Add secondary pulsar app staging to the second elixir storage for all production instances

[martindemko]

The idea is about having at least two instances of the pulsar app (possibly on the same VM), staging job working data on both Elixir storages (praha5 and brno11) separately. One is the main, and the other is a backup for now. This will need to be configured in TPV via individual destinations and MQ channels. A simple way (manual reconfiguration) would be in the time of crisis of the main storage to change one line under tools: default: scheduling: require: pulsar[12]. Automatic way would need to use some check and decision point...TBD