v.24.8Experimental Feature

Use Adaptive Read Task Size Calculation for Parallel Replicas

Official CH changelog text
Use adaptive read task size calculation method (adaptive meaning it depends on read column sizes) for parallel replicas. #60377 (Nikita Taranov).
Adaptive read task size calculation for parallel replicas in ClickHouse, where the task size depends on the size of the read columns.

Why it matters

This feature aims to optimize the performance of parallel replicas by dynamically adjusting the read task size based on the size of the columns being read. It helps in better load balancing and efficient resource utilization during data replication and querying.

How to use it

The adaptive read task size calculation is applied automatically for parallel replicas. No explicit user action is required to enable this feature; it activates by detecting column size dependencies during read operations.

Related resources

PreviousNext