Partially successful scale-out requests causes subsequent requests to fail.

Products

• VMware Aria Suite

Issue/Introduction

Cleanup a partially successful scale-out request.

Symptoms:

• A scale-out request competes successfully,  but the additional machines are not created.  Subsequent requests will fail with a cluster size error. 
• A scale-out request fails with the error of

  The data specified within the request is invalid.  Invalid cluster size (8) provided for component [machine-name] of blueprint [blueprint-name] during scale-out operation.  New value must be greater than current cluster size of (10)

Environment

VMware vRealize Automation 7.x

Resolution

1. SSH to the vRealize Automation Primary Appliance

2. Login as root

3. Login to the the vPostgres database vcac as user postgres

   su postgres
   
   psql vcac
   

4. Run the below query to determine the gap in the cluster within the table comp_component_res
   
   select id, name, cafe_resource_id, cluster_index from comp_component_res where dep_id= '<deployment_id>' and name = '<component_name>' order by name;
   
   To find '<deployment_id>':
   
   select * from cat_resource where name = '<deployment_name>' and parentresource_id is null;
   
   Make sure you get one and only return; If yes, then below query:
   
   select id from comp_deployment where name = '<deployment_name>';
   The returned id is '<deployment_id>'.
   
   
   To find '<component_name>':
   You should be able to know the name from UI: Deployments - Components, pick the component name that the scale-out was applied to. See screenshot components.png

Example Output:
 <UUID> | test00002947 | <UUID> | 0
 <UUID> | test00002948 | <UUID> | 1
 <UUID> | test00002949 | <UUID> | 2
 <UUID> | test00002950 | <UUID> | 3
 <UUID> | test00002951 | <UUID> | 4
 <UUID> | test00002952 | <UUID> | 5
 <UUID> | test00002953 | <UUID> | 6
 <UUID> | test00002955 | <UUID> | 8
 <UUID> | test00002960 | <UUID> | 9
 <UUID> | test00002985 | <UUID> | 10

5. In the example output, the gap in the cluster starts with the number "7".

6. Request a scale-in to the cluster number that is missing in the table.

7. Using the example: Request a new scale-in request of "7" to destroy virtual machines 8, 9 and 10 fixing the gap.

8. Using the example:  Request a new scale-out to a value greater than 7 using the vRealize Automation UI.