Realtime System Uptime Tracking for 2M req/min
Submitted by Amiruddin Nagri (@amir) on Monday, 26 February 2018
@GO-JEK, we serve more than 3M+ customers daily. This involves rigorous discipline around system reliability and availability. One of the tools that helps us achieve this goal is our Realtime System Uptime Tracking.
- Vanilla ELK stack and its shortcomings
- Moving logs to Kafka
- Realtime aggregation using Flink
- Scaling InfluxDB
- Scaling Elastic Search
- Monitoring and Alerting using Grafana
- Advantages over ELK
Amir works as Data Engineer@GO-JEK. He has donned many hats, and currently loves working on distributed systems and problems involving data and scaling.