Ability to work with anti-affinity rule finitely

Registered by Prashant Zanwar

Currently ServerGroupAntiAffinity works indefinitely, goes on till all hypevisors are exhausted in the aggregate and match the meta properties of flavors.
1. Logical grouping of physical compute hypervisors, which would restrict work load between n-number of hypervisors.
2. Ability to use same anti affinity group multiple times, which would avoid creating groups in same host aggregate group multiple times ..

1. current ServerGroupAntiAffinity can extend it's defination to include hypervisors under it.
2. While deploying application instances, there should be tag for application as meta policy like - myapp:prod or myapp:stage, myapp:dev

It will look like below -

ID Name Policies Members Matadata Hypervisors Application-Metadata
XXX XXX-Anti anti-affinity/affinity VMs {} x,y,z,a,b myapp:prod
XXX XXX-Anti anti-affinity/affinity VMs {} x,y myapp:dev
XXX XXX-Anti anti-affinity/affinity VMs {} a,b,c,d myapp:stage

so in the end it would be possible to use same rule, with different apps, and same or different hypervisors.
shall be possible to either add hypervisors on command line, or list a range. Or apply rule to whole aggregate.

Blueprint information

Status:
Complete
Approver:
Sergey Lukjanov
Priority:
Undefined
Drafter:
Prashant Zanwar
Direction:
Needs approval
Assignee:
None
Definition:
Obsolete
Series goal:
None
Implementation:
Unknown
Milestone target:
None
Completed by
Vitalii Gridnev

Related branches

Sprints

Whiteboard

(?)

Work Items

This blueprint contains Public information 
Everyone can see this information.

Subscribers

No subscribers.