Skip to content
Romain edited this page Sep 11, 2019 · 5 revisions

Disco-Kafka

Overview

Live detection of outage events using RIPE Atlas disconnection events. Kafka is used for all data processing and storage.

Architecture Diagram

Running everything

live

  • Run the producer for probe information frequently (probeDataProducer.py)
  • Run forever eventProducer.py (getting the connect/disconnect events)
  • Run src/run.py -t 7

with historical data

  • run eventProducer with a specific topic name and the start/end times (don't forget to set a large retention time for old data. By default retention time in eventProducer is one year)

  • change topic for probe archive?

  • start run with corresponding dates and topic

Clone this wiki locally