Files
landing_page/qdrant-landing/content/documentation/ops-monitoring/memory-usage.md
T
Abdon PijpelinkandClaude Sonnet 4.6 dc7d66545e Add Monitor Collection Memory Usage page (v1.18.0) (#2300)
* Add Monitor Collection Memory Usage page (v1.18.0)

New page in Monitoring & Telemetry documenting the Web UI Memory tab and
the GET /collections/{name}/memory API endpoint added in qdrant/qdrant#8606
and qdrant/qdrant-web-ui#372.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Update web-ui docs link

* Review feedback

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-04-30 20:24:27 +02:00

2.4 KiB
Raw Blame History

title, weight
title weight
Memory Usage 7

Monitor Collection Memory Usage

Available as of v1.18.0

Qdrant lets you inspect a collection's disk space, RAM, and OS page cache usage, summed up across the whole cluster and broken down by component. Use this to plan capacity, diagnose memory pressure, or understand which parts of a collection are resident in memory.

This information is available in the Qdrant Web UI and through the API.

Web UI

Open the collection detail page and select the Memory tab. It shows the memory breakdown for the collection, updated on demand.

The Memory tab shows the breakdown of disk, RAM, and cached usage for each component of the collection.

Understanding the Fields

The breakdown covers these components:

Component Description
Total Aggregate across all components.
Vectors Per dense and multi-dense vector: storage, index, and optionally quantization.
Sparse Vectors Per sparse vector: storage and index.
Payload Payload storage.
Payload Index Per payload field index.
ID Tracker Maps external point IDs to internal ones.

Each component reports four values:

Field Description
Disk Total file sizes on disk.
RAM Non-evictable heap RAM: in-memory data structures not backed by memory-mapped files.
Cached Evictable RAM: file pages currently resident in the OS page cache.
Expected Cache The amount of data that should ideally be cached for best performance. Compare this against Cached to see how much of the working set is warm.

API

You can retrieve the same data though Qdrant's API:

curl http://localhost:6333/collections/{collection_name}/memory

Accuracy

The reported values are estimates. RAM usage is typically underestimated by 10–15% because memory allocated by third-party libraries and the allocator itself isn't accounted for.

On non-Unix systems, Cached is always reported as 0.