hadoop - Calling htable.get(rowID) in reduce task returns NULL for random rows -
hadoop - Calling htable.get(rowID) in reduce task returns NULL for random rows -
in hadoop mapper task, create [key,value] = [sortkey,rowid] scanning hbase table in cut down task, when result=htable.get(rowid), returns null, randomly, diffrent rowids. on querying in hbase shell, able fetch same rowid. input hbase table row count of 1000 (mapper able fetch these records), around 100 records returns htable.get(rowid) null (in reducer), these 100 records available in table.
any thought why 'get' on htable homecoming null randomly? caching or memory tweaking solve this?
am using hadoop-1.0.3, hbase-0.92.1
i using hashmap store text rowids, , cos of that, referring same location within loop. since doing updates/deletes within loop, row ids not found. either changing hashmap string, or creating new text object , adding hashmap solved issue.
hadoop mapreduce hbase
Comments
Post a Comment