Ingestion Rate
Ingestion rate is the speed at which a database can accept and store incoming data, usually measured in records or rows per second. For high-volume workloads like telemetry, metrics, and market data, it is one of the most important performance numbers.
Why ingestion rate matters
When data arrives in a constant stream, the database has to keep up or data backs up and gets dropped. A fleet of sensors, a busy application emitting events, or a market data feed can generate millions of records per second. If the database cannot ingest that fast, you are forced to sample, aggregate early, or drop data.
Ingestion rate depends on how the system handles writes. Designs that funnel every write through a single process or rebuild an index on every insert hit a ceiling quickly. Designs that buffer and write in column-oriented batches scale much higher.
High ingestion rate is what lets you keep full-resolution data instead of throwing detail away to survive the volume.
How Arc handles Ingestion Rate
Arc sustains 19.9 million records per second on a single instance because writes are buffered and written as columnar Parquet rather than funneled through one stateful process. That headroom is what lets teams stop sampling and keep every record.
Arc is a high-performance columnar database. Open Parquet on storage you own, single Go binary, production-ready in 30 seconds.