hadoop - running a custom shell script distributed shell apache yarn -
i have been going through apache hadoop yarn book hortonworks, wherein have explained 2 ways of running yarn task.
my intent run shell script ( compiles , runs various java , python scripts ) , runs set of these scripts/patches various folders. easy metaphor :- " unzipping 100 folders , logging 'ls' "
now want parallelize flow , such container runs 1-2 folders , ask 50 such containers.
how do using distributed shell ? have seen examples of ls / whoami / uptime / hostname not want. want run script takes / iterated on argument path , want run in distributed fashion on yarn. ?
Comments
Post a Comment