hadoop-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Vijay Thakorlal <vijayj...@hotmail.com>
Subject RE: Reg Too many fetch-failures Error
Date Fri, 01 Feb 2013 17:52:52 GMT
Hi Manoj,


As you may be aware this means the reduces are unable to fetch intermediate
data from TaskTrackers that ran map tasks - you can try:

* increasing tasktracker.http.threads so there are more threads to handle
fetch requests from reduces. 

* decreasing mapreduce.reduce.parallel.copies : so fewer copy / fetches are
performed in parallel


It could also be due to a temporary DNS issue.


See slide 26 of this presentation for potential causes for this message:


Not sure why you did not receive the problem before but was it the same data
or different data? Did you have other jobs running on your cluster?


Hope that helps





From: Manoj Babu [mailto:manoj444@gmail.com] 
Sent: 01 February 2013 15:09
To: user@hadoop.apache.org
Subject: Reg Too many fetch-failures Error


Hi All,


I am getting Too many fetch-failures exception.

What might be the reason for this exception, For same size of data i dint
face this error earlier and there is change in code.

How to avoid this?


Thanks in advance.




View raw message