No live upstreams while connecting to upstream message in nginx log

Install & Config
#1

Hi all. Can anybody help me clarify source of this errors in the nginx error log? This error messages repeats more times in bulks during day (maybe related to higher server load at this time), this is only shot example:

2020/04/07 09:26:50 [error] 27228#27228: *378977 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 127.0.0.1, server: meet.domain.eu, request: “POST /http-bind?room=strejcekfrantisek HTTP/2.0”, upstream: “http://127.0.0.1:5280/http-bind?room=strejcekfrantisek”, host: “meet.domain.eu”, referrer: “https://meet.domain.eu/StrejcekFrantisek”
2020/04/07 09:26:53 [error] 27227#27227: *453493 connect() failed (111: Connection refused) while connecting to upstream, client: 85.248.11.76, server: 0.0.0.0:443, upstream: “127.0.0.1:4445”, bytes from/to client:0/0, bytes from/to upstream:0/0
2020/04/07 09:27:05 [error] 27228#27228: *438696 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 127.0.0.1, server: meet.domain.eu, request: “POST /http-bind?room=fulier HTTP/2.0”, upstream: “http://[::1]:5280/http-bind?room=fulier”, host: “meet.domain.eu”, referrer: “https://meet.domain.eu/Fulier”
2020/04/07 09:27:05 [error] 27228#27228: *419209 no live upstreams while connecting to upstream, client: 127.0.0.1, server: meet.domain.eu, request: “POST /http-bind?room=munk HTTP/2.0”, upstream: “http://localhost/http-bind?room=munk”, host: “meet.domain.eu”, referrer: “https://meet.domain.eu/munk”
2020/04/07 09:27:05 [error] 27228#27228: *445335 no live upstreams while connecting to upstream, client: 127.0.0.1, server: meet.domain.eu, request: “POST /http-bind?room=schlarmannovazoologia1 HTTP/2.0”, upstream: “http://localhost/http-bind?room=schlarmannovazoologia1”, host: “meet.domain.eu”, referrer: “https://meet.domain.eu/Schlarmannovazoologia1”
2020/04/07 09:27:05 [error] 27228#27228: *443382 no live upstreams while connecting to upstream, client: 127.0.0.1, server: meet.domain.eu, request: “POST /http-bind?room=schlarmannovazoologia1 HTTP/2.0”, upstream: “http://localhost/http-bind?room=schlarmannovazoologia1”, host: “meet.domain.eu”

For today we have so far:
meet:~/scripts# cat /var/log/nginx/error.log|grep “no live upstreams while connecting to upstream, client: 127.0.0.1”|wc -l
1301

It seem this errors are aligned with messages in prosody.log:

Apr 07 09:27:05 mod_bosh info Client tried to use sid ‘975b5495-51b1-4998-9f0f-350ea3f25953’ which we don’t know about
Apr 07 09:27:05 mod_bosh info Client tried to use sid ‘975b5495-51b1-4998-9f0f-350ea3f25953’ which we don’t know about
Apr 07 09:27:05 mod_bosh info Client tried to use sid ‘975b5495-51b1-4998-9f0f-350ea3f25953’ which we don’t know about
Apr 07 09:27:05 mod_bosh info Client tried to use sid ‘975b5495-51b1-4998-9f0f-350ea3f25953’ which we don’t know about
Apr 07 09:27:05 mod_bosh info Client tried to use sid ‘975b5495-51b1-4998-9f0f-350ea3f25953’ which we don’t know about

We are using secure domain with LDAP2 auth setup on Debian 10. Any hints on it please?

Thank you in advance!

Kind regards,

Milan

Server load for many simultaneous meetings
#2

same problem here, using jwt with MUC approach

2020/04/08 07:19:58 [error] 965#965: *132381 recv() failed (104: Connection reset by peer) while proxying and reading from upstream, client: some_ip, server: 0.0.0.0:443, upstream: “127.0.0.1:4445”, bytes from/to client:582/1026, bytes from/to upstream:1026/1134
2020/04/08 07:20:09 [error] 965#965: *131915 recv() failed (104: Connection reset by peer) while proxying and reading from upstream, client: some_ip, server: 0.0.0.0:443, upstream: “127.0.0.1:4445”, bytes from/to client:1674/872, bytes from/to upstream:872/2226
2020/04/08 07:20:25 [error] 965#965: *135674 recv() failed (104: Connection reset by peer) while proxying and reading from upstream, client: some_ip, server: 0.0.0.0:443, upstream: “127.0.0.1:4445”, bytes from/to client:422/3963, bytes from/to upstream:3963/718
2020/04/08 07:20:48 [error] 965#965: *140916 recv() failed (104: Connection reset by peer) while proxying and reading from upstream, client: some_ip, server: 0.0.0.0:443, upstream: “127.0.0.1:4445”, bytes from/to client:7/2951, bytes from/to upstream:2951/178
2020/04/08 07:22:32 [error] 965#965: *151378 recv() failed (104: Connection reset by peer) while proxying and reading from upstream, client: some_ip, server: 0.0.0.0:443, upstream: “127.0.0.1:4445”, bytes from/to client:7/2951, bytes from/to upstream:2951/178
2020/04/08 07:23:24 [error] 965#965: *156217 recv() failed (104: Connection reset

Apr 08 07:25:08 mod_bosh info Client tried to use sid ‘e8645024-355c-412d-a284-60c23086f90f’ which we don’t know about
Apr 08 07:25:09 mod_bosh info Client tried to use sid ‘e8645024-355c-412d-a284-60c23086f90f’ which we don’t know about
Apr 08 07:25:10 mod_bosh info Client tried to use sid ‘e8645024-355c-412d-a284-60c23086f90f’ which we don’t know about
Apr 08 07:25:10 mod_bosh info Client tried to use sid ‘e8645024-355c-412d-a284-60c23086f90f’ which we don’t know about
Apr 08 07:25:10 mod_bosh info Client tried to use sid ‘e8645024-355c-412d-a284-60c23086f90f’ which we don’t know about
Apr 08 07:25:11 mod_bosh info Client tried to use sid ‘e8645024-355c-412d-a284-60c23086f90f’ which we don’t know about
Apr 08 07:25:11 mod_bosh info Client tried to use sid ‘e8645024-355c-412d-a284-60c23086f90f’ which we don’t know about

#3

There are several topics in the community about this;

but not a specific solution for all i think to get rid of
2020/04/17 11:25:20 [error] 949#949: *1124533 recv() failed (104: Connection reset by peer) while proxying and reading from upstream, client: 81.215.xxx.xxx, server: 0.0.0.0:443, upstream: “127.0.0.1:4445”, bytes from/to client:7/2951, bytes from/to upstream:2951/178

kind of nginx error logs…

changing to
/etc/nginx/modules-enabled/60-jitsi-meet.conf
map $ssl_preread_alpn_protocols $upstream {
~\bh2\b web;
~\bhttp/1. web;
default turn;
}
also did not help…

I have installed jitsi-meet unstable from scratch in two methods, one with nginx installation first and second just jitsi-meet but both installations also have same error.log especially when i connect from an iOS device

any other suggestions??