Hi, Harsh. Sorry, I failed to write to user@hadoop.apache.org
(http://mail-archives.apache.org/mod_mbox/hadoop-user/201301.mbox/%3CCAOcnVr2A%2BmDQ%2BVY3iovxuP13VTXk2Wog7ui_o6x%3DW0et5igUSg%40mail.gmail.com%3E)
The logic of combiner is the same as reducer's one - summation of some
statistical para
Can you add in what exactly is your combiner logic performing?
On Wed, Jan 23, 2013 at 3:09 PM, s2323 wrote:
> Hi!
>
> When I run job with this options:
> -Dmapred.map.child.java.opts=-Xmx2048M
> -Dio.sort.mb=1424
> -Dio.sort.record.percent=0.08
> all tasks fails on combiner step with:
>
Hi!
When I run job with this options:
-Dmapred.map.child.java.opts=-Xmx2048M
-Dio.sort.mb=1424
-Dio.sort.record.percent=0.08
all tasks fails on combiner step with:
...
2013-01-23 12:20:28,143 INFO org.apache.hadoop.mapred.MapTask: io.sort.mb = 1424
2013-01-23 12:23:03,772 INFO org.apache.ha