Mailing-List: contact user-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hadoop.apache.org
Received-SPF: pass (athena.apache.org: domain of harsh@cloudera.com designates
 209.85.217.176 as permitted sender)
MIME-Version: 1.0
In-Reply-To: 
 <CAFWufDN01c_b4T7Ljz7Tc+=4dQevSo+wK819RHWvD=WD2VmV5w@mail.gmail.com>
References: 
 <CAFWufDP=jqkuKt7GU26eoeEF77FfQnabDRuR3do8VA9D1TCTxQ@mail.gmail.com>
 <CAOcnVr2PE6+_1cTH_FOdRMmrpSfukgFmUXeXqnTLQhnChvujcg@mail.gmail.com>
 <CAFWufDN01c_b4T7Ljz7Tc+=4dQevSo+wK819RHWvD=WD2VmV5w@mail.gmail.com>
From: Harsh J <harsh@cloudera.com>
Date: Fri, 23 Nov 2012 02:20:54 +0530
Message-ID: 
 <CAOcnVr0-mr01rkHDfxR68bzeDFdKTuTb5U0SMEqBbwdsxH9Mxw@mail.gmail.com>
Subject: Re: FileNotFoundExcepion when getting files from DistributedCache
To: "<user@hadoop.apache.org>" <user@hadoop.apache.org>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable

DistributedCache files in tasks are located locally (not on HDFS), so
use the LocalFileSystem or java.io.File if you prefer that, to read
them from within tasks.

On Fri, Nov 23, 2012 at 2:16 AM, Barak Yaish <barak.yaish@gmail.com> wrote:
> Thanks for the quick response.
>
> I wanted to use DistributedCache to localized the files in interest to al=
l
> nodes, so which API should I use in order to be able to read all those
> files, regardless the node running the mapper?
>
>
> On Thu, Nov 22, 2012 at 10:38 PM, Harsh J <harsh@cloudera.com> wrote:
>>
>> You pointed that you use:
>>
>> FSDataInputStream fs =3D FileSystem.get( context.getConfiguration() ).op=
en(
>> path )
>>
>> Note that this (FileSystem.get) will return back a HDFS FileSystem by
>> default and your path is a local one. You can either use simple
>> java.io.File APIs or use
>> FileSystem.getLocal(context.getConfiguration()) [1] to get a local
>> filesystem handle that can look in file:/// FSes rather than hdfs://
>> paths.
>>
>> [1]
>> http://hadoop.apache.org/docs/current/api/org/apache/hadoop/fs/FileSyste=
m.html#getLocal(org.apache.hadoop.conf.Configuration)
>>
>> On Fri, Nov 23, 2012 at 2:04 AM, Barak Yaish <barak.yaish@gmail.com>
>> wrote:
>> > Hi,
>> >
>> > I=92ve 2 nodes cluster (v1.04), master and slave. On the master, in
>> > Tool.run()
>> > we add two files to the DistributedCache using addCacheFile(). Files d=
o
>> > exist in HDFS. In the Mapper.setup() we want to retrieve those files
>> > from
>> > the cache using FSDataInputStream fs =3D FileSystem.get(
>> > context.getConfiguration() ).open( path ). The problem is that for one
>> > file
>> > a FileNotFoundException is thrown, although the file exists on the sla=
ve
>> > node:
>> >
>> > attempt_201211211227_0020_m_000000_2: java.io.FileNotFoundException:
>> > File
>> > does not exist:
>> >
>> > /somedir/hdp.tmp.dir/mapred/local/taskTracker/distcache/-7769715304990=
780/master/tmp/analytics/1.csv
>> >
>> > ls =96l on the slave:
>> >
>> > [hduser@slave ~]$ ll
>> >
>> > /somedir/hdp.tmp.dir/mapred/local/taskTracker/distcache/-7769715304990=
780/master/tmp/
>> > analytics/1.csv
>> > -rwxr-xr-x 1 hduser hadoop 42701 Nov 22 10:18
>> >
>> > /somedir/hdp.tmp.dir/mapred/local/taskTracker/distcache/-7769715304990=
780/master/tmp/analytics/1.csv
>> > [hduser@slave ~]$
>> >
>> > My questions are:
>> >
>> > Shouldn't all files exist on all nodes?
>> > What should be done to fix that?
>> >
>> > Thanks.
>>
>>
>>
>> --
>> Harsh J
>
>


--=20
Harsh J