NextCloud Integration error: 502 Bad Gateway for healthcheck

[farewarr]

Nextcloud integration with OnlyOffice fails due to the following nextcloud errror:

###Error from Nextcloud:

Error	onlyoffice	ServerExceptionServer error: GET https://onlyoffice.nonooculusnas.com/healthcheck resulted in a 502 Bad Gateway response: <title>502 Bad Gateway</title> 502 Bad Gateway ngin (truncated...)healthcheckRequest on check error

• Error is seen when clicking save on the integration settings and in the nextcloud logs page.

###NGINX logs from onlyoffice:
nginx logs again: 2024/11/13 05:52:13 [error] 755#755: *2 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "GET /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:52:13 [error] 755#755: *2 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "GET /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:52:25 [error] 755#755: *5 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "GET /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:52:25 [error] 755#755: *5 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "GET /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:52:52 [error] 755#755: *8 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "GET /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:52:52 [error] 755#755: *8 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "GET /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:53:42 [error] 755#755: *13 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "POST /ConvertService.ashx HTTP/1.1", upstream: "http://127.0.0.1:8000/ConvertService.ashx", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:53:42 [error] 755#755: *13 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "POST /ConvertService.ashx HTTP/1.1", upstream: "http://127.0.0.1:8000/ConvertService.ashx", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:53:45 [error] 755#755: *16 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "POST /ConvertService.ashx HTTP/1.1", upstream: "http://127.0.0.1:8000/ConvertService.ashx", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:53:45 [error] 755#755: *16 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "POST /ConvertService.ashx HTTP/1.1", upstream: "http://127.0.0.1:8000/ConvertService.ashx", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:54:23 [error] 755#755: *21 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "HEAD /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:54:23 [error] 755#755: *21 connect() failed (111: Unknown error) while connecting to upstream, client: 172.25.0.1, server: , request: "HEAD /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "onlyoffice.nonooculusnas.com" 2024/11/13 05:59:54 [error] 1552#1552: *34 connect() failed (111: Unknown error) while connecting to upstream, client: 127.0.0.1, server: , request: "HEAD /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "localhost" 2024/11/13 05:59:54 [error] 1552#1552: *34 connect() failed (111: Unknown error) while connecting to upstream, client: 127.0.0.1, server: , request: "HEAD /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "localhost" 2024/11/13 06:02:01 [error] 1552#1552: *41 connect() failed (111: Unknown error) while connecting to upstream, client: 127.0.0.1, server: , request: "HEAD /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "127.0.0.1" 2024/11/13 06:02:01 [error] 1552#1552: *41 connect() failed (111: Unknown error) while connecting to upstream, client: 127.0.0.1, server: , request: "HEAD /healthcheck HTTP/1.1", upstream: "http://127.0.0.1:8000/healthcheck", host: "127.0.0.1"

It seems that https:///healthcheck is being checked on port 8000 even though the port config in compose file uses port 80

###Compose File:

version: "3.9"
services:
onlyoffice:
image: onlyoffice/documentserver:latest
container_name: onlyoffice-server
hostname: OnlyOfficeServer
networks:
- nextcloud_network
ports:
- 8467:80
volumes:
- /volume1/docker/onlyoffice/logs:/var/log/onlyoffice:rw
- /volume1/docker/onlyoffice/data:/var/www/onlyoffice/Data:rw
- /volume1/docker/onlyoffice/config:/etc/onlyoffice/documentserver/:rw
environment:
JWT_ENABLED: true
JWT_SECRET: secret
rejectUnauthorized: false
restart: always
labels:
- com.centurylinklabs.watchtower.enable=true

networks:
nextcloud_network:
external: true

###from default.json we see coauthoring port is 8000
"services": {
"CoAuthoring": {
"server": {
"port": 8000,
"workerpercpu": 1,
"mode": "development",
"limits_tempfile_upload": 104857600,
"limits_image_size": 26214400,
"limits_image_download_timeout": {
"connectionAndInactivity": "2m",
"wholeCycle": "2m"
},

###Curl Command:
ash-4.4# curl -I http://:80/healthcheck
HTTP/1.1 502 Bad Gateway
Server: nginx
Date: Wed, 13 Nov 2024 06:18:43 GMT
Content-Type: text/html
Content-Length: 150
Connection: keep-alive

###curl command ran from inside container:
ash-4.4# docker exec -it onlyoffice-server curl -I http://127.0.0.1:80/healthcheck
HTTP/1.1 502 Bad Gateway
Server: nginx
Date: Wed, 13 Nov 2024 06:18:43 GMT
Content-Type: text/html
Content-Length: 150
Connection: keep-alive

Downgrading to onlyoffice/documentserver:8.2.0 resolved the issue. As of this post the 'latest' tag was updated 16 hours ago.

###Curl command ran after downgrading:
ash-4.4# curl -I http://onlyoffice.nonooculusnas.com:80/healthcheck
HTTP/1.1 404 Not Found
Server: nginx
Date: Wed, 13 Nov 2024 06:51:37 GMT
Content-Type: text/html; charset=iso-8859-1
Connection: keep-alive
Keep-Alive: timeout=20
Vary: Accept-Encoding

I did not know to check this before the issue started and even though its coming back 404 not found, the integration in nextcloud is now working.

Also, I am not sure if the coauthoring port is correct or not, but the main issue here is the healthceck url is coming back as 502 bad gateway.

[igwyd]

Also, I am not sure if the coauthoring port is correct or not, but the main issue here is the healthceck url is coming back as 502 bad gateway.

I get your docker compose file, by the way, it has mistakes:

1. Forbidden to create volume /volume1/docker/onlyoffice/config:/etc/onlyoffice/documentserver/:rw
2. We havent variable rejectUnauthorized: false, use instead USE_UNAUTHORIZED_STORAGE, all configuration parameters here https://github.com/ONLYOFFICE/Docker-DocumentServer?tab=readme-ov-file#available-configuration-parameters

Ran curl inside container:
root@OnlyOfficeServer:/# curl -I http://127.0.0.1:80/healthcheck
HTTP/1.1 200 OK
Server: nginx
Date: Fri, 29 Nov 2024 06:12:29 GMT
Content-Type: text/plain; charset=utf-8
Content-Length: 4
Connection: keep-alive
ETag: W/"4-X/5TO4MPCKAyY0ipFgr6/IraRNs"

And outside (Port 8467 because you map it 8467:80)
root@oods:~/NC-test# curl -I http://ip-address:8467/healthcheck
HTTP/1.1 200 OK
Server: nginx
Date: Fri, 29 Nov 2024 06:15:10 GMT
Content-Type: text/plain; charset=utf-8
Content-Length: 4
Connection: keep-alive
ETag: W/"4-X/5TO4MPCKAyY0ipFgr6/IraRNs"

Everything is ok as you see.

I see you are using an external DNS for request https://onlyoffice.nonooculusnas.com/healthcheck. Is there some proxy before onlyoffice? I think the issue should be there.

[maltokyo]

Did anyone solve this, in terms of how to update the reverse proxy to get a clean return from /healthcheck url?

[igwyd]

Hello @maltokyo we have the proxy examples for onlyoffice documentserver https://helpcenter.onlyoffice.com/installation/docs-community-proxy.aspx

[maltokyo]

Thank you. But this doesn't have a /healthcheck entry at all. This issue for me is that /welcome works fine, but this integration relies on /healthcheck

[igwyd]

It is not there because it is an internal check of the document server, the proxy will redirect it to our internal service. I just checked this config with my docker installation and external nginx - healthcheck shows true as expected.