Ok, I am answering my own question. This was just a noob being noob…, anyway after struggling for almost two days with this the answer to the question is simple: The events are constantly firing, probably, at the same time, this caused to look like if only one of them was executing, nothing further from the truth. The solution was just to add the corresponding branch node to validate, using a boolean variable, which event stream of execution should reach the socket offset update, in other words, a switch.