Public bug reported: With the change of https://review.openstack.org/#/c/465160, NUMA related features like CPU pinning, hugepages, and realtime are now explicitly disabled when using libvirt driver with `virt_type=qemu`, and compute hosts with libvirt/qemu driver are filtered out with NUMATopologyFilter.
This is because qemu with the TCG(tiny code generator) backend doesn’t support cpu pinning and currently nova uses cpu pinning implicitly when NUMA related features are specified in image/flavor. However, qemu with the TCG backend potentially has capability to support some NUMA related features such as NUMA topology and hugepages. So, we should change codes to enable libvirt/qemu driver to support these NUMA features without cpu pinning. Steps to reproduce ================== - deploy nova with virt_type=qemu and enabled_filters=NUMATopologyFilter in nova.conf. - set hw:numa_nodes=1 or hw:mem_page_size=large in the nova flavor. - boot an instance with this modified flavor. Expected result =============== VM starts with a virtual numa topology of 1 numa node(hw:numa_nodes=1) or with hugepage backed memory (hw:mem_page_size=large). Actual result ============= nova reports no valid host error because of the NUMATopologyFilter. Note: Remenber to update at least these documents when they are enabled. https://docs.openstack.org/nova/pike/admin/huge-pages.html https://docs.openstack.org/nova/pike/admin/cpu-topologies.html ** Affects: nova Importance: Undecided Status: New -- You received this bug notification because you are a member of Yahoo! Engineering Team, which is subscribed to OpenStack Compute (nova). https://bugs.launchpad.net/bugs/1737449 Title: [libvirt] virt_type=qemu doesn't support NUMA related features Status in OpenStack Compute (nova): New Bug description: With the change of https://review.openstack.org/#/c/465160, NUMA related features like CPU pinning, hugepages, and realtime are now explicitly disabled when using libvirt driver with `virt_type=qemu`, and compute hosts with libvirt/qemu driver are filtered out with NUMATopologyFilter. This is because qemu with the TCG(tiny code generator) backend doesn’t support cpu pinning and currently nova uses cpu pinning implicitly when NUMA related features are specified in image/flavor. However, qemu with the TCG backend potentially has capability to support some NUMA related features such as NUMA topology and hugepages. So, we should change codes to enable libvirt/qemu driver to support these NUMA features without cpu pinning. Steps to reproduce ================== - deploy nova with virt_type=qemu and enabled_filters=NUMATopologyFilter in nova.conf. - set hw:numa_nodes=1 or hw:mem_page_size=large in the nova flavor. - boot an instance with this modified flavor. Expected result =============== VM starts with a virtual numa topology of 1 numa node(hw:numa_nodes=1) or with hugepage backed memory (hw:mem_page_size=large). Actual result ============= nova reports no valid host error because of the NUMATopologyFilter. Note: Remenber to update at least these documents when they are enabled. https://docs.openstack.org/nova/pike/admin/huge-pages.html https://docs.openstack.org/nova/pike/admin/cpu-topologies.html To manage notifications about this bug go to: https://bugs.launchpad.net/nova/+bug/1737449/+subscriptions -- Mailing list: https://launchpad.net/~yahoo-eng-team Post to : yahoo-eng-team@lists.launchpad.net Unsubscribe : https://launchpad.net/~yahoo-eng-team More help : https://help.launchpad.net/ListHelp
