Вот исправленный и улучшенный вариант вашего сообщения:
Versions of the related components and OS (frontend, hypervisors, VMs):
OpenNebula 6.6.1.1, Ubuntu 22.04
Steps to reproduce:
I’m trying to limit the bandwidth used for VM migration between hosts. The storage is local, so during migration the VM consumes the entire available network bandwidth, affecting other VMs.
I added the following line:
exec_and_log "virsh --connect $LIBVIRT_URI migrate-setspeed $DEPLOY_ID --bandwidth 500" \
"Setspeed to 500M is set!"
to /var/lib/one/remotes/vmm/kvm/migrate and /var/lib/one/remotes/vmm/kvm/migrate_local. This works for live migration.
However, when the migration is not live (i.e., save-restore), the bandwidth limit doesn’t seem to apply.
What should I modify to limit the migration speed for non-live (save-restore) migrations as well?