Linux VPS

Proxmox troubleshooting: fix common VE errors

Solve frequent Proxmox VE problems: locked VMs, web UI not loading, VMs that will not start, full storage, lost quorum, and network loss after configuration changes.

IntermediateProxmox VE2 min read

This Proxmox troubleshooting guide covers the errors admins meet most often and the commands that fix them. Always read the task log first: double-click the failed task at the bottom of the web UI for the full error.

Before you start

  • Root shell access through SSH, the web UI shell or IPMI console.
  • The VM ID and the exact error text.

"VM is locked"

A backup, snapshot or migration left a lock behind. Make sure no task is still running, then:

qm unlock 105

For containers use pct unlock 105.

Web UI on port 8006 does not load

  1. Check the services: systemctl status pveproxy pvedaemon pve-cluster.
  2. Restart the proxy: systemctl restart pveproxy.
  3. Check root disk space with df -h; a full root filesystem stops cluster services.
  4. Check that the firewall still allows your IP on 8006 (firewall setup).

VM fails to start

  • "KVM virtualisation configured, but not available": enable VT-x/AMD-V in the BIOS.
  • Not enough memory: free RAM on the node or lower the VM's memory; check free -g.
  • Storage not available: an NFS share or disk is offline; check Datacenter → Storage.
  • ISO missing: the CD drive points to a deleted ISO; set it to Do not use any media.

Storage full

  • Check thin pools with lvs and ZFS with zpool list.
  • Delete old snapshots and old backups, and enable discard in VMs so freed space returns to the pool.
  • A full thin pool can corrupt VMs; act before it reaches 100%.

Cluster without quorum

  1. Check with pvecm status.
  2. Bring the missing nodes back or fix the corosync network.
  3. Only in an emergency, on a single surviving node, lower expected votes with pvecm expected 1 to regain write access; undo the situation properly afterwards.

Warning: Forcing quorum while other nodes are still running can cause split-brain. Use it only when you are sure the other nodes are down.

Lost network after a change

From the console, review /etc/network/interfaces, fix the bridge or IP, and apply with ifreload -a. See bridges and VLANs.

Frequently asked questions

Where are Proxmox logs?

Use journalctl -xe and the task log in the web UI; syslog is under Node → System → Syslog.

Is it safe to restart pveproxy?

Yes, running VMs are not affected.

Why does apt fail with 401?

The enterprise repository is enabled without a subscription (repositories).

Next steps

Was this guide helpful?

Still stuck?

Ask MdmHost AI for a quick answer, or open a ticket and our team will take a look. We are here 24/7.