Orthanc Performance Issues with OHIF While Loading Large CT/MR Studies

General
1

Hello Orthanc Community,

I hope you are doing well.

I am currently facing a performance issue with Orthanc and would greatly appreciate any guidance or recommendations from the community.

Current Setup

  • AWS EC2 Instance: c6a.xlarge

  • vCPUs: 4

  • RAM: 8 GB

  • Storage: 48 GB gp3 EBS

  • Network Bandwidth: 12.5 Gbps

  • Orthanc: Running inside a Docker container

  • Viewer: OHIF Viewer

Issue Description

When users open CT or MR studies through OHIF, Orthanc becomes very slow and sometimes appears to get stuck or become unresponsive while serving images. This is especially noticeable with larger studies.

At the moment, I have not applied any CPU or memory limits to the Docker container, so Orthanc is free to use the available host resources.

Questions

I would be grateful for your advice on the following:

  1. Is a c6a.xlarge instance (4 vCPUs, 8 GB RAM) sufficient for handling CT and MR studies in a production environment?

  2. What Docker resource allocation or limits would you recommend for Orthanc?

  3. What would be the recommended architecture and infrastructure sizing for an Orthanc + OHIF deployment serving large imaging studies?

I would also appreciate any suggestions on monitoring, performance tuning, or best practices that could help identify the bottleneck.

Thank you very much for your time and assistance. Any advice or shared experiences would be greatly appreciated.

2

Hello @Akash123

We run nearly the same setup as you and have no limits on our Orthanc container.

We used docker stats while viewing studies as an indicator and found that Orthanc is more memory bound than CPU intensive.

When we switched to r6 class we found an immediate improvement in performance. You may find benefit in switching to an r6a.large instance (cheaper too than c6a.xlarge).