Detailed bug description:
In case a 'deployment' or 'provisioning' task is in progress and an operator tries to start another deployment, this attempt is validated but the error message is too generic for the operator to understand what happened.
Steps to reproduce:
1. Create an HA env, e.g. 3 controllers + 1 compute
2. Shutdown a controller
3. Wait until the node goes offline
4. Remove the node and start deploying changes, using API call
curl -X "DELETE" http://localhost:8000/api/nodes/<NODE_ID> -H "X-Auth-Token: <AUTH_TOKEN>")
5. Do not wait for deployment end, add a new controller node and start deploying changes
> fuel --env <ENV_ID> node set --node <NODE_ID> --role controller
> fuel --env <ENV_ID> deploy-changes
Expected result: a user friendly error message is thrown giving the operator understanding of what error was made.
Actual result: a generic error message is thrown, which does not describe what happened - '500 Server Error: Internal Server Error for url: http://10.109.0.2:8000/api/v1/clusters/1/changes?dry_run=0 (Unexpected exception, please check logs)'
Description of the environment:
MOS 9.0, ISO #467.
Fix proposed to branch: master
Review: https:/