hadoop-common-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Harsh J <ha...@cloudera.com>
Subject Re: Hadoop order of operations
Date Sat, 06 Aug 2011 03:50:16 GMT
Premal,

Didn't go through your entire thread, but the right order is: "map"
(N) -> "partition" (N) -> "combine" (0…N).

On Sat, Aug 6, 2011 at 4:04 AM, Premal <premal.j.shah@gmail.com> wrote:
>
> According to the attached image found on yahoo's hadoop tutorial, the order
> of operations is map > combine > partition which should be followed by
> reduce
>
> Here is my an example key emmited by the map operation
>
>    LongValueSum:geo_US|1311722400|E        1
>
> Assuming there are 100 keys of the same type, this should get combined as
>
>    geo_US|1311722400|E     100
>
> Then i'd like to partition the keys by the value before the first pipe(|)
> http://hadoop.apache.org/common/docs/r0.20.2/streaming.html#A+Useful+Partitioner+Class+%28secondary+sort%2C+the+-partitioner+org.apache.hadoop.mapred.lib.KeyFieldBasedPartitioner+option%29
>
>    geo_US
>
> Here's the streaming command
>
>    hadoop jar
> /usr/local/hadoop/contrib/streaming/hadoop-streaming-0.20.203.0.jar \
>    -D mapred.reduce.tasks=8 \
>    -D stream.num.map.output.key.fields=1 \
>    -D mapred.text.key.partitioner.options=-k1,1 \
>    -D stream.map.output.field.separator=\| \
>    -file mapper.py \
>    -mapper mapper.py \
>    -file reducer.py \
>    -reducer reducer.py \
>    -combiner org.apache.hadoop.mapred.lib.aggregate.ValueAggregatorReducer
> \
>    -partitioner org.apache.hadoop.mapred.lib.KeyFieldBasedPartitioner \
>    -input input_file \
>    -output output_path
>
>
> This is the error I get
>    java.lang.NumberFormatException: For input string: "1311722400|E    1"
>        at
> java.lang.NumberFormatException.forInputString(NumberFormatException.java:48)
>        at java.lang.Long.parseLong(Long.java:419)
>        at java.lang.Long.parseLong(Long.java:468)
>        at
> org.apache.hadoop.mapred.lib.aggregate.LongValueSum.addNextValue(LongValueSum.java:48)
>        at
> org.apache.hadoop.mapred.lib.aggregate.ValueAggregatorReducer.reduce(ValueAggregatorReducer.java:59)
>        at
> org.apache.hadoop.mapred.lib.aggregate.ValueAggregatorReducer.reduce(ValueAggregatorReducer.java:35)
>        at org.apache.hadoop.mapred.Task$OldCombinerRunner.combine(Task.java:1349)
>        at
> org.apache.hadoop.mapred.MapTask$MapOutputBuffer.sortAndSpill(MapTask.java:1435)
>        at
> org.apache.hadoop.mapred.MapTask$MapOutputBuffer.flush(MapTask.java:1297)
>        at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:436)
>        at org.apache.hadoop.mapred.MapTask.run(MapTask.java:371)
>        at org.apache.hadoop.mapred.Child$4.run(Child.java:259)
>        at java.security.AccessController.doPrivileged(Native Method)
>        at javax.security.auth.Subject.doAs(Subject.java:396)
>        at
> org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1059)
>        at org.apache.hadoop.mapred.Child.main(Child.java:253)
>
> I looks like the partitioner is running before the combiner. Any thoughts?
> --
> View this message in context: http://old.nabble.com/Hadoop-order-of-operations-tp32205781p32205781.html
> Sent from the Hadoop core-user mailing list archive at Nabble.com.
>
>



-- 
Harsh J

Mime
View raw message