VPS crashing on Proxmox is usually a resource or config issue, and the logs will tell you exactly which. Generic guides skip the Proxmox-specific stuff, so here’s what to check.
Start with the qemu-server logs
For KVM VMs, the task log is the first place to look. In the web UI, click the VM, then Task History, and open the failed start or stop. The raw log lives in /var/log/pve/tasks/ and the per-VM config is /etc/pve/qemu-server/<vmid>.conf. If the VM crashed while running, check journalctl -u qemu-server@<vmid> for the actual error. Often it’s a missing disk or a bad PCI passthrough.
LXC containers and the OOM killer
Containers crash differently. If an LXC dies silently, it’s probably the OOM killer. Check dmesg | grep -i oom on the host. You’ll see lines like oom-kill:constraint=CONSTRAINT_MEMCG with the container’s cgroup. That means the container hit its memory limit, not the host’s. Raise the memory limit or fix the leak inside the container. Also check pct status <vmid> and the container’s log at /var/log/lxc/<vmid>.log.
Host-level issues that look like VPS crashes
Sometimes the whole node stutters. Check dmesg for hardware errors, journalctl -p err for service failures, and free -h to see if swap is thrashing. If you set swappiness to 1 on a small RAM VM, that can cause OOM inside the guest—see Swappiness 1 Small RAM VM Proxmox: Downsides and OOM Risk. Also verify your storage isn’t full: df -h and zpool list if you’re on ZFS.
Config validation before you reboot
If you made changes and then the VPS started crashing, validate /etc/pve before rebooting. A typo in a config file can stop a VM from starting at all. See Proxmox configuration validation: check /etc/pve files before you reboot.
What I’d do
Check the task log first. If it’s an LXC, look for OOM. If it’s a VM, check the qemu-server journal. Don’t guess—the logs are right there.
