Hadoop 0.20: "job setup" task -


i not sure if has been fixed newer releases of hadoop, i'm locked running hadoop 0.20 (legacy code).

here's issue: when launch hadoop job, there "job setup" task needs run first. seems me hadoop randomly picks task either map task or reduce task.

we have more capacity map tasks configured reduce tasks, whenever unlucky , have reduce startup task, takes forever long job start running. ideas how overcome this?

hadoop job first complete mapper task. once mapper task completed go across network , shuffling , sorting , after reducer task start processing. guess there possibly other delay.


Comments

Popular posts from this blog

shopping cart - Page redirect not working PHP -

php - How to modify a menu to show sub-menus -

python - Installing PyDev in eclipse is failed -