-
Download the executable from
bin/
(be sure to select the one for your OS) -
Download
config.yml
andprefixes.yml
and place them in the same directory of the executable -
Modify
prefixes.yml
and add the prefixes you want to monitor -
Run the executable
-
See the alerts in
logs/reports-YYYY-MM-DD.log
In config.yml
you can find other reporting mechanisms (e.g. email and slack) in addition to logging to files.
Please uncomment the related section and configure according to your needs.
You can compose the tool with 3 main components: connectors, monitors, and reports.
Connectors retrieve/listen to the data from different sources and transform them to a common format. Monitors analyze the data flow and produce alerts. Different monitors try to detect different issues. Reports send/store the alerts, e.g. by email or to a file.
Possible connectors are:
-
connectorRIS, for real-time data from RIPE RIS (https://ris-live.ripe.net/)
-
connectorTest, for testing purposes, it provokes all types of alerting
Possible monitors are:
-
monitorHijack, for monitoring hijacks
-
monitorVisibility, for monitoring prefixes' visibility (you will get notified when withdrawals make monitored routes disappear). A threshold can be specified in config.yml to trigger an alert only if the issue is visible from a certain amount of peers.
-
monitorNewPrefix, for monitoring if new more specifics (of the monitored prefixes) start to be announced
Possible reports are:
-
reportEmail, to send alerts by email. SMTP configurations are in config.yml
-
reportFile, to log the alerts to files. File directory, format, and log rotation configurations are in config.yml
-
reportSlack, to send alerts in Slack. Hook URL is configurable in config.yml
To start development:
-
git clone this repo
-
install Node.js (version >= 10.16) and npm (installing node and npm)
-
execute
npm install
oryarn
to install all dependencies (installing yarn) -
run
npm run watch-and-serve
to run the application. At every file change it will self-reload.
-
npm run watch-and-serve
to run the application from source code and monitor for file changes -
npm run serve
to run the application from the source -
npm run test
to run the tests -
npm run build
to compile and build OS native applications
You can compose the tool with 3 main components: connectors, monitors, and reports. All connectors must extend the class Connector. Monitors extend the class Monitor. Reports extend the class Report. From the superclass they will inherit various generic methods while some specifically for the particular component have to be implemented. Reports don't receive only alerts but also the data that provoked such alerts (so you can store the data and replay the accident later).
In config.yml
, for each collection of components:
-
file
- refers to the file name which contains the class -
channels
- refer to what channel(s) will be used to send or receive messages -
params
- whatever parameters may be needed by the component at creation time