Details
-
Improvement
-
Status: Closed
-
Minor
-
Resolution: Fixed
-
1.5
-
None
-
None
Description
The fetcher has an identity reducer but uses the configure method from the mapper, this means large filters are loaded in the reducer as well eating up lots of unnecessary heap space.
Attachments
Issue Links
- is fixed by
-
NUTCH-2375 Upgrade the code base from org.apache.hadoop.mapred to org.apache.hadoop.mapreduce
- Closed