1. Describe your incident:
Hi there. I’m new to Graylog and have tried searching these forums as well as wider Google searches but not coming up with the same issue. So trying here.
I successfully got Graylog up in running using Docker Compose on a test Alma Linux 9 VM with no issues. However when trying the same process on a Debian 12 (Proxmox Backup Server) Graylog doesn’t start.
When I start Docker Compose without -d I can see the startup processing which is full errors (see below).
2. Describe your environment:
-
OS Information:
Debian GNU/Linux 12 (bookworm) [PBS] -
Package Version:
mongo:
image: mongo:6.0.5-jammy
opensearch:
image: opensearchproject/opensearch:2
graylog:
image: graylog/graylog:6.2
- Service logs, configurations, and environment variables:
mongodb | {"t":{"$date":"2025-05-15T14:09:06.883+00:00"},"s":"E", "c":"CONTROL", "id":20558, "ctx":"initandlisten","msg":"std::exception in initAndListen, terminating","attr":{"error":"open: Permission denied"}}
opensearch | BindTransportException[Failed to bind to 0.0.0.0:[9300-9400]]; nested: ChannelException[Unable to create Channel from class class org.opensearch.transport.CopyBytesServerSocketChannel]; nested: InvocationTargetException; nested: NoClassDefFoundError[Could not initialize class sun.nio.ch.ServerSocketChannelImpl]; nested: ExceptionInInitializerError[Exception java.lang.ExceptionInInitializerError [in thread "main"]];
opensearch | Likely root cause: java.lang.ExceptionInInitializerError: Exception java.lang.ExceptionInInitializerError [in thread "main"]
Plus lots of Java errors:
graylog | 2025-05-15 17:09:23,027 ERROR: org.graylog2.bootstrap.CmdLineTool - Startup error:
graylog | java.lang.ExceptionInInitializerError: null
graylog | at java.base/sun.nio.ch.FileChannelImpl.<init>(Unknown Source) ~[?:?]
graylog | at java.base/sun.nio.ch.FileChannelImpl.open(Unknown Source) ~[?:?]
graylog | at java.base/sun.nio.fs.UnixChannelFactory.newFileChannel(Unknown Source) ~[?:?]
&
Caused by: java.io.IOException: Permission denied
It does give this:
"For complete error details, refer to the log at /usr/share/opensearch/logs/docker-cluster.log"
But being the container (which is now stuck in a boot loop) I can’t get to the docker-cluster.log fast enough to read/grab it.
3. What steps have you already taken to try and solve the problem?
I’ve run docker compose down and then up again after removing the local files to see if that helps but as these are all self contained Docker images I’m not sure what else I can do?
Other than changing the time zone in the docker-compose.yml I didn’t configure my previous successful deployment so I must be missing something here…
4. How can the community help?
I am hoping someone recognises this and can point me in the right direction?
Like I said as far as I can determine the only difference between my succesfull deployment and this unsuccessful one is that the first was an Alma 9 VM whilst this is Debian 12 physical (whilst actually a Proxmox Backup Server).
On that: I’ll try the deployment steps on a plain Debian 12 VM to rule out Proxmox being the problem here. Will update with the result shortly.
Thank you.
