Red Hat Bugzilla – Bug 1261880
[RFE] Consider hard affinity group migration, as opposed to current state (blocking migration)
Last modified: 2016-11-30 06:42:59 EST
Description of problem:
Available to put host to maintenance, when it have at least two vms, that included in hard positive affinity group.
In 3.4 we blocked such possibility, to prevent race condition between vms migration, but now we queue of pending vms and it possible to avoid race condition.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Create hard positive affinity group with two vms(vm_1, vm_2)
2. Start both vms(run on the same host)
3. Put host to maintenance
Detail: [Cannot switch Host(s) to Maintenance mode.
The following Enforcing Affinity Group(s) have running VMs and can break the affinity rule.
maintenance_hard_positive_affinity_group (golden_env_mixed_virtio_0 ,golden_env_mixed_virtio_1)
Please manually migrate the VMs, or change Affinity Group's enforcing to false.]
Action succeed and vms migrated to same host
We need take in account, that destination host will not have resources to run both vms, in such case we need to failed host maintenance with appropriate message