Fix: provision config join after leader reboot

[suizman]

Currently after rebooting the leader in our AWS based cluster re-election occurs and some of the followers become a leader.
The problem is that the previous leader now is unable to join the cluster because the join configuration is missing.

We should fix the join cluster configuration and include all the nodes addresses in order to fix this issue and allow all nodes rejoining the cluster after an ordinary service stop/restart.