6. inGraph
● inGraphs is a visualization frontend for a linkedin’s metrics
● Agent is a library used by all java application
● Data is transported over kafka
● Collector consume and write to rrd
● Features
○ Rest based
○ Self service
○ Little over 500,000 metrics collected per minute
○ Each RRD is roughly 815k. Each RRD is written to two collectors
to maintain data integrity
○ We currently have 1.4TB of SSD in production colo
7. ELK at linkedin
● ELK (Elasticsearch, Logstash,
Kibana)
○ E -> Near Real-Time Search &
Analysis
○ L -> Plumbing and Glue for all
your data
○ K -> Line graphs, pie charts,
dashboard
● Logstash kafka plugin consumes
logs and end up written into
elasticsearch
8. Why Kafka
1. Near real time delivery
2. Supports multiple consumers groups aka queuing
3. No overhead on clients that comes up logstash agent
4. Kafka scales horizontally
5. Supports REST out of the box
6. In house support