You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@whirr.apache.org by Light Reader <li...@gmail.com> on 2012/05/08 19:46:07 UTC
Broken pipe from hadoop browse
Hi,
Looks like I am past some of my older issues. when i do a simple hadoop fs
-ls on my new cluster I get the following error:
[root@ip-10-118-190-121 WCluster1]# ~/hadoop-1.0.1/bin/hadoop fs -ls /
12/05/08 17:33:03 WARN conf.Configuration: DEPRECATED: hadoop-site.xml
found in the classpath. Usage of hadoop-site.xml is deprecated. Instead use
core-site.xml, mapred-site.xml and hdfs-site.xml to override properties of
core-default.xml, mapred-default.xml and hdfs-default.xml respectively
Bad connection to FS. command aborted. exception: Call to
ec2-23-22-79-169.compute-1.amazonaws.com/10.122.25.228:8020 failed on local
exception: java.net.SocketException: Broken pipe
I am guessing this is more of a Hadoop issue but dunno which setting is
causing this to happen.
Thanks,
Arni
Re: Broken pipe from hadoop browse
Posted by Andrei Savu <sa...@gmail.com>.
On Tue, May 8, 2012 at 10:37 PM, Light Reader <li...@gmail.com> wrote:
> Quick question which I could not find in FAQ or guide. If I make config
> changes (under /usr/local/hadoop/conf) on a hadoop node (say master) how do
> i restart my hadoop services?
>
Unfortunately there is no easy way with 0.7.x. This is something we are
working on for 0.8.0 - too be released as soon.
You are more than welcome to contribute:
https://issues.apache.org/jira/browse/WHIRR
To learn how to restart daemons by hand take a look at
start_hadoop_daemon() in:
https://github.com/apache/whirr/blob/branch-0.7/services/hadoop/src/main/resources/functions/configure_hadoop.sh
-- Andrei
Re: Broken pipe from hadoop browse
Posted by Light Reader <li...@gmail.com>.
I am a newbie but analysis of large data sets is my use case. Trying to
figure out the best way to do it right now.
Quick question which I could not find in FAQ or guide. If I make config
changes (under /usr/local/hadoop/conf) on a hadoop node (say master) how do
i restart my hadoop services?
On Tue, May 8, 2012 at 3:21 PM, Andrei Savu <sa...@gmail.com> wrote:
> Awesome!
>
> How are you using the Hadoop clusters? Are you planning to add custom
> services to Whirr?
>
> -- Andrei
>
>
> On Tue, May 8, 2012 at 10:16 PM, Light Reader <li...@gmail.com>wrote:
>
>> Slapping forehead.
>>
>> All good now.Thanks!!
>>
>>
>> On Tue, May 8, 2012 at 3:13 PM, Andrei Savu <sa...@gmail.com>wrote:
>>
>>> I suspect a Hadoop version mismatch. You need to set
>>> whirr.hadoop.tarball.url=...hadoop-1.0.1.tar.gz
>>>
>>> Can you try to login to the namenode and run hadoop fs -ls / ? That
>>> should work.
>>>
>>> -- Andrei Savu
>>>
>>>
>>> On Tue, May 8, 2012 at 8:46 PM, Light Reader <li...@gmail.com>wrote:
>>>
>>>> Hi,
>>>> Looks like I am past some of my older issues. when i do a simple hadoop
>>>> fs -ls on my new cluster I get the following error:
>>>>
>>>> [root@ip-10-118-190-121 WCluster1]# ~/hadoop-1.0.1/bin/hadoop fs -ls /
>>>> 12/05/08 17:33:03 WARN conf.Configuration: DEPRECATED: hadoop-site.xml
>>>> found in the classpath. Usage of hadoop-site.xml is deprecated. Instead use
>>>> core-site.xml, mapred-site.xml and hdfs-site.xml to override properties of
>>>> core-default.xml, mapred-default.xml and hdfs-default.xml respectively
>>>> Bad connection to FS. command aborted. exception: Call to
>>>> ec2-23-22-79-169.compute-1.amazonaws.com/10.122.25.228:8020 failed on
>>>> local exception: java.net.SocketException: Broken pipe
>>>>
>>>> I am guessing this is more of a Hadoop issue but dunno which setting is
>>>> causing this to happen.
>>>>
>>>> Thanks,
>>>> Arni
>>>>
>>>
>>>
>>
>
Re: Broken pipe from hadoop browse
Posted by Andrei Savu <sa...@gmail.com>.
Awesome!
How are you using the Hadoop clusters? Are you planning to add custom
services to Whirr?
-- Andrei
On Tue, May 8, 2012 at 10:16 PM, Light Reader <li...@gmail.com> wrote:
> Slapping forehead.
>
> All good now.Thanks!!
>
>
> On Tue, May 8, 2012 at 3:13 PM, Andrei Savu <sa...@gmail.com> wrote:
>
>> I suspect a Hadoop version mismatch. You need to set
>> whirr.hadoop.tarball.url=...hadoop-1.0.1.tar.gz
>>
>> Can you try to login to the namenode and run hadoop fs -ls / ? That
>> should work.
>>
>> -- Andrei Savu
>>
>>
>> On Tue, May 8, 2012 at 8:46 PM, Light Reader <li...@gmail.com>wrote:
>>
>>> Hi,
>>> Looks like I am past some of my older issues. when i do a simple hadoop
>>> fs -ls on my new cluster I get the following error:
>>>
>>> [root@ip-10-118-190-121 WCluster1]# ~/hadoop-1.0.1/bin/hadoop fs -ls /
>>> 12/05/08 17:33:03 WARN conf.Configuration: DEPRECATED: hadoop-site.xml
>>> found in the classpath. Usage of hadoop-site.xml is deprecated. Instead use
>>> core-site.xml, mapred-site.xml and hdfs-site.xml to override properties of
>>> core-default.xml, mapred-default.xml and hdfs-default.xml respectively
>>> Bad connection to FS. command aborted. exception: Call to
>>> ec2-23-22-79-169.compute-1.amazonaws.com/10.122.25.228:8020 failed on
>>> local exception: java.net.SocketException: Broken pipe
>>>
>>> I am guessing this is more of a Hadoop issue but dunno which setting is
>>> causing this to happen.
>>>
>>> Thanks,
>>> Arni
>>>
>>
>>
>
Re: Broken pipe from hadoop browse
Posted by Light Reader <li...@gmail.com>.
Slapping forehead.
All good now.Thanks!!
On Tue, May 8, 2012 at 3:13 PM, Andrei Savu <sa...@gmail.com> wrote:
> I suspect a Hadoop version mismatch. You need to set
> whirr.hadoop.tarball.url=...hadoop-1.0.1.tar.gz
>
> Can you try to login to the namenode and run hadoop fs -ls / ? That should
> work.
>
> -- Andrei Savu
>
>
> On Tue, May 8, 2012 at 8:46 PM, Light Reader <li...@gmail.com>wrote:
>
>> Hi,
>> Looks like I am past some of my older issues. when i do a simple hadoop
>> fs -ls on my new cluster I get the following error:
>>
>> [root@ip-10-118-190-121 WCluster1]# ~/hadoop-1.0.1/bin/hadoop fs -ls /
>> 12/05/08 17:33:03 WARN conf.Configuration: DEPRECATED: hadoop-site.xml
>> found in the classpath. Usage of hadoop-site.xml is deprecated. Instead use
>> core-site.xml, mapred-site.xml and hdfs-site.xml to override properties of
>> core-default.xml, mapred-default.xml and hdfs-default.xml respectively
>> Bad connection to FS. command aborted. exception: Call to
>> ec2-23-22-79-169.compute-1.amazonaws.com/10.122.25.228:8020 failed on
>> local exception: java.net.SocketException: Broken pipe
>>
>> I am guessing this is more of a Hadoop issue but dunno which setting is
>> causing this to happen.
>>
>> Thanks,
>> Arni
>>
>
>
Re: Broken pipe from hadoop browse
Posted by Andrei Savu <sa...@gmail.com>.
I suspect a Hadoop version mismatch. You need to set
whirr.hadoop.tarball.url=...hadoop-1.0.1.tar.gz
Can you try to login to the namenode and run hadoop fs -ls / ? That should
work.
-- Andrei Savu
On Tue, May 8, 2012 at 8:46 PM, Light Reader <li...@gmail.com> wrote:
> Hi,
> Looks like I am past some of my older issues. when i do a simple hadoop fs
> -ls on my new cluster I get the following error:
>
> [root@ip-10-118-190-121 WCluster1]# ~/hadoop-1.0.1/bin/hadoop fs -ls /
> 12/05/08 17:33:03 WARN conf.Configuration: DEPRECATED: hadoop-site.xml
> found in the classpath. Usage of hadoop-site.xml is deprecated. Instead use
> core-site.xml, mapred-site.xml and hdfs-site.xml to override properties of
> core-default.xml, mapred-default.xml and hdfs-default.xml respectively
> Bad connection to FS. command aborted. exception: Call to
> ec2-23-22-79-169.compute-1.amazonaws.com/10.122.25.228:8020 failed on
> local exception: java.net.SocketException: Broken pipe
>
> I am guessing this is more of a Hadoop issue but dunno which setting is
> causing this to happen.
>
> Thanks,
> Arni
>