Databricks makes Real-Time Mode generally available for Spark Declarative Pipelines
Overview
Databricks announces general availability of Real-Time Mode for Spark Declarative Pipelines, which it says brings millisecond end-to-end latency to SDP pipelines.
The company says RTM is up to 92% faster than Apache Flink on customer feature-computation workloads. SDP can now also write to custom sinks such as Lakebase and Redis.
According to the Databricks blog, Real-Time Mode requires an update flow.
Written by AI from the articles below · updated Oct 29, 7:27 PM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
- Databricks BlogOfficialPickDatabricks makes real-time mode generally available for Spark Declarative Pipelines
AIDatabricks announces general availability of Real-Time Mode for Spark Declarative Pipelines, bringing millisecond end-to-end latency to SDP pipelines. The company says RTM is up to 92% faster than Apache Flink on customer feature-computation workloads, and that SDP can now write to custom sinks such as Lakebase and Redis.
Heat Trend
Not enough continuous observations to show a trend yet.