hadoop-mapreduce-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Hari Sreekumar <hsreeku...@clickable.com>
Subject Re: setup() and cleanup() methods
Date Tue, 14 Dec 2010 04:25:10 GMT
ah yeah..  was looking at the wrong log directory

Thanks,
Hari

On Tue, Dec 14, 2010 at 12:37 AM, Ken Goodhope <kengoodhope@gmail.com>wrote:

> If you look in your logs directory, you will see a subdir called userlogs.
> Inside there you will see a directory for every task.  Those directories
> contain stdout, stderr, and system.
>
>
> On Mon, Dec 13, 2010 at 9:26 AM, Hari Sreekumar <hsreekumar@clickable.com>wrote:
>
>> Hi,
>>     I am using the setup() and cleanup() methods as follows:
>>
>>        @Override
>>         protected void setup(Context context) throws IOException,
>> InterruptedException {
>>             HBaseConfiguration conf = new
>> HBaseConfiguration(context.getConfiguration());
>>             System.out.println("Opening table NwKeyword");
>>             keywordTable = new HTable(conf, "NwKeyword");
>>             System.out.println("Table opened");
>>         }
>>
>>         @Override
>>         protected void cleanup(Context context) throws IOException,
>> InterruptedException {
>>             keywordTable.close();
>>             System.out.println("Table closed");
>>         }
>>
>>         @Override
>>         public void reduce(Text key, Iterable<Text> values, Context
>> context) throws IOException, InterruptedException {
>>             ......
>>
>> I know that the setup method is being called because I can use the
>> keywordTable in my reduce function, which I cannot do if I don't initialize
>> the table in setup(). But the System.out.println statement is not getting
>> printed. I checked the sysout as well as the job logs. Where does/should it
>> get printed?
>>
>> Thanks,
>> hari
>>
>
>

Mime
View raw message