I just tried it on EMR and it worked as expected.Is there a verbose mode
which can help debug what's going on with our EC2-based configuration?
Ranjan
On Dec 16, 2011, at 10:55 AM, Mark Grover wrote:
> Hi Ranjan,
> I agree with Igor. I consider it a good practice to point the location to th
Thanks Alex, I cleaned up my hdfs file system, and now the query works now.
-Andrew
On Fri, Dec 16, 2011 at 1:29 AM, alo alt wrote:
> Hi,
>
> looks like the user who uses the statement has not the correct rights.
> org.apache.hadoop.fs.permission.FsPermission$2.())'
>
> - Alex
>
> On Fri, Dec 1
Hi Ranjan,
I agree with Igor. I consider it a good practice to point the location to the
directory containing the file instead of the file itself.
It's probably some config option that's causing you this problem (You can try
using s3 instead of s3n in your paths). It's definitely possible to fin
It sounds like you are using Apache Hive. I don’t think it supports querying
data on S3, does it?
https://groups.google.com/a/cloudera.org/group/cdh-user/browse_thread/thread/34693af0fa6a9101
On Dec 16, 2011, at 10:43 AM, Ranjan Bagchi wrote:
> Following up with more information:
>
> * The ha
Following up with more information:
* The hadoop cluster is on EC2, not EMR, but I'll try bringing it up on EMR.
* I don't see a job conf on the tracker page -- I'm semi-suspicious it never
makes it that far.
* Here's the extended explain plan: it doesn't look glaringly wrong.
Totally appreciate
Ranjith
You can definitely change the number of buckets in a hive table even after
its creation. You need to issue an alter table command that contains the
CLUSTERED BY and/or SORTED BY clauses used by your table. For example if I have
a table whose DDL looks like this
CREATE EXTERNAL TABLE
Hi,
looks like the user who uses the statement has not the correct rights.
org.apache.hadoop.fs.permission.FsPermission$2.())'
- Alex
On Fri, Dec 16, 2011 at 8:59 AM, jingjung Ng wrote:
> Hi,
>
> I have simple hive select count(*) query,which results in the following
> exception. I am using Clo
Hi,
I have simple hive select count(*) query,which results in the following
exception. I am using Cloudera cdh3u1 ( hadoop/hbase/hive). However I am
able to do "select * from t1" from hive CLI.
Here is output after running "select count(*) from t1".
hive> select count(*) from t1;
Total MapReduce