Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
MapReduce, mail # user - Write an object into hadoop hdfs issue


Copy link to this message
-
Re: Write an object into hadoop hdfs issue
Chris Mawata 2013-12-30, 13:35
Not unique to hdfs. The same thing would happen on your local file system
or anywhere and any way you store the state of the object outside of the
JVM.  That is why singletons should not be serializable.
Chris
On Dec 30, 2013 5:46 AM, "unmesha sreeveni" <[EMAIL PROTECTED]> wrote:

> I am trying to write an object into hdfs .
>
>
> public static Split *currentsplit *= new Split();
>
> *Split currentsplit = new Split();*
> *Path p = new Path("C45/mysavedobject");*
> *ObjectOutputStream oos = new ObjectOutputStream(fs.create(p));*
> *oos.writeObject(currentsplit);*
> *oos.close();*
>
> But I am not able to get the exact object value.
>
> *Path p = new Path("C45/mysavedobject");*
> *Split curntsplit = null;*
> *ObjectInputStream ois = new ObjectInputStream(fs.open(p));*
> *try {*
> *    curntsplit = (Split) ois.readObject();*
> *} catch (ClassNotFoundException e) {*
> *    // TODO Auto-generated catch block*
> *    e.printStackTrace();*
> *}*
> * ois.close();*
>
> Showing duplicate objects value.
>
> In driver i am getting :
>
> *objjjjjjj in driver pck.Split@14da173c*
>
> After writing the object into file and reading back within driver itself I
> am getting an alternate object value.
>
> *objjjjjj in mysavedobject pck.Split@62eade0*
>
>
> I want the driver object along my Mapper code.
>
> Why is it so?
>
> --
> *Thanks & Regards*
>
> Unmesha Sreeveni U.B
>
>