Hello everyone,
I am reaching out regarding an issue I am encountering while deploying version 3.5.0 using AWS HA.
Specifically, the OPENVIDU_URL variable is associated with a unique IP address, but it is not returning any data. This behavior is causing the error shown in the attached file.
Could you please help me investigate why this is happening?
Best regards, Matteo.
You are using the Angular component, right?
I think this issue is related to this: OpenVidu 3.5: Components (Angular) can’t join in split-VM setup
Take a look to the workaround of the first message.
Hi,
yes, I’m using the “Openvidu platform” angular components solution.
I’ll take a look at the link you suggested, however are there any problem on my deployment?
Is it correct that if I launch my OPENVIDU_URL value in the browser I got this?
Shouldn’t I be seeing the Meet App or something else?
Thank you, Matteo.
No, It should open OpenVidu Meet.
I would check if every docker container is running correctly on every Master Node.
SSH into every Master Node and execute:
sudo su
docker ps
See if there is any restarting service.
In that case, restart with:
systemctl restart openvidu
EDIT:
The ERR_TIMED_OUT also suggest me that maybe the DNS is not configured to point to the LoadBalancer? Verify it.
Hi,
it was the DNS that hasn’t kicked in yet and now I see the Meet App.
I tried to modify my caddy.yaml configuration inside the Master Node 1, however I don’t know where to change the configuration, could you help me please?
# Caddy configuration file
# ----------------------
# NOTES:
# ----------------------
# WARNING: Take into account that this file will be overwritten by OpenVidu when you upgrade it.
#
# This file uses the same interpolation rules as Docker Compose.
# For more details, refer to the documentation: https://docs.docker.com/compose/compose-file/12-interpolation/
#
# - ${openvidu.CONFIG_PARAMETER} must be used to access environment variables from openvidu.env.
# - ${master_node.CONFIG_PARAMETER} must be used to access environment variables from master_node.env.
# ----------------------
admin:
listen: unix//var/run/caddy/admin.sock
storage:
module: redis
client_type: failover
address:
- master-node-1:${master_node.REDIS_SENTINEL_INTERNAL_PORT:?mandatory}
- master-node-2:${master_node.REDIS_SENTINEL_INTERNAL_PORT:?mandatory}
- master-node-3:${master_node.REDIS_SENTINEL_INTERNAL_PORT:?mandatory}
- master-node-4:${master_node.REDIS_SENTINEL_INTERNAL_PORT:?mandatory}
master_name: openvidu
password: ${master_node.REDIS_PASSWORD:?mandatory}
password_sentinel: ${master_node.REDIS_PASSWORD:?mandatory}
apps:
layer4:
servers:
rtmp:
listen:
- ":${openvidu.CADDY_RTMPS_PUBLIC_PORT:?mandatory}"
routes:
- handle:
- handler: proxy
# @id is used to dinamically update upstreams
# by the operator
"@id": ov-rtmp
upstreams:
- dial: []
turn:
listen:
- ":${openvidu.LIVEKIT_TURN_TLS_INTERNAL_PORT:?mandatory}"
routes:
- handle:
- handler: proxy
# @id is used to dinamically update upstreams
# by the operator
"@id": ov-turn
upstreams:
- dial: []
http:
http_port: ${openvidu.CADDY_HTTP_PUBLIC_PORT:?mandatory}
https_port: ${openvidu.CADDY_HTTPS_PUBLIC_PORT:?mandatory}
servers:
# Redirect HTTP to HTTPS
redirect:
listen:
- ":${openvidu.CADDY_HTTP_PUBLIC_PORT:?mandatory}"
routes:
- handle:
- handler: static_response
status_code: 301
headers:
Location:
- https://{http.request.host}:${openvidu.CADDY_HTTPS_PUBLIC_PORT:?mandatory}{http.request.uri}
public:
listen:
- ":${openvidu.CADDY_HTTP_INTERNAL_PORT:?mandatory}"
logs:
default_logger_name: default
automatic_https:
disable: true
routes:
- handle:
- handler: subroute
# @id is used to dinamically update openvidu route
# by the operator
"@id": "ov-public-handler"
routes: []
# Ingress WHIP service
- handle:
- handler: subroute
routes:
- handle:
- handler: reverse_proxy
# @id is used to dinamically update upstreams
# by the operator
"@id": ov-whip
health_checks:
active:
expect_status: 404
interval: 5s
timeout: 2s
uri: /
load_balancing:
retries: 2
selection_policy:
policy: round_robin
upstreams: []
match:
- path:
- /whip
# Dashboard
- handle:
- handler: subroute
routes:
- handle:
- handler: rewrite
strip_path_prefix: /dashboard
- handler: reverse_proxy
health_checks:
active:
expect_status: 200
interval: 5s
timeout: 2s
uri: /dashboard/api/healthcheck
load_balancing:
retries: 2
selection_policy:
policy: round_robin
upstreams:
- dial: master-node-1:${openvidu.DASHBOARD_INTERNAL_PORT:?mandatory}
- dial: master-node-2:${openvidu.DASHBOARD_INTERNAL_PORT:?mandatory}
- dial: master-node-3:${openvidu.DASHBOARD_INTERNAL_PORT:?mandatory}
- dial: master-node-4:${openvidu.DASHBOARD_INTERNAL_PORT:?mandatory}
match:
- path:
- /dashboard/*
- /dashboard
# Minio Console
- handle:
- handler: subroute
routes:
- handle:
- handler: static_response
headers:
Location:
- "/minio-console/"
status_code: 302
match:
- path:
- /minio-console
- handle:
- handler: subroute
routes:
- handle:
- handler: rewrite
strip_path_prefix: /minio-console
- handler: reverse_proxy
health_checks:
active:
expect_status: 200
interval: 5s
timeout: 2s
uri: /minio-console/
load_balancing:
retries: 2
selection_policy:
policy: round_robin
upstreams:
- dial: master-node-1:${openvidu.MINIO_CONSOLE_INTERNAL_PORT:?mandatory}
- dial: master-node-2:${openvidu.MINIO_CONSOLE_INTERNAL_PORT:?mandatory}
- dial: master-node-3:${openvidu.MINIO_CONSOLE_INTERNAL_PORT:?mandatory}
- dial: master-node-4:${openvidu.MINIO_CONSOLE_INTERNAL_PORT:?mandatory}
match:
- path:
- /minio-console/*
# OpenVidu v2 compatibility API
- handle:
- handler: subroute
routes:
- handle:
- handler: reverse_proxy
health_checks:
active:
expect_status: 401
interval: 5s
timeout: 2s
uri: /openvidu/api/health
load_balancing:
retries: 2
selection_policy:
policy: ip_hash
upstreams:
- dial: master-node-1:${openvidu.OPENVIDU_V2COMPAT_INTERNAL_PORT:?mandatory}
- dial: master-node-2:${openvidu.OPENVIDU_V2COMPAT_INTERNAL_PORT:?mandatory}
- dial: master-node-3:${openvidu.OPENVIDU_V2COMPAT_INTERNAL_PORT:?mandatory}
- dial: master-node-4:${openvidu.OPENVIDU_V2COMPAT_INTERNAL_PORT:?mandatory}
match:
- path:
- /openvidu/api/*
- /openvidu/recordings/*
- /openvidu/ws/*
# Grafana
- handle:
- handler: subroute
routes:
- handle:
- handler: reverse_proxy
health_checks:
active:
expect_status: 200
interval: 5s
timeout: 2s
uri: /api/health
load_balancing:
retries: 2
selection_policy:
policy: ip_hash
upstreams:
- dial: master-node-1:${openvidu.GRAFANA_INTERNAL_PORT:?mandatory}
- dial: master-node-2:${openvidu.GRAFANA_INTERNAL_PORT:?mandatory}
- dial: master-node-3:${openvidu.GRAFANA_INTERNAL_PORT:?mandatory}
- dial: master-node-4:${openvidu.GRAFANA_INTERNAL_PORT:?mandatory}
match:
- path:
- /grafana/*
- /grafana
# Health check Caddy
- handle:
- handler: subroute
routes:
- handle:
- handler: static_response
status_code: 200
body: "OK"
match:
- path:
- /health/caddy
# OpenVidu Meet
- handle:
- handler: subroute
routes:
- handle:
- handler: reverse_proxy
health_checks:
active:
expect_status: 200
interval: 5s
timeout: 2s
uri: /
load_balancing:
retries: 2
selection_policy:
policy: ip_hash
upstreams:
- dial: master-node-1:${openvidu.MEET_INTERNAL_PORT:?mandatory}
- dial: master-node-2:${openvidu.MEET_INTERNAL_PORT:?mandatory}
- dial: master-node-3:${openvidu.MEET_INTERNAL_PORT:?mandatory}
- dial: master-node-4:${openvidu.MEET_INTERNAL_PORT:?mandatory}
match:
- path:
- /*
minio:
listen:
- ":${openvidu.CADDY_MINIO_PUBLIC_PORT:?mandatory}"
logs:
default_logger_name: default
automatic_https:
disable: true
routes:
- handle:
- handler: reverse_proxy
health_checks:
active:
expect_status: 200
interval: 5s
timeout: 2s
uri: /minio/health/live
load_balancing:
retries: 2
selection_policy:
policy: round_robin
upstreams:
- dial: master-node-1:${openvidu.MINIO_API_INTERNAL_PORT:?mandatory}
- dial: master-node-2:${openvidu.MINIO_API_INTERNAL_PORT:?mandatory}
- dial: master-node-3:${openvidu.MINIO_API_INTERNAL_PORT:?mandatory}
- dial: master-node-4:${openvidu.MINIO_API_INTERNAL_PORT:?mandatory}
logging:
logs:
default:
level: WARN
# Comment the previous logging section and
# uncomment the following lines to enable detailed Caddy logs
# Disabled by default to avoid performance issues
# ----------------------
# logging:
# logs:
# default:
# level: INFO
# encoder:
# format: filter
# wrap:
# format: json
# fields:
# "request>headers":
# filter: delete
# "resp_headers":
# filter: delete
# "request>uri":
# filter: query
# actions:
# - parameter: "access_token"
# type: delete
# include:
# - http
# writer:
# output: stdout
# layer4access:
# level: DEBUG
# include:
# - layer4
# writer:
# output: stdout
Thank you, Matteo.
Ah, sorry, that /rtc parameter is configured via the operator in PRO versions
It is not a mandatory change, is it working now with the DNS?
If not I will build an operator with that specific change
Hi, no, unfortunately it still not working.
I get the same error shared in the image in the previous messages.
Thank you, Matteo.
PS: It works using the Meet App, but if I want to use my customized Openvidu Platform that uses Angular Components I still got the error.
I’ll prepare an updated image with the operator fix and let you know here once it’s ready with instructions on how to deploy it.
I’ve created an image docker.io/openvidu/openvidu-operator:3.5.1 with the caddy rules changed.
You need to:
/opt/openvidu/docker-compose.yamloperator service and change the image to: docker.io/openvidu/openvidu-operator:3.5.1: operator:
image: docker.io/openvidu/openvidu-operator:3.5.1
sudo su
cd /opt/openvidu
docker compose pull
systemctl restart openvidu
You need to change the docker-compose.yaml and restart every Master Node.
Hi, now it works!
Thank you again for the fix, have a nice day.