Graylog server repeatedly crashing on startup - TCPNIOBindingHandler issue

Graylog Central (peer support)
1

Hi Community,

I am running into an issue where our Graylog server keeps shutting down immediately after attempting to start.

The /var/log/graylog-server/server.log indicates a port binding conflict. The server throws a stack trace originating from sun.nio.ch.Net.bind0 and org.glassfish.grizzly.nio.transport.TCPNIOBindingHandler.bindToChannelAndAddress, followed immediately by a graceful shutdown signal.

Here is a quick overview of our environment pulled from the logs:

  • Graylog Version: 4.3.15

  • MongoDB: 4.4.30

  • Elasticsearch: 6.8.23

  • OS/Java: Ubuntu 22.04 / Java 1.8.0_482

The MongoDB logs (/var/log/mongodb/mongod.log) appear normal, mostly just showing the Graylog client repeatedly connecting and dropping as the service restarts in a loop.

Could you help me verify which specific port is likely causing this conflict (e.g., the REST API or Web Interface port) and the best way to resolve this in the server.conf?

I have included snippets of the stack trace and the startup sequence below for reference.

Graylog server.log snippet:

at sun.nio.ch.Net.bind0(Native Method) ~[?:1.8.0_482]
at sun.nio.ch.Net.bind(Net.java:461) ~[?:1.8.0_482]
...
at org.glassfish.grizzly.nio.transport.TCPNIOBindingHandler.bindToChannelAndAddress(TCPNIOBindingHandler.java:107) ~[graylog.jar:?]
...
2026-05-22T15:29:43.710Z INFO  [Server] SIGNAL received. Shutting down.
2026-05-22T15:29:43.725Z INFO  [GracefulShutdown] Graceful shutdown initiated.

Thanks for your help.

Best regards,

Ahnaf Tahmeed

2

Hey @ahnaf702,

Where doe the pre-flight log reach before hitting this issue? Within the server.conf you could try altering the bind address as a starting point.