Prevent duplicate MC search peers

[jmeixensperger]

Similar issue described in: #864

The monitoring console role performs 2 peering steps:

1. When indexer clustering is enabled, peer the cluster master. This automatically registers indexers as search peers
2. Manually add every instance as a search peer

This caused duplicate search peers for the indexers, where 1 set of search peers always has the status Down / Sick. This ultimately causes checks in adding_peers.yml to fail.

This PR makes the following changes:

• change the order of above steps so that peering the cluster master happens first
• before adding the remaining search peers, check for existing peers and remove them from our target list