In this case, if net_kernel:start/1 is asynchronous or if it's otherwise possible to continue (which sometimes failing tests would lead me to believe), there is a chance that node() will return it's previous value and in that case the erlang:set_cookie/2 call will fail.
If this turns out to be the case, then a slightly modified version would make sense:
I suspect that in
test_fmke_operations_travis:start/0
there is an issue with the code that sets the node cookie:In this case, if
net_kernel:start/1
is asynchronous or if it's otherwise possible to continue (which sometimes failing tests would lead me to believe), there is a chance thatnode()
will return it's previous value and in that case theerlang:set_cookie/2
call will fail.If this turns out to be the case, then a slightly modified version would make sense: