CRITICAL: MongoDB failure on Proxmox LXC after Kernel 7.0.2 update

Graylog Central (peer support)
1

Hi all,

I wanted to share a critical finding for anyone running Graylog on Proxmox LXCs. Following a host update to Kernel 7.0.2-2-pve, my MongoDB service (v8.0.21) failed to start, effectively taking down the entire Graylog stack.

The Symptom

  • Graylog web interface is unreachable.

  • mongod.service fails with status=1/FAILURE.

  • Crucially: No logs are written to /var/log/mongodb/mongod.log. The process exits before it can initialize logging.

The Cause

This appears to be a direct incompatibility between the Linux 7.0.2 kernel memory management and MongoDB’s storage engine (related to the SERVER-121912 JIRA issue). When running in an LXC, the containerized MongoDB process cannot handle the memory-mapping changes introduced in this kernel version, leading to an immediate silent crash.

The Proof

I verified this by migrating the Graylog container from my production server to a test machine:

  • Production (Failing): Proxmox 9.1.9 w/ Kernel 7.0.2-2-pve

  • Test (Working): Proxmox 9.1.9 w/ Kernel 6.17.13-3-pve

The exact same LXC started perfectly on the 6.x kernel without any configuration changes.

Recommended Action for Graylog Admins

If you are planning updates this week:

  1. Check your Kernel: If your update includes a jump to Kernel 7.x, be prepared for MongoDB to fail.

  2. Hold/Rollback: If you’ve already updated and are seeing silent MongoDB crashes, reboot your host and select a 6.1x.x kernel from the GRUB menu.

  3. LXC Settings: If you must stay on the 7.x kernel, you may need to set the container to unconfined and disable THP (Transparent Huge Pages) on the host, though a kernel rollback is currently the most stable fix.

Has anyone else encountered this on other distros using the 7.0+ kernel, or is this limited to the PVE implementation?

2

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.