id: PG-05
category: vm-lifecycle
title: ACTIVE in Nova, frozen for users?
situation: >-
  The instance store of a compute is full. QEMU pauses each instance at its next write. Nova keeps showing
  `ACTIVE`, and the guest reports no error: its write just stops.

diagram:
  type: compare
  caption: Three views of the same instance
  columns:
    - kicker: Nova status
      title: ACTIVE
      icon: cloud
      tone: mute
      points:
        - What the instance list shows
        - "Nothing in `nova-compute.log`"
    - kicker: Nova power state
      title: Paused
      icon: pause
      tone: warn
      points:
        - The only signal in Nova
        - "Shown by `server list --long`"
    - kicker: libvirt on the compute
      title: paused (I/O error)
      icon: hard-drive
      tone: stop
      points:
        - The reason, from the hypervisor
        - "Shown by `virsh domstate`"
  note: >-
    Placement counts the disk of the flavors, not the space guests use. The scheduler keeps sending instances
    to the full compute.

checks:
  - label: ACTIVE but paused instances
    hint: openstack server list --all-projects --long | grep -i paused
  - label: Is the instance store full?
    hint: df -h /var/lib/nova/instances
  - label: Why each domain is paused
    hint: virsh domstate <domain> --reason
  - label: What filled it
    hint: "du on the store: instance disks, _base, logs, stray files"

caution:
  - level: destructive
    text: >-
      Never delete or truncate an instance disk, or a `_base` image that a running instance still uses as
      its backing file.
  - level: know
    text: >-
      Freeing space does not resume them, and `server unpause` does not apply. Resume each domain with
      `virsh resume`.

lesson: Nova does not watch the disk. Monitor the instance store and the power state.

runbook:
  path: lifecycle/compute-disk-full-instance-paused.md
  status: lab-verified
  release: "2026.1"

scope: >-
  Lab-verified on OpenStack 2026.1 with Kolla-Ansible, local qcow2 instance disks and CirrOS guests. Not
  tested with Ceph or other shared instance storage, volume-backed instances, or raw and LVM image backends.
  The pause comes from QEMU's default write-error policy; other releases were not checked. With
  Kolla-Ansible the store is the nova_compute Docker volume, and virsh runs in the nova_libvirt container.

linkedin:
  title: "ACTIVE in Nova, frozen for the user: the silent full disk"
  caption: |
    A user says their VM "hangs". Nova says ACTIVE. Nothing in the nova-compute log.

    The compute's instance store is full. QEMU paused the instance on its next write, and it will wait there, losing nothing, until someone resumes it.

    Three views of the same instance:
    → Nova status: ACTIVE
    → Nova power state: Paused (the only signal in Nova)
    → libvirt: paused (I/O error)

    Placement does not help: it counts the disk of flavors, not the space guests actually use, so the scheduler keeps placing instances on the full host.

    Freeing space is not enough. The domains stay paused until they are resumed.

    Monitor the instance store, and alert on "ACTIVE but Paused".
  hashtags: [OpenStack, Nova, Libvirt, Monitoring, CloudOperations, OzarLand]
