Mattermost plugin failure

I am trying to connect to documentserver via MM plugin.
The server is working inkl example) and is reacheable via registered domain.
When I am pointing in my plugin to https:// dcoumen.domain.net with correct Header, secret, and Prefix, I see an 500 internal error.
MM and OO are on the same server in different dockers. NGINX is providing the LE Certificates and pointing to the http: ports in both appl.
Any Hints?
Thank You

DOC

Hello @doc

Please share versions of used connector app, Document Server and Mattermost.
Also, please check mutual accessibility of both servers with wget <address>, execute from Document Server to Mattermost and vice versa.

By the way, do I understand that both apps are behind the NGINX proxy?

1 Like

Thank you.
Both are behind nginx.
Nginx provides LE certificate for MM and points to localIp:8065.
Onlyoffice is in different container with portmapping 8199 to 80 and 7443 to 443.

Problem: nr 1
I have a Mattermost backup server Teams edition 192.168.0.171:8065.
Documentserver answers in Browser pointing to URL http:// 192.168.200.22:8199 perfectly.

When pointing to 192.168.200.22:8199 (onlyoffice) from Mattermost ( 192.168.0.171) the plugin works fine.

The second MM server, Professional licence, is located 192.168.200.22:8065
When pointing to http:// 192.168.200.22:8199 (onlyoffice) from Mattermost Prof Ed. (192.168.200.22) the plugin starts loading and gives 500 Error.

Problem n2

In nginx i simply point the domain as proxyhost to 192.168.200.22:8065 and everything works fine. Inkl. LE.

So it doesn’t in Onlyoffice.
The domain never connects to onlyoffice. I spent hours read your website OO behind nginx.
Nothing.

Ngingx is located by 192.168.200.22:81 and gets poots 81, 80 and 443.

Mattermost latest 10.9
Onlyoffice latest 9.
Plugin latest 1.2

Thanks for your help

meanwhile i am able to connect to my domain via http
and get the welcome page of OO.
But:
Plugin says:
This Plugin could not be started. [ONLYOFFICE CMD]: could not parse document server version

but only via
Server in http:
Https: not yet testable (too may tries to get certificate)

Thank You Klaus

Since i was desperatly trying to get MM plugin to work behind nginx. I gave up. an installed OnlyOffice on an separate Server.
This was more or less flawless.
ist is reacheabl and ssl:

https;//d.domain.net

MM Plugin installs normaly and configuring is normal.

But when I try to open any text file i get an internal error.

this is my mattermost log anonymized:

info […] Installing extracted plugin caller=“app/plugin_install.go:429” plugin_id= version=1.2.0
info […] Persisting plugin to filestore caller=“app/plugin_install.go:230” plugin_id=
warn […] No signature when persisting plugin to filestore caller=“app/plugin_install.go:233” plugin_id=
warn […] Unrecognized config permissions tag value. caller=“api4/config.go:448” path=/api/v4/config/patch request_id= ip_addr= user_id= method=PUT tag_value=“sysconsole_write_read"
warn […] Unrecognized config permissions tag value. caller=“api4/config.go:448” path=/api/v4/config/patch request_id= ip_addr= user_id= method=PUT tag_value="sysconsole_write
read"
error […] RPC call to OnActivate plugin failed. caller=“plugin/client_rpc.go:272” plugin_id= error=“unexpected EOF”
error […] plugin process exited caller=“plugin/hclog_adapter.go:79” plugin_id= wrapped_extras=“… exit status 2”
error […] RPC call OnConfigurationChange to plugin failed. caller=“plugin/client_rpc_generated.go:68” plugin_id= error=“connection is shut down”
warn […] Health check failed for plugin caller=“plugin/health_check.go:59” id= error=“plugin RPC connection is not responding”
warn […] error closing client during Kill caller=“plugin/hclog_adapter.go:70” plugin_id= wrapped_extras=“errconnection is shut down”
warn […] plugin failed to exit gracefully caller=“plugin/hclog_adapter.go:72” plugin_id=
error […] RPC call to OnActivate plugin failed. caller=“plugin/client_rpc.go:272” plugin_id= error=“unexpected EOF”
error […] plugin process exited caller=“plugin/hclog_adapter.go:79” plugin_id= wrapped_extras=“… exit status 2”
warn […] Health check failed for plugin caller=“plugin/health_check.go:59” id= error=“plugin RPC connection is not responding”
warn […] error closing client during Kill caller=“plugin/hclog_adapter.go:70” plugin_id= wrapped_extras=“errconnection is shut down”
warn […] plugin failed to exit gracefully caller=“plugin/hclog_adapter.go:72” plugin_id=
error […] RPC call to OnActivate plugin failed. caller=“plugin/client_rpc.go:272” plugin_id= error=“unexpected EOF”
error […] plugin process exited caller=“plugin/hclog_adapter.go:79” plugin_id= wrapped_extras=“… exit status 2”
warn […] Health check failed for plugin caller=“plugin/health_check.go:59” id= error=“plugin RPC connection is not responding”
warn […] error closing client during Kill caller=“plugin/hclog_adapter.go:70” plugin_id= wrapped_extras=“errconnection is shut down”
warn […] plugin failed to exit gracefully caller=“plugin/hclog_adapter.go:72” plugin_id=
warn […] Unrecognized config permissions tag value. caller=“api4/config.go:448” path=/api/v4/config/patch request_id= ip_addr= user_id= method=PUT tag_value="sysconsole_write
read"
warn […] Unrecognized config permissions tag value. caller=“api4/config.go:448” path=/api/v4/config/patch request_id= ip_addr= user_id= method=PUT tag_value="sysconsole_write
_read”
error […] RPC call to OnActivate plugin failed. caller=“plugin/client_rpc.go:272” plugin_id= error=“unexpected EOF”
error […] plugin process exited caller=“plugin/hclog_adapter.go:79” plugin_id= wrapped_extras=“… exit status 2”
error […] RPC call OnConfigurationChange to plugin failed. caller=“plugin/client_rpc_generated.go:68” plugin_id= error=“connection is shut down”
error […] Plugin failed to ServeHTTP, RPC call failed caller=“plugin/client_rpc.go:456” plugin_id= error=“connection is shut down”
error […] Plugin failed to ServeHTTP, muxBroker couldn’t accept connection caller=“plugin/client_rpc.go:411” plugin_id= serve_http_stream_id=3 error=“timeout waiting for accept”
error […] Plugin failed to ServeHTTP, muxBroker couldn’t Accept request body connection caller=“plugin/client_rpc.go:430” plugin_id= error=“timeout waiting for accept”
warn […] Health check failed for plugin caller=“plugin/health_check.go:59” id= error=“plugin RPC connection is not responding”
warn […] error closing client during Kill caller=“plugin/hclog_adapter.go:70” plugin_id= wrapped_extras=“errconnection is shut down”
warn […] plugin failed to exit gracefully caller=“plugin/hclog_adapter.go:72” plugin_id=
error […] RPC call to OnActivate plugin failed. caller=“plugin/client_rpc.go:272” plugin_id= error=“unexpected EOF”
error […] plugin process exited caller=“plugin/hclog_adapter.go:79” plugin_id= wrapped_extras=“… exit status 2”

And now? MM 10.9 Professional Ed. Plugin latest version

Have you checked logs of Document Server? Also, what is the installation type of Document Server (Docker, DEB/RPM packages, EXE/MSI)?

Yes. Dokumentserver is on docker and uses its on Resources.
It resolves perfectly via https with own public domain

Mattermost plugin is v1.2
Mattermost is 10.9.1

Plugin crashes in the moment, when one is trying to open document 500 internal error.
Mattermost is on docker as well, behind
nginx, which gives https.
Both have their lets encrypt certificates.

mattermost log:

error [2025-06-25 20:37:49.764 Z] RPC call to OnActivate plugin failed. caller=“plugin/client_rpc.go:272” plugin_id=com.onlyoffice.mattermost error=“unexpected EOF”
error [2025-06-25 20:37:49.764 Z] plugin process exited caller=“plugin/hclog_adapter.go:79” plugin_id=com.onlyoffice.mattermost wrapped_extras=“pluginplugins/com.onlyoffice.mattermost/server/dist/plugin-linux-amd64id205errorexit status 2”
error [2025-06-25 20:37:49.772 Z] RPC call OnConfigurationChange to plugin failed. caller=“plugin/client_rpc_generated.go:68” plugin_id=com.onlyoffice.mattermost error=“connection is shut down”
error [2025-06-25 20:38:05.175 Z] Plugin failed to ServeHTTP, RPC call failed caller=“plugin/client_rpc.go:456” plugin_id=com.onlyoffice.mattermost error=“connection is shut down”
error [2025-06-25 20:38:10.178 Z] Plugin failed to ServeHTTP, muxBroker couldn’t Accept request body connection caller=“plugin/client_rpc.go:430” plugin_id=com.onlyoffice.mattermost error=“timeout waiting for accept”
error [2025-06-25 20:38:10.178 Z] Plugin failed to ServeHTTP, muxBroker couldn’t accept connection caller=“plugin/client_rpc.go:411” plugin_id=com.onlyoffice.mattermost serve_http_stream_id=3 error=“timeout waiting for accept”
warn [2025-06-25 20:38:11.823 Z] Health check failed for plugin caller=“plugin/health_check.go:59” id=com.onlyoffice.mattermost error=“plugin RPC connection is not responding”
warn [2025-06-25 20:38:11.823 Z] error closing client during Kill caller=“plugin/hclog_adapter.go:70” plugin_id=com.onlyoffice.mattermost wrapped_extras=“errconnection is shut down”
warn [2025-06-25 20:38:11.823 Z] plugin failed to exit gracefully caller=“plugin/hclog_adapter.go:72” plugin_id=com.onlyoffice.mattermost
error [2025-06-25 20:38:12.568 Z] RPC call to OnActivate plugin failed. caller=“plugin/client_rpc.go:272” plugin_id=com.onlyoffice.mattermost error=“unexpected EOF”
error [2025-06-25 20:38:12.568 Z] plugin process exited caller=“plugin/hclog_adapter.go:79” plugin_id=com.onlyoffice.mattermost wrapped_extras=“pluginplugins/com.onlyoffice.mattermost/server/dist/plugin-linux-amd64id214errorexit status 2”

document server logs are empty

Plugin crashes on startup

Would it be possible to check out if the issue reproduces without proxy in front of Mattermost? There should be no problem connecting HTTPS Document Server to HTTP Mattermost for the test.

So i did this:
I have a second teams Edition - this works fine
The professional Edition accessed via local IP
Crashs.
When I test a MM entrprise edition with plugin and https:// documentserver it works.
So thr seems to be a preoblem with nginx.

btw. are ther known any incompatibilities with other plugins?
the server has theese pugins installed:

Any Suggestions w r. to nginx?

this is my config:



grafik

Any help is welcome.

here some mor log entrys from the mattermost log:
panic: runtime error: invalid memory address or nil pointer dereference caller=“plugin/hclog_adapter.go:54” plugin_id=com.onlyoffice.mattermost
…
RPC call to OnActivate plugin failed … error=“unexpected EOF”
…
plugin process exited … exit status 2
…
Health check failed for plugin … error=“plugin RPC connection is not responding”
Deactivating plugin due to multiple crashes
…
Access to route for non-existent plugin … error=“plugin not found: com.onlyoffice.mattermost”

Thank you for the test. Indeed, it seems to be related to the proxy configuration.

Just to clarify my understanding: the setup without proxy works, is that correct? The one that crashes has proxy in front, right?

No, not that I am aware of.


In general, this could be an issue with request forwarding via HTTP, considering internally both recourses communicate over HTTP. You can check out Mattermost community for hints, for instance:

Another issue to address: some systems does not allow using Authorization as a JWT secret. You can also try changing it to see if was the case, for instance, to AuthorizationJwt or something like that.

Setup 1 ist like This:
Mattermost Server1 - only local Mattermost Team Ed → Webinterface via http:// lokal-IP:8065.

Onlyoffice Server : bulidt only with inbuildt nginx gets IP via DYNS und is publicly availalbe:
https://document.servername.de
in Mattermost Server 1: Plugin is configuered with credentials and point to https://document.servername.de

Setup 1 is working.

Setup 2:
Mattermost Server2 is behind nginx with Nginx Proxamanager (gets LE Certificates). and publicly available. https://MM.Servername.de

in Nginx proxy manager , I added under location the bove mentioned Proxy header.

Plugin is configured as Server 1

This plugin crashes

All the best and thank You

@Constantine

Dear All,

After installing the OnlyOffice plugin for Mattermost (com.onlyoffice.mattermost) on our Mattermost Professional instance (v10.9.1, Ubuntu 24.04, glibc 2.39), the plugin fails to start. The logs show a recurring segmentation fault and “invalid memory address or nil pointer dereference” (SIGSEGV) on plugin activation. Mattermost repeatedly tries to restart the plugin, but each time it crashes immediately with the same error. Health checks then disable the plugin.

Relevant log entries:

panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 ...]
RPC call to OnActivate plugin failed. error="unexpected EOF"
plugin process exited ... errorexit status 2
Health check failed for plugin id=com.onlyoffice.mattermost error="plugin RPC connection is not responding"

We have checked file permissions and ownership. The issue persists.
Please advise if this is a known incompatibility or limitation, or if a specific build is needed for the Professional edition

Please let me know if you want it even more concise or with specific environment/build info!

log in the att.

Just to be clear – earlier you mentioned that the setup without NGINX proxy is working, does it mean that this particular setup 2 works without proxy or it was related to the setup 1?

@Constantine these are two different computers.

Setup 1 is holding the onlyoffice Server, a PostgresSQL for MM and has a local Mattermost Teams edition for backup and testing purposes.
Onlyoffice is directly connectet to the internet. It gets Ports 80 and 443.

It is publicly reacheable by https://

Mattermost ist not connectet to the internet. It has the same Postgresdatabase and the same MMdata as setup 2 has.
To connect to this MM I am using → MMlocalIP:8065. the Plugin is configured https://<document.Onlyofficeserver.de>, MySecret. AuthorizationJWT and Bearer .

This setup has no Problems to connect to OO and the Plugin runs smoothly.

Setup 2 is on a different computer. It only holds a Mattermost Professional Edition, Postfix, and the PostgresSQL for this MM. they are all behind Nginx, managed by Nginx Porxy Manager.
This is may productive Installation for MM.
Member reach this Plattform via https://<forum.MMserver.de>
the Plugin is installed and configured by copy - paste like the one in Setup1.
This plugin crashes on startup and gives the above mentioned errors in the Mattermost log.
There are no errors in the logs of Nginx, or in the Onlyoffice logs.

panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 …]
RPC call to OnActivate plugin failed. error=“unexpected EOF”
plugin process exited … errorexit status 2
Health check failed for plugin id=com.onlyoffice.mattermost error=“plugin RPC connection is not responding”

It is the same Plugin version as it is in Setup1. V1.2.0.
And it should connect to the OnlyOffice Server form Setup1

Both Computers Setup1 and Setup2 are identical, with the same Softwareversion and hardware version.

I really hope this helps.!
Thank Ypu for advice

@Constantine

Now in Setup 1 (Testsetup)

Mattermost and OO are behind 1 Nginx.
Both are accessibel from outside via https:// and the Plugin is working fine.
Issue must be unrelated to nginx.
Testsytsem will become productive System - just need to change Databases an Licence

Hope this resolves the issue

I am glad to know. Should I mark the topic as solved then?

@Constantine
Partially. The crash is related to the professional licence in mattermost.
When i take the license away the plugin works even in the productive system.
When i add the license again.
It crashes.

Thank you for the sum up. We are checking the situation with tariffs, I’ll provide an update once it becomes available.