Cpu model is not correct on Aarch64/Qemu/Custom mode
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
OpenStack Compute (nova) |
Fix Released
|
Undecided
|
Unassigned |
Bug Description
Background:
We'd like to setup Nova and Devstack Upstream CI. Should support launch vm via Qemu virt_type. But host-passthrough mode(default in Aarch64) doesn't work on Qemu. So we can just use "custom" and specify CPU-model.
But actually Aarch64 don't return available cpu models list from Libvirt side. Ref:https:/
So we should use the specified ones from config file. And add default ones to it on Aarch64.
Nova-cpu.conf:
[libvirt]
live_migration_uri = qemu+ssh:
cpu_mode = custom
virt_type = qemu
cpu_model = cortex-a57
WARNING nova.virt.
WARNING nova.virt.
WARNING nova.virt.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
ERROR oslo_service.
Changed in nova: | |
assignee: | nobody → Kevin Zhao (kevin-zhao) |
status: | New → In Progress |
Changed in nova: | |
assignee: | Kevin Zhao (kevin-zhao) → Marcin Juszkiewicz (hrw) |
libvirt+qemu