The MAD Podcast with Matt Turck Why Influx Rebuilt Its Database for the IoT and Robotics Explosion
May 8, 2025
Evan Kaplan, CEO of InfluxData, shares the remarkable journey of InfluxDB, the leading open-source time series database, which surged from 3,000 to over 1.3 million users. He discusses the major architectural overhaul to InfluxDB 3.0, tackling the cardinality problem and the shift to object storage. Kaplan reveals insights about the competitive landscape against Databricks and Snowflake, and the tough lessons learned in monetizing their vast user base. The evolution of IoT and its critical role in analytics also takes center stage in this engaging conversation.
AI Snips
Chapters
Transcript
Episode notes
Influx Pivoted to Database
- Influx started as a SaaS service for server monitoring but pivoted to focus on its database which became more important.
- The database, built in Go, grew from 3,000 users in 2015 to over 1.3 million daily users today.
Why Specialized Time Series Databases
- Time series data is telemetry captured over time, ideally suited to sensor analytics.
- Specialized databases optimize high-volume time-stamped data ingestion and low-latency queries unlike general purpose databases.
Reasons Behind InfluxDB 3 Rebuild
- The main challenges that led to InfluxDB 3.0 rewrite were the cardinality problem, linked compute and storage, and lack of SQL adoption.
- They rewrote the database in Rust and moved to an object storage model on S3 for scalability and performance.