@bbaldino
live streaming fail
ffmpeg.0.txt (38.4 KB) jicofo.log (38.5 KB) log.0.txt (39.7 KB)
when restart server it is working
According to the ffmpeg log, it works. It seems that you modified the Jibri stream behaviour and it streams to a custom RTMP server.
yeah @emrah
i followed this instruction Stream to any OR Multiple RTMP Destinations + Record Simultaneously
but some time jibri is working sometime it is not working.
We are doing both streaming and recording to happen at the same time. We setup jibri with autoscaling groups in aws. Now the issue is jicofo knows the available jibri servers and healthcheck is working and it states that the servers as idle. When jibri servers are spinning up they are recognized as available servers. Suppose there are 3 servers. 2 servers are busy with streaming. But when the 3rd tries to stream, it is shown as busy. But actually the 3rd one is free. The healthcheck shows it is free. In the jibri log it shows as idle as well. When the streaming is stopped in the other two conferences from jitsi end and restarted it works. This issues happens from time to time.
What is the Java version? Try with Java8 if it is not the case
And check the RAM usage while streaming
@emrah i am using java8. ram usage is fine. first time jicofo request fail and jibri is busy after that. jitsi meet say streaming fail.
It seems that Jibri starts streaming but crashes after a while. It remains locked/busy after that point.
@emrah actually first time jicofo say jibri start fail but jibri is trying to streaming and being busy. after restart jibri. it is working without any issue. only these issue happen servers are spinup
If both the Java 11 and Java 8 are installed, double check which one is started. I still suspect that RAM or some resource may be insufficient for this container/process.