

Latest commit

 

History

History
209 lines (142 loc) · 8.81 KB

installation.md

File metadata and controls

209 lines (142 loc) · 8.81 KB
Raw

Installation

Requirements

BGPalerter requires a machine with 4GB of RAM and swap enabled. If you are using small AWS EC2 instances (or similar), remember to activate the swap.

Running BGPalerter from binaries - Quick Setup

  1. Download the binary here (be sure to select the one for your OS)

  2. Execute the binary (e.g., chmod +x bgpalerter-linux-x64 && ./bgpalerter-linux-x64)
    The first time you run it, the auto-configuration will start.

Running BGPalerter from binaries - All steps

Linux

  1. Download the binary: wget https://github.com/nttgin/BGPalerter/releases/latest/download/bgpalerter-linux-x64

  2. Download config.yml.example as config.yml (in the same directory of the binary)

  3. Make the binary executable (e.g., chmod +x bgpalerter-linux-x64)

  4. Auto-configure it: ./bgpalerter-linux-x64 generate -a _YOUR_ASN_ -o prefixes.yml -i -m

  5. Run it: ./bgpalerter-linux-x64
    Or use nohup ./bgpalerter-linux-x64 & to leave it running after you close the terminal

Additionally, you can configure BGPalerter to run as a Linux Serivce

Mac

  1. Download the binary here.

  2. Download config.yml.example as config.yml (in the same directory of the binary)

  3. Make the binary executable (e.g., chmod +x bgpalerter-macos-x64)

  4. Auto-configure it: ./bgpalerter-macos-x64 generate -a _YOUR_ASN_ -o prefixes.yml -i -m

  5. Run it: ./bgpalerter-macos-x64

Windows

  1. Download the binary here.

  2. Download config.yml.example as config.yml (in the same directory of the binary)

  3. Open cmd (press ctrl + R and type cmd) and cd on the directory where you downloaded the binary (usually cd C:\Users\_USER_\Downloads)

  4. Run it: bgpalerter-win-x64.exe

Running BGPalerter from the source code

  1. Git clone this repo.

  2. Install Node.js (version >= 12) and npm (installing node and npm).

  3. Execute npm install to install all dependencies.

  4. Run npm run serve to run the application.

Build Debian Package

The git repo contains the debian scripts to build a debian package. Follow these steps:

  1. Git clone this repo;

  2. Build the package with your preferred tool; e.g., dpkg-buildpackage -us -uc;

  3. Install the package dpkg -i ../node-bgpalerter_x.x.x-1_all.deb.

Running BGPalerter in Docker

BGPalerter is available in Docker Hub here.

There are two main builds:

  • latest stable version for production monitoring;
  • dev reflects the last commit in the dev branch. Use this only for development purposes.

Additionally, each release has its own build in case you want to revert back to an older version.

To run the latest stable version of BGPalerter in Docker, do:

docker run -i --name bgpalerter \
  -v $(pwd)/volume:/opt/bgpalerter/volume \
  nttgin/bgpalerter:latest run serve -- --d /opt/bgpalerter/volume/

Alternatively, you can use docker-compose. Create a file named docker-compose.yml with the following content:

version: "3.8"

services:
  bgpalerter:
    image: nttgin/bgpalerter:latest
    command: run serve -- --d /opt/bgpalerter/volume/
    container_name: bgpalerter
    volumes:
      - "$(pwd)/volume:/opt/bgpalerter/volume"
    restart: always

After that you have to invoke sudo docker-compose up -d and the container should start up. You might want to have a look at the logs by using sudo docker-compose logs.

With this command, a new directory ./volume will be created in the current position. Such directory will contain all the persistent data that BGPalerter will generate, including configuration and alert logs.
You can specify another directory by changing the directory before the colon in the -v flag (e.g., -v _LOCATION_YOU_WANT_/volume:/opt/bgpalerter/volume).

The command above runs BGPalerter in interactive mode (-i flag), which is necessary if you want to run the auto configuration.

You should replace the flag -i with the flag -d, when:

  • You already have the configuration files config.yml and prefixes.yml, or you plan to create them by hand. Just place them into the volume directory.
  • You executed BGPalerter with the -i flag and the volume directory and the configuration files have been already generated.

For production monitoring we suggest to monitor the uptime of BGPalerter.
In case you want to monitor the uptime by using the uptimeApi (read more), you need to map the port in the docker configuration with the following command:

docker run -i --name bgpalerter \
  -v $(pwd)/volume:/opt/bgpalerter/volume \
  -p 8011:8011 \
  nttgin/bgpalerter:latest run serve -- --d /opt/bgpalerter/volume/

Alternatively, you can use docker-compose. Create a file named docker-compose.yml with the following content:

version: "3.8"

services:
  bgpalerter:
    image: nttgin/bgpalerter:latest
    command: run serve -- --d /opt/bgpalerter/volume/
    container_name: bgpalerter
    ports:
      - '8011:8011'
    volumes:
      - "$(pwd)/volume:/opt/bgpalerter/volume"
    restart: always

After that you have to invoke sudo docker-compose up -d and the container should start up. You might want to have a look at the logs by using sudo docker-compose logs.

The uptimeApi module has to be enabled in volume/config.yml as described here. Now you can monitor http://127.0.0.1:8011/status (e.g., in Nagios) to check the status of the BGPalerter monitoring. Such API may return a negative result when there is a misconfiguration or when BGPalerter failed to connect to the data repository.

Optionally, you can specify a health check in docker to auto-restart the container in case of prolonged failure.

docker run -i --name bgpalerter \
  -v $(pwd)/volume:/opt/bgpalerter/volume \
  --health-cmd='wget --quiet --tries=1 --spider http://127.0.0.1:8011/status || exit 1' \
  --health-timeout=2s \
  --health-retries=15 \
  --health-interval=60s \
  --restart unless-stopped \
  -p 8011:8011 \
  nttgin/bgpalerter:latest run serve -- --d /opt/bgpalerter/volume/

Alternatively, you can use docker-compose. Create a file named docker-compose.yml with the following content:

version: "3.8"

services:
  bgpalerter:
    image: nttgin/bgpalerter:latest
    command: run serve -- --d /opt/bgpalerter/volume/
    container_name: bgpalerter
    ports:
      - '8011:8011'
    volumes:
      - "$(pwd)/volume:/opt/bgpalerter/volume"
    restart: always
    healthcheck:
      test: ["CMD-SHELL", "wget --quiet --tries=1 --spider http://127.0.0.1:8011/status || exit 1"]
      interval: 60s
      timeout: 2s
      retries: 15

After that you have to invoke sudo docker-compose up -d and the container should start up. You might want to have a look at the logs by using sudo docker-compose logs.

This option does NOT replace proper monitoring. Just restarting the container will not assure you that the monitoring is working properly or that it will work again. You should always investigate failures and fix possible misconfiguration.

BGPalerter parameters

The execution of BGPalerter supports some parameters

Parameter Description
-v Show version number
-h Show help
-c To specify the config file to load (default ./config.yml)
-d To specify a directory that will contain all the files used by BGPalerter. See here before modifying this.
-t To test the configuration by generating fake BGP updates. This will start sending alerts on all the reports listening the hijack channel.

You can also use the same parameters with npm (if you are running the source code), in the following format npm run serve -- --h (replace h with the parameter you need).

Volume

BGPalerter writes/reads some files on disk (e.g., logs). The positions of these files is set in config.yml, where both absolute and relative paths can be used. However, the volume parameter can be used to modify this behavior and confine BGPalerter to a specific directory. This is mostly useful in two occasions: (1) when you want to create "virtual environments", e.g., you want to have a single installation of BGPalerter running multiple instances each operating confined in a different directory; or (2) when you are using docker and you want to create a persistent volume (see above).