Upstream connection issue

[spkcompute]

Hi Team,

I am getting following error in default setup

websocket-tcp-relay[31604]: 3.7.237.29:56904 connected
3.7.237.29:56904 connected to upstream
3.7.237.29:56904 disconnected: #<IO::Error:Error reading socket: Connection reset by peer>

[main]
upstream = tcp://127.0.0.1:5672
bind = 0.0.0.0
port = 443
proxy-protocol = false
webroot = /var/www/html
tls-cert = /etc/letsencrypt/live/polkadex.trade/fullchain.pem
tls-key = /etc/letsencrypt/live/polkadex.trade/privkey.pem

[spkcompute]

Error from rabbitmq log

2022-02-18 17:36:46.329907+00:00 [info] <0.31586.0> accepting AMQP connection <0.31586.0> (127.0.0.1:43634 -> 127.0.0.1:5672)
2022-02-18 17:36:46.330080+00:00 [erro] <0.31586.0> closing AMQP connection <0.31586.0> (127.0.0.1:43634 -> 127.0.0.1:5672):
2022-02-18 17:36:46.330080+00:00 [erro] <0.31586.0> {handshake_timeout,handshake}

[carlhoerberg]

That's 0.2ms between the connection and the timeout. Have you configred the handshake timeout manually? Because the default is more like 10 seconds.

[spkcompute]

No I am using default setup
handshake_timeout = 10000
Should I change it?

Also when I test message submission directly to rabbitmq it is working fine.

[carlhoerberg]

Which client library/version were you connecting with?

[shubhenducw]

@carlhoerberg

I am facing the same issue

2024-03-12T17:14:50.914146378Z 152.58.208.213:0 disconnected: #<IO::Error:Closed stream>
2024-03-12T17:14:52.571178383Z 152.58.208.213:0 connected
2024-03-12T17:14:52.578574932Z 152.58.208.213:0 connected to upstream

and with this we noticed there's been increase in the memory usage with time.

Created an issue for this : #16