Description of problem:

Opening case on behalf of customer:

In our traditional infrastructure, we had to double all elements to guarantee high availability. With the PaaS, this is no longer required as the system will recreate containers when a node goes down.  We would like in case of failover to specify which containers should take priority
  
4. Why does the customer need this? (List the business requirements here)  
If we need to provide enough capacity for all containers to fail over that way, we still have to reserve double capacity. We would like to indicate that in case of a node failure, first our production workloads are recreated. If no capacity is left, it should start removing acceptatation or development containers to free up resources.
  
5. How would the customer like to achieve this? (List the functional requirements here)  
A priority per project. When higher priority projects do not meet their required state, containers with lower priority should be removed/moved to free up resources.
  
6. For each functional requirement listed, specify how Red Hat and the customer can test to confirm the requirement is successfully implemented.  
* Create 2 projects : 1 high prio, 1 low prio
* fill a node to its max with several replica's pods of both projects
* kill a node
* see that all replica's of the high prio are recreated and that pods of the low prio are scaled down

8. Does the customer have any specific timeline dependencies and which release would they like to target (i.e. RHEL5, RHEL6)?  
Openshift 3.4
    
10. List any affected packages or components.  
Project structure + container scheduling algorithm
  
11. Would the customer be able to assist in testing this functionality if implemented? 
Yes.

Version-Release number of selected component (if applicable):

Openshift Container Platform