Jitsi-meet-docker Server bandwidth

Server bandwidth is 1G.
However, when 4~6 people are connected, the log shows that the bandwidth is insufficient as shown below.

INFO: Endpoints were suspended due to insufficient bandwidth (bwe=196472 bps)

In the log, bwe=196472 bps is not 0.1megabyte?

And Is there a way to check the bandwidth per participant?

Also, are there any settings to limit the bandwidth?

What about participants bandwidth?
These are estimations about participants bandwidth.

Is there any way to check the participant bandwidth?

https://test.webrtc.org/
There is a Video bandwidth at the end with some estimations, run it on the client side.

The result is shown below.
What does it have to do with bandwidth?
I want to measure the bandwidth.

You need to click on the last green bar and there click on video bandwidth option to see the results.

The result is shown below.

Which items should I look at?

For the participant where you ran these tests you constantly see in jvb logs: bwe=196472 bps ?

Is the bps in the jvb log bytes per second as shown below? Or is it bits per second?
INFO: Endpoints were suspended due to insufficient bandwidth (bwe=106848 bps)

I see that it is bits per second

And here is bwe=106848 bps exceeded? What does the bps appear over there mean?

That there is no sufficient bandwidth for that participants so a videos are not sent to that participant.

What I’m curious about is bwe = 106848 bps I want to know a clear and clear meaning.

Does 106848bps represent the participant’s bandwidth?
Or maybe 106848bps isn’t enough?

Info: The endpoint was paused due to insufficient bandwidth (bwe = 106848bps).

Yes this is the bandwidth that was measured on the participant’s side.

The inactive black screen does not appear on the jitsi site (meet.jit.si).

It only appears on the selfhosting server.

In this case, isn’t it a problem with the server?

Seems like … is the server and the participant close enough, maybe that participant doesn’t have good connection to that server.

I experience the same issue, context is self-hosting on docker with coturn as TURN server.
This seems not to happen on meet.jit.si but when self hosting. Even a small number of participants does not work out.

If using a different setup with multiple devices works great (no p2p in any case) I wonder if TURN could be somehow involved.

Another hypothesis: As I am experiencing the same issue as Firefox and Safari: endless loop with 'toUnifiedPlan'? this could be related to not using websockets for bridge channel.

Are there any settings related to server bandwidth?

@damencho @qwerts89 I made a solution using Jitsi APIs & I only have 3 participants in the conference and those are kept getting connected and disconnected on every seconds, when I see the logs it kept getting filled with messages like

BandwidthAllocator.allocate#361: Endpoints were suspended due to insufficient bandwidth (bwe=80576 bps): 3a49461c

is there any solution of this problem ? something we can do on the server side ?

This is not enough bandwidth on the client side detected … Nothing to be done on serverside