[MAPREDUCE-5160] Aggregatewordcount and aggregatewordhist in hadoop-1 examples can not find their inner classes when running on Yarn

• Type: Sub-task
• Status: Closed
• Priority: Major
• Resolution: Duplicate
• Fix Version/s: 2.1.0-beta
• Assignee: Zhijie Shen
• Reporter: Zhijie Shen
• Created: 17/Apr/13 18:17
• Updated: 27/Aug/13 22:22
• Resolved: 08/May/13 19:05

Description

Aggregatewordcount and Aggregatewordhist of hadoop-1 cannot run on hadoop-2 due to org.apache.hadoop.mapred.lib.aggregate.ValueAggregatorJob.createValueAggregatorJob(String args[], Class<?> caller) is not available on hadoop-2 (see MAPREDUCE-5159).

After I changed Aggregatewordcount and Aggregatewordhist to use createValueAggregatorJob(String args[], Class<? extends ValueAggregatorDescriptor>[] descriptors), which is available on hadoop-2, the two examples could be accepted and run on Yarn.

However, the two examples still failed, because their inner classes, WordCountPlugInClass and AggregateWordHistogramPlugin, cannot be found in runtime, respectively. Both the plugin classes extend org.apache.hadoop.mapred.lib.aggregate.ValueAggregatorBaseDescriptor.

Issue Links

• is blocked by: MAPREDUCE-5159 Aggregatewordcount and aggregatewordhist in hadoop-1 examples are not binary compatible with hadoop-2 mapred.lib.aggregate
• is duplicated by: MAPREDUCE-5159 Aggregatewordcount and aggregatewordhist in hadoop-1 examples are not binary compatible with hadoop-2 mapred.lib.aggregate
• relates to: MAPREDUCE-5159 Aggregatewordcount and aggregatewordhist in hadoop-1 examples are not binary compatible with hadoop-2 mapred.lib.aggregate