Please fill in the following information:

Issue description

The step 6 does not result as expected. 
**

Expected result:
**

computehci0.overcloud.example.com

Received result:
**

compute0.overcloud.example.com

Additional details:

learner also reported following issues:{}

1. As /var/log/containers/nova/nova-scheduler.log, computehci0 host is excluded at the beginning, and only compute0 and compute1 two hosts are left for nova scheduler filter.
   [image attached]
2. Checked on dashboard, found there is an unknown instance on computehci0, created 3 years ago.
   [image attached]
3. It looks the strange instance occupied resource on computehci0, so I deleted it. And re-create my instance hci-server1 again.
   This time my instance is created to computehci0 successfully, but this is NOT chosen by aggregate and flavor property settings, but by compute host weight.
   computehci0 happened to get a higher weight, so it was chosen.
   [image attached]
4. After search on the internet, there are 2 additional problmes:
   1. Add AggregateInstanceExtraSpecsFilter to enabled_filters in nova.conf
   2. Add aggregate_instance_extra_specs: prefix to the property of flavor.
   For example, openstack flavor set default-hci --property aggregate_instance_extra_specs:computehci=true

Further details indicate that the case is also impacting https://rol.redhat.com/rol/app/courses/cl210-16.1/pages/ch08s07

Do let me know if any additional details are required.