Lost Connection - Reload - Lost Connection - Reload ... Repeat

We have a lot of users who report issues with Jitsi Meet reporting connection issues but they have no problem using other video services.

When they reload, Jitsi works again for approx 15 seconds or less sometimes and then they need to reload again. These are one on one sessions, usually, the other person is still there even though video and audio stops working. After reload it works again but only for a short time…

It’s very counter productive for sessions and we still don’t have any solutions. Any update on this issue? I know it’s been reported before. Is there a bug with the reconnecting after a connection issue detection?

Thanks

I have the exact same issue. I regularly take meetings on Skype for business, MS Teams, Webex Teams.

I have 1Gbps at home because I work for the ISP. All my switches are Gb switches.

I setrup jitsi on a physical server at home and started having meetings with 4-5 friends.

It works for a few minutes then connections start dropping randomly. Eventually you’re left with just your own portrait. So you reload and everyone comes back. Same goes for all the other participants, if they reload they can see everyone for a short while before it all stops working again.

What is this? It’s clearly something wrong with my setup or with jitsi.

We’re using another setup that someone wrote themselves with JS and that works better than jitsi.

If your jitsi-meet setup work for two users but looses connection with 3 or more users then you likely have a misconfigured videobridge.

When using 2 users jitsi uses peer-2-peer mode, when the third user connects jitsi switches to using the videobridge.

check the /var/log/jitsi/jvb.log and the /var/log/jitsi/jicofo.log and make sure the videobridge connect and work.
If your videobridge server is behind a NAT then make sure you have updated
/etc/jitsi/videobridge/sip-communicator.properties
https://github.com/jitsi/jitsi-meet/blob/master/doc/quick-install.md#advanced-configuration

to ease debugging of the videobridge issues you can disable the p2p mode entirely so that all conferences uses the videobridge even with two users.
p2p: {
enabled: false,
}

My JVB is behind a NAT in regards to the internet, but jicofo and the other jitsi services are on the same server as jvb so from their perspective it’s not behind a NAT.

Nevertheless I did set the sip-communicator.properties settings as recommended by the setup docs for when you’re using a NAT. I assumed it meant from the perspective of the users.

My issue is intermittent loss of connection. Sometimes I can see all 5 participants but only for a short while.

Does that still mean jvb fails to connect? Because the logs are quite extensive.

Here’s 5 minutes of the meeting, I couldn’t upload anymore than 5 minutes because the log is so verbose no pastebin service will take it.

https://paste.centos.org/view/fe3895e7

I’ve tried reading the log myself and without any experience with jvb I can at least see that it’s establishing connections to peers at certain points but then perhaps losing them or losing packets to them. Not sure what’s happening.

Hey, did someone solved this problem? Can anyone help me with it?

@Stefan_Midjich @Askar_Musaev Did any of you able to resolve this issue ? I am facing this similar issue & I think I would need to alter some sort of videobridge configuration to make it work. Any help in this regard would be very appreciated.

I don’t know why it worked, but i just created new conference(other name) and it worked well(couldn’t debug old one), but it happens for me for sometimes.

Hi dear friend
Sometimes this problem is due to the lack of prosody updates.