Apache Pig Process Killed After Putty Session timeout

Question

i launched a pig process over a folder of 71 GB in putty . after 45 minutes,the putty session gives "software connection abort" and then on job tracker the pig job status is "killed".

I tried increasing the session timeout of putty(Seconds between keppalives:1800),but no help.

Please suggest if i am going in right direction and that my concern regarding session timeout affecting pig job is correct or not.

also is there any other way to launch pig job apart from putty. I have cloudera on a 8 node cluster.

Answer 1

You should use the Shell command "screen". So you could now even shutdown your computer then reconnect to the distant computer and use screen -r in order to find the id of the screen in which you started your pig :)