A Whirlwind Tour of Etsy's Monitoring Stack

Slide 1

Slide 1 text

A Whirlwind Tour of Etsy's Monitoring Stack Daniel Schauenberg [email protected] @mrtazz

Slide 2

Slide 2 text

No content

Slide 3

Slide 3 text

@mrtazz

Slide 4

Slide 4 text

@mrtazz

Slide 5

Slide 5 text

@mrtazz Item by TheBackPackShoppe

Slide 6

Slide 6 text

How comfortable are you deploying a change right now?

Slide 7

Slide 7 text

“If this is your first day at Etsy, you deploy the site”

Slide 8

Slide 8 text

No content

Slide 9

Slide 9 text

No content

Slide 10

Slide 10 text

@mrtazz Ganglia • System level metrics • Instance per DC/environment • > 220k RRD files • Fully configured through Chef role attributes

Slide 11

Slide 11 text

@mrtazz Rainbow Graphs!

Slide 12

Slide 12 text

@mrtazz StatsD • Single instance on one server • Traffic mostly from 70 Web & 24 API servers • Node.js • Heavy Sampling • Graphite as backend

Slide 13

Slide 13 text

@mrtazz

Slide 14

Slide 14 text

@mrtazz Graphite • Application level metrics • 96G RAM, 20 Cores, 7.3T SSD RAID 10 • 525k metrics/minute • Mirrored Master/Master Setup • Functionally sharded relays

Slide 15

Slide 15 text

@mrtazz CNAME relays relays caches caches statsdtimers statsdcounts statsd chef logster fqld search generic

Slide 16

Slide 16 text

@mrtazz

Slide 17

Slide 17 text

@mrtazz

Slide 18

Slide 18 text

@mrtazz Syslog-Ng • Web, Search, Gearman, Photos, Nagios, Network, VPN • 1.2GB written/minute • Chef role attribute based config • Rule ordering!

Slide 19

Slide 19 text

No content

Slide 20

Slide 20 text

@mrtazz github.com/etsy/logster • Extract metrics from log files • Written in Python • Runs every minute via cron

Slide 21

Slide 21 text

@mrtazz Splunk • Indexes all of our log files • Easy search for patterns • Saved searches for interesting ones • Basically using it as a glorified grep

Slide 22

Slide 22 text

@mrtazz Logstash • Experiment status • Makes it easier integrate different sources • Easy to set up in dev environment • Trying to figure out where/how it fits into our infrastructure

Slide 23

Slide 23 text

@mrtazz Eventinator • Tracks all events in our infrastructure • Chef runs and changes • DNS changes • Network • Deploys • Server provisioning and decommissioning • ~ 12 million events in the last 2 years

Slide 24

Slide 24 text

@mrtazz

Slide 25

Slide 25 text

@mrtazz Chef • rules everything around me • Same cookbooks on prod and dev • every node runs Chef every 10 minutes • ton of knife plugins and handlers

Slide 26

Slide 26 text

@mrtazz

Slide 27

Slide 27 text

@mrtazz > 120 recipes

Slide 28

Slide 28 text

@mrtazz

Slide 29

Slide 29 text

@mrtazz Nagios

Slide 30

Slide 30 text

@mrtazz Nagios • 2 instances in each DC/environment • Fully Chef generated configuration • Service checks and contacts in git • Notifications via email->SMS gateway • ~75% ops on-call

Slide 31

Slide 31 text

@mrtazz github.com/lozzd/nagdash

Slide 32

Slide 32 text

@mrtazz

Slide 33

Slide 33 text

@mrtazz

Slide 34

Slide 34 text

@mrtazz

Slide 35

Slide 35 text

@mrtazz Nagios Herald • Add context to nagios alerts • What are the first 5 things you do when you get paged? • You already have the phone in your hand • nagios notification handler

Slide 36

Slide 36 text

@mrtazz

Slide 37

Slide 37 text

@mrtazz The Toys are real

Slide 38

Slide 38 text

@mrtazz There’s another side of heaven

Slide 39

Slide 39 text

@mrtazz Ops Weekly

Slide 40

Slide 40 text

@mrtazz Ops Weekly

Slide 41

Slide 41 text

@mrtazz Summary • Set of trusted tools • Enhance where they come short • Try out new things • Write tools where applicable • Continuous monitoring and adaptation

Slide 42

Slide 42 text

@mrtazz codeascraft.com etsy.com/codeascraft/talks etsy.github.com etsy.com/careers

Slide 43

Slide 43 text

@mrtazz Questions?

Slide 44

Slide 44 text

A Whirlwind Tour of Etsy's Monitoring Stack Daniel Schauenberg [email protected] @mrtazz