Graphite

882 views

Published on

Graphite presentation at Triagle Devops Meetup host on 9/19/2013

Published in: Technology, Business
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
882
On SlideShare
0
From Embeds
0
Number of Embeds
3
Actions
Shares
0
Downloads
11
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Graphite

  1. 1. “Collect Everything!” GRAPHITE
  2. 2. Why collect data? - Problem Determination - Root Cause Analysis - Performance Optimization - Cost optimization - Reporting - Forecasting
  3. 3. Why Graphite? - Powerful UI - Highly scalable - No central configuration - Lots of Supporting tools - Collectd - Statsd - Dashboards
  4. 4. Powerful UI - Pulling data - Correlating data - Aggregating data - Manipulating Data
  5. 5. Highly Scalable - Metrics are queued - I/O nice - 40,000+ metrics/minute on 1CPU 3G VM
  6. 6. How to scale? - Big $$$ storage - Mount multiple independent RAID partitions - Multiple Graphite instances - Deduplication
  7. 7. No Central Configuration - Push metrics from anywhere - Clients available for many programming languages - Write data to socket in standard format <Metric.name> <value> <timestamp>
  8. 8. Supporting tools Collectd - Metric collection daemon - Plugin based - Support for… - Apache - JMX - Load - Memory - Disk I/O and usage - Etc.
  9. 9. Supporting tools Statsd - Metrics are submitted to buckets - Buckets flush on a regular interval - Aggregated metrics are written to graphite - Useful for application metrics - Support for… - Counters - Timers - Guages - Sets
  10. 10. Supporting tools Dashboards - Graphite - Descartes - Cubism - Tasseo
  11. 11. Glenn Poston - Twitter: gposton - Github: gposton - Blog: www.glennposton.com

×