Slide 1

Slide 1 text

Dr Nic Williams VP Developer Evangelism @drnic drnicwilliams.com

Slide 2

Slide 2 text

Open Source Logging?

Slide 3

Slide 3 text

Logstash

Slide 4

Slide 4 text

Kibana

Slide 5

Slide 5 text

From Servers to Search

Slide 6

Slide 6 text

Logstash Kibana

Slide 7

Slide 7 text

Logs & Events sys

Slide 8

Slide 8 text

Store & Search

Slide 9

Slide 9 text

Locate & Display

Slide 10

Slide 10 text

Why use Ruby? • Author’s preferred language at the time • Originally in PHP; rewritten in Ruby • Sinatra framework • Closer integration with Logstash

Slide 11

Slide 11 text

Why use JRuby? • Actual threads, better CPU usage • Debugging - jvisualvm, jstack • Single executable - including ElasticSearch • Some better libraries - Joda Time

Slide 12

Slide 12 text

Performance “Many users report 20k events/sec on a single m1.xlarge”

Slide 13

Slide 13 text

Data per day • 20k events per second (say 500b each) • 10Kb per second • 864Mb per day • Splunk is $2000/year for 500Mb per day, plus hosting costs

Slide 14

Slide 14 text

10-min walk thru • Go to logstash.net (DOT NET, OH NO!) • Click “docs” • Click “10-minute walkthrough”

Slide 15

Slide 15 text

Live better • Use great tools • Live better with your apps • Find many more great tools with JRuby/Ruby