Description of problem:

When user scale the node pool (in pending state after installation of cluster) with new node count value, the UI throws 500 undefined error as below

This was seen with a HCP cluster with multi-az data plane model.

CLI status of machine pool as below

$ rosa list machinepool -c mzap10jayahy

ID             AUTOSCALING  DESIRED REPLICAS  CURRENT REPLICAS  INSTANCE TYPE  LABELS    TAINTS    AVAILABILITY ZONE  SUBNET                    VERSION  AUTOREPAIR  MESSAGE                       
workers-0-1    No           1                 0                 m5.xlarge                          us-east-2c         subnet-03bd2369e65cbf1bb  4.12.9   Yes         WaitingForAvailableMachines   
workers-0-1-2  No           1                 0                 m5.xlarge                          us-east-2b         subnet-0f13e32ba4115fa32  4.12.9   Yes         WaitingForAvailableMachines   
workers-0      No           1                 1                 m5.xlarge                          us-east-2a         subnet-020da5b3818696c4f  4.12.9   Yes    
 

Steps to reproduce:

1. Create a VPC with 3 private subnets & public subnets. 
2. Install ROSA hypershift cluster via CLI (Provide all the subnet values during installation) 
3. Once installation is successful , Launch OCM UI staging.
4. Navigate to machine pools tab.
5. Select one of the node pool listed for a AZ and  click on  the Scale option
6. Update the node count value to valid value(ex: 3)
7. Click on "Apply" button view the behavior.

Actual results:
The node/machine pool changes i.e. node count value failed to update against the selected node/machine pool. The UI thrown 500 undefined error.
Note: The machine pool status shown as "WaitingForAvailableMachines".

Expected results:
The node/machine pool changes i.e. update on node count value should be successful for the selected node/machine pool.

Attached screen capture for the issue NodepoolSA.mp4