Try or ByeTry or Bye

Qdrant Out of Memory (OOMKilled / High RAM Usage)

Share
Workaround available Self-hosted, Docker, Kubernetes Updated November 14, 2026Based on 4 user reportsHow we research

Quick answer

By default Qdrant keeps vectors and the index in RAM, so large collections hit container limits. Move vectors (and the index) to disk with mmap, use quantization, and give the container enough memory.

Most likely fix: Confirm it is an OOM kill

How to fix it

Try these in order — the most common fix is first.

  1. 1. Confirm it is an OOM kill

    1. docker inspect <container> shows OOMKilled: true, exit code 137
    Did this fix it?

  2. 2. Store vectors on disk

    1. Create the collection with vectors on_disk: true so vectors use mmap
    2. Qdrant's own benchmark shows this lets the same data run in less RAM, but throughput drops sharply at very low RAM
    Did this fix it?

  3. 3. Move the index and payload to disk or quantize

    1. Set hnsw_config on_disk: true for the index
    2. Use on_disk_payload for large payloads
    3. Enable scalar or binary quantization to keep compressed vectors in RAM
    4. Check the current optimize guide for exact field names for your version
    Did this fix it?

  4. 4. Raise limits

    1. Increase the container memory limit
    2. Upload in smaller batches to reduce optimizer spikes
    Did this fix it?

Symptoms

Container exits with code 137 / OOMKilled, RAM usage keeps growing, or the node restarts under load with a big collection.

Why it happens

  • Vectors and HNSW index held fully in RAM
  • Docker or Kubernetes memory limit below the working set
  • Large payload kept in memory
  • Indexing and optimizer jobs spiking memory during bulk upload

This guide also covers

  • qdrant out of memory
  • qdrant high memory usage
  • qdrant oomkilled docker
  • qdrant ram usage large collection

Still not working?

If Qdrant keeps failing, you're not alone. See every Qdrant problem users report or the alternatives people switch to.

Have an app idea? Our small team builds any software: web and mobile apps, games, desktop, AI tools. Price agreed together.

Talk to us
Fixed it? Share this guide

The same problem in other apps

All slow, laggy or resource-hungry guides →

Other Qdrant errors