RTSP SETUP request using TCP handled with error: waiting for response timed out

MIP
, , ,
1

Hello,

I’m currently using Milestone AI Bridge v1.7.2, running via Docker Compose, and I’m encountering an issue when attempting to stream multiple RTSP feeds simultaneously from the AI container.

Specifically, I’m launching 9 parallel RTSP streams from the container. Sometimes, all streams connect successfully, but other times one or more streams fail to establish. In the container logs, I consistently see the following error messages:

RTSP SETUP request using TCP handled with error: waiting for response timed out

RTSP session closed (not in use)

What’s particularly problematic is that once this issue occurs, restarting the AI Bridge container does not resolve it — the affected streams remain inaccessible even after a restart.

To ensure the issue isn’t on the RTSP source side, I tested the exact same setup using MediaMTX RTSP server with 9 simultaneous streams. In that environment, all streams work correctly without timeouts or disconnections.

This leads me to believe the issue may be related to how AI Bridge handles multiple RTSP sessions, possibly due to session cleanup, connection limits, or socket handling.

Could you please advise on:

  • Whether there are any known limitations in AI Bridge v1.7.2 regarding concurrent RTSP streams?
  • Any configuration settings or workarounds to improve stability in this use case?

Thank you in advance for your support.

Best regards,

ABDEL RAHMEN

2

Hello Abdel,

Thanks for reaching out. I would strongly suggest you to upgrade your setup to use AI Bridge 2.0.1 recently published on AWS ECR.

To get the docker compose files for this version you will need to download the Milestone AI Bridge resources file from here.

On version 2.0.1 AI Bridge uses a newer and more stable version of the rtsp packets. You can find the release notes here.

Please, let us know how it goes on this new setup.

Regards,

Fer