celestia-node: Kernel Panic

Celestia Node version

0.9.5, but we experienced the same problem also on previous versions

OS

Ubuntu 22.04.2 LTS

Install tools

The node is running on a proxmox container using a systemd service with this the following ExecStart:

/usr/local/bin/celestia light start --core.ip https://grpc-blockspacerace.pops.one/ --core.grpc.port 9090 --keyring.accname my_celes_key --metrics.tls=false --metrics --metrics.endpoint otel.celestia.tools:4318 --gateway --gateway.addr 0.0.0.0 --gateway.port 26659 --p2p.network blockspacerace

Actual result

As we already told here (https://medium.com/@openbitlab/our-celestia-light-node-performance-in-the-itn-9a7291b9b90b), our Celestia dedicated server is randomly crashing with a kernel panic; this time we were able to log the klog using netconsole. Even if we changed (for this reason) the server, the farm and the provider, the issue is still happening on the new server.

A similar issue is happening also to another user: https://github.com/celestiaorg/celestia-node/issues/2097#issuecomment-1523015566

If @blockonaut solved the problem by using an Intel CPU, we suspect is an issue with AMD Ryzen CPUs.

Relevant log output

https://gist.github.com/dakk/3fb35a850ce29ad6e45178503fb4cba2

Notes

No response