Configuration

These pages cover the settings and runtime options available for customizing and administering a Qdrant deployment.

Configuration

Configuration describes how to customize Qdrant’s behavior using config files and environment variables, covering storage paths, network interfaces, performance parameters, and feature flags.

Resource Quotas

Resource Quotas explains how to cap node memory and disk usage across a cluster, what happens when a node reaches a limit, and how to find the node that is full.

Administration

Administration covers runtime administration tools that let you modify instance behavior without restarting, including recovery mode for resolving out-of-memory situations, low memory mode, and strict mode.

Memory Tiers

Memory Tiers explains how the memory parameter (cold, cached, pinned) controls RAM residency for vectors, indexes, quantized data, and payloads.

Usage Statistics

Usage Statistics explains what anonymized telemetry the open-source Qdrant image collects, why it’s collected, and how to opt out.

Running with GPU

Running with GPU explains how to enable GPU-accelerated vector indexing using dedicated Docker images for NVIDIA and AMD hardware, available from Qdrant v1.13.0.