Uploaded image for project: 'Red Hat OpenStack Services on OpenShift'
  1. Red Hat OpenStack Services on OpenShift
  2. OSPRH-21987

Consistent timeouts waiting for libvirt event during disk detach

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Unresolved
    • Icon: Undefined Undefined
    • None
    • 2026.1 (G), 2025.1 (Epoxy), 2024.2 (Dalmatian), 2024.1 (Caracal), rhos-17.1.z, rhos-18.0.14 FR 4, 2025.2 (Flamingo)
    • openstack-nova
    • None
    • Important

      Downstream tracker for upstream bugs:

      Shift on Stack observed that in https://github.com/kubernetes/cloud-provider-openstack/pull/2742#issuecomment-3135940269 that since the Caracal OpenStack release device detach is significantly (~100%) slower than before as the nova-compute libvirt driver relies on timeout instead of libvirt notifications to detect when the device is removed from the running VM.

      During eventlet removal work of nova-compute service it was noticed that the libvirt event handling of the libvirt driver does not work on a freshly deployed compute but it works after the compute is restarted at least once.

              rh-ee-bgibizer Balazs Gibizer
              rh-ee-bgibizer Balazs Gibizer
              Stephen Finucane
              rhos-workloads-compute
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

                Created:
                Updated: