project reactor - ParallelFlux vs flatMap() for a Blocking I/O task -

i have project reactor chain includes blocking task (a network call, need wait response). i'd run multiple blocking tasks concurrently.

it seems either parallelflux or flatmap() used, bare-bone examples:

flux.just(1)     .repeat(10)     .parallel(3)     .runon(schedulers.elastic())     .doonnext(i -> blockingtask())     .sequential()     .subscribe() 

or

flux.just(1)     .repeat(10)     .flatmap(i -> mono.fromcallable(() -> {blockingtask(); return i;}).subscribeon(schedulers.elastic()), 3)     .subscribe(); 

what merits of 2 techniques? 1 preferred on other? there alternatives?

parallel tailored parallelization of tasks performance purposes, , dispatching of work between "rails" or "groups", each of own execution context scheduler pass runon. in short, put cpu cores work if cpu intensive work. you're doing i/o bound work...

so in case, flatmap better candidate. use of flatmap parallelization more orchestration.

these pretty 2 alternatives, if don't count different flavor of flatmap flatmapsequential (concatmap doesn't allow parallelization).