Build resilient systems at scale
28–30 October 2015 • Amsterdam, The Netherlands

Show me the metrics: OpenTSDB and OpenTSP at Betfair

James Brooks (Betfair)
16:10–16:50 Thursday, 29/10/2015
Location: Emerald Room
Average rating: ***..
(3.65, 23 ratings)
Slides:   1-PPTX 

Prerequisite Knowledge

An interest in monitoring systems

Description

Betfair have been using OpenTSDB to form one of the pillars of their internal monitoring systems for around three years. This talk will cover an introduction to the architecture of OpenTSDB, with illustrated examples of its usefulness as a troubleshooting and analytics tool.

In the second part of the talk, you will be introduced to OpenTSP, which is a new open source application developed at Betfair to work alongside OpenTSDB and work around some limitations that we discovered in it.

OpenTSP is a tool designed to manage gathering time series metrics from across an entire estate, enriching them with additional metadata, and then sending the aggregated stream of data to multiple endpoints.

OpenTSP (Time Series Pipeline) features include:

  • High compatibility with tcollector, the original TSDB collector
  • Forwarding throughput of >100k data points per second
  • User-defined routing topology with multiple write allows you to feed the datastream into many analysers at the same time
  • Centralised control and configuration of selected aspects of the pipeline
  • Collector plugins can be written in any language
Photo of James Brooks

James Brooks

Betfair

James Brooks is a member of the Site Reliability Engineering team at Betfair.