Optimization

These pages cover strategies and internal mechanisms for improving search performance, reducing memory usage, managing how Qdrant organizes data on disk, and troubleshooting read-write contention.

Optimize Performance

Optimize Performance walks through three main optimization scenarios — high-speed search, high-precision search, and low memory usage — and the configuration choices that achieve each one.

Optimizer

Optimizer explains how Qdrant’s background optimizer rebuilds segments to keep data structures efficient, including the vacuum, merge, and indexing stages and the thresholds that trigger them.

Troubleshoot Read-Write Contention

Troubleshoot Read-Write Contention walks through a step-by-step set of configuration changes to reduce contention between the background optimizer and search queries under continuous write load.