I need these orchestrators to be able to call each others' workflows remotely when needed.
I tried multi-node plugin, but I couldn't make it to work stable. First of all, when bidirectional communication (that is when each orchestrator is both - Master and Slave for its counterpart) is configured, then orchestrators start to slow down until they are totally unresponsive, and only service restart helps to bring them out from this state for a while. But even when communication is unidirectional (Master -> Slave) then from time to time Master refuses to call any remote workflow until Slave is restarted. Error message says: "Unable to retrieve remote workflow". Finally, there are situations, when Slave was down and is starting up, and if Master tries to communicate to Slave during that time, or Administrator tries to access Multi-Node plugin inventory, when Slave is starting, then Master itself hangs.
That's why I'm investigating other possibilities.
