Streaming into Iceberg at Scale: What Breaks and How We Fixed It

​What happens when Iceberg has to keep up with millions of events per second?

​At scale, streaming into Iceberg hits a less obvious bottleneck: writers multiply, commits collide, metadata piles up, and freshness starts to slip.

​In this 60-minute live webinar, join Aravindh Sridharan from e6data for a technical deep dive into how we rethought the Iceberg write path, separating data production from publishing and redesigning how commits and metadata are handled at scale.

​We’ll unpack the architecture that reached 1.5 million events/sec into a single Iceberg table, with data queryable in ~30 seconds, and explore why commit coordination, not raw write throughput, becomes the real challenge at scale.

​Expect architecture, engineering trade-offs, and 20 minutes of live Q&A.

------------------------------------------------------------

Interested in joining?

​Register on this page to receive the webinar joining details.

​Who should attend?

​Data, Platform & Backend Engineers, Architects, and Engineering Leaders working with streaming data, Iceberg, or modern data infrastructure.