[BUG] When custom repo is used backup/recovery stops working

[sk4zuzu]

Describe the bug
Backup/recovery uses resolved_repository_hostname which does not point to the correct VM when custom repository is used.

How to reproduce
Steps to reproduce the behavior:

1. use custom repository during epicli apply
2. enable something (anything really) to backup and run epicli backup

Expected behavior
No error, ansible stores backup artifacts correctly on a proper node.

Config files
N/A

Environment

• Cloud provider: all
• OS: all

Additional context
N/A

[rpudlowski93]

• Tested, works as expected ✅

I created new cluster with custom repository and postgresql + load balancer machines and after that I executed backup of the postgresql and load balancer instances and everything passed well, backups are stored on master node.