Knowledge Base
Troubleshooting and solution know-how — search, or ask AI Assistant.
Change the management network gateway on a hostBack up the configuration, change the management bridge gateway in sequence, and restart networking to update a host's management network gateway.Updated: 2026-09-17VM creation fails because lvmlockd leaks VG shared locks after storage disconnectsAfter a temporary storage disconnect, clear unreleased VG shared locks in lvmlockd that prevent exclusive locks from being acquired and restore volume and VM creation.Updated: 2026-09-17Upgrade ZStack Cube HCI to 2.4.17 in a dual-management-node HA environmentPerform prechecks, backups, and HA toolkit and cloud-platform upgrades to upgrade ZStack Cube HCI to 2.4.17 in a dual-management-node HA environment.Updated: 2026-09-17Resynchronizing a dual-MN database and upgrading the zsha2 HA suiteResynchronize an out-of-sync database between two ZStack Cloud management nodes, then upgrade the zsha2 HA suite and verify platform health.Updated: 2026-08-17Physical NIC disconnect and reconnect alarms caused by an unsupported SFP moduleResolve repeated physical NIC disconnect and reconnect alarms by identifying an unsupported SFP module and replacing it with a vendor-approved module.Updated: 2026-08-17VM startup fails with "mediated device not found" after vGPU repartitioningWhen a VM bound to a vGPU (virtual GPU) fails to start with a libvirt \"device not found: mediated device\" error, the broken vGPU binding can be removed and the VM started on a healthy vGPU.Updated: 2026-08-17Volume snapshot creation fails with "unable to find volume" due to a path mismatchFix a mismatched primary-storage UUID and installPath between the management database and the actual qcow2 path on the host, then merge old snapshots.Updated: 2026-08-17Deleting an abnormal empty storage pool from a ZStone clusterWhen a ZStone Ceph cluster shows HEALTH_WARN because of an empty storage pool with a malformed name, the pool can be removed with ceph osd pool delete after the customer confirms the change plan, global HA is disabled, and the database is backed up.Updated: 2026-08-17