Mariano,
Pig 0.12.0 does work with Hadoop-2.2.0, but you need to recompile Pig first.
In your $PIG_HOME, run the following command to rebuild Pig:
`ant clean jar-withouthadoop -Dhadoopversion=23`
Then, try re-running your script.
On 1/9/14, 5:13 PM, Mariano Kamp wrote:
Hi,
I am trying to run the first pig sample from the Hadoop - The definitive guide
- book.
Unfortunately that doesn't work for me.
I downloaded 0.12.0 and got the impression it should work with Hadoop 2.2.
http://pig.apache.org/releases.html#14+October%2C+2013%3A+release+0.12.0+available
14 October, 2013: release 0.12.0 available
This release include several new features such as ASSERT operator, Streaming UDF,
new AvroStorage, IN/CASE >operator, BigInteger/BigDecimal data type, support
for Windows.
Note
This release works with Hadoop 0.20.X, 1.X, 0.23.X and 2.X
I use Hadoop 2.x.
snow:bin mkamp$ which hadoop
/Users/mkamp/hadoop-2.2.0/bin//hadoop
snow:bin mkamp$ echo $HADOOP_HOME
/Users/mkamp/hadoop-2.2.0
But no matter if HADOOP_HOME is set or not I get a couple of errors and it
doesn't work if I run the script:
records = LOAD 'micro-tab/sample.txt'
AS (year:chararray, temperature:int, quality:int);
DUMP records;
All hell breaks lose and there is a lot of output, but most seems meaningless,
warnings about settings that are deprecated in Hadoop, but still delivered by
default this way.
Hard to say what is relevant. Here are some excerpts, full output attached as
file.
From the logfile:
Unexpected System Error Occured: java.lang.IncompatibleClassChangeError: Found
interface >org.apache.hadoop.mapreduce.JobContext, but class was expected
at
>org.apache.pig.backend.hadoop.executionengine.mapReduceLayer.PigOutputFormat.setupUdfEnvAndStores(PigO>utputFormat.java:225)
ERROR 1066: Unable to open iterator for alias records
From the console:
2014-01-09 22:24:45,976 [main] WARN org.apache.pig.backend.hadoop20.PigJobControl
- falling back to default >JobControl (not using hadoop 0.20 ?)
java.lang.NoSuchFieldException: runnerState
at java.lang.Class.getDeclaredField(Class.java:1918)
But as a little googling indicated, this is business as usual?
2014-01-09 22:24:49,228 [JobControl] ERROR
org.apache.hadoop.mapreduce.lib.jobcontrol.JobControl - Error >while trying to
run jobs.
java.lang.IncompatibleClassChangeError: Found interface
org.apache.hadoop.mapreduce.JobContext, but class >was expected
Input(s):
Failed to read data from "hdfs://localhost/user/mkamp/micro-tab/sample.txt"
That last one looks interesting. Maybe I am using it wrong and the reported
errors are not related? I wanted to read from the local file system.
So I also changed the script to read from hdfs, but that didn't change the
error.
Any ideas where to go from here?
Is it possible to run the latest Hadoop binary download and the latest Pig
binary download together?
