Re: CsvBulkLoadTool not populating Actual Table & Local Index Table when '-it' option specified

2016-07-05 Thread Vamsi Krishna 
Thanks Rajeshbabu.

On Tue, Jul 5, 2016 at 5:59 AM rajeshb...@apache.org <
chrajeshbab...@gmail.com> wrote:

> Hi Vamsi,
>
> There is a bug with local indexes in 4.4.0 which is fixed in 4.7.0
> https://issues.apache.org/jira/browse/PHOENIX-2334
>
> Thanks,
> Rajeshbabu.
>
> On Tue, Jul 5, 2016 at 6:21 PM, Vamsi Krishna 
> wrote:
>
>> Team,
>>
>> I'm working on HDP 2.3.2 (Phoenix 4.4.0, HBase 1.1.2).
>> When I use '-it' option of CsvBulkLoadTool neither Acutal Table nor Local
>> Index Table is loaded.
>> *Command:*
>> *HADOOP_CLASSPATH=/usr/hdp/current/hbase-master/lib/hbase-protocol.jar:/etc/hbase/conf
>> yarn jar /usr/hdp/current/phoenix-client/phoenix-client.jar
>> org.apache.phoenix.mapreduce.CsvBulkLoadTool
>> -Dmapreduce.job.queuename=$QUEUE_NAME -s VAMSI -t TABLE_A -c COL1,COL2,COL3
>> -it IDX_TABLE_A_COL2 -i test/test_data.csv -d ',' -z $ZOOKEEPER_QUORUM*
>>
>> When I use the same command without specifying the '-it' option it
>> populates the Actual Table but not Local Index Table (Which is as expected).
>> *Command:*
>> *HADOOP_CLASSPATH=/usr/hdp/current/hbase-master/lib/hbase-protocol.jar:/etc/hbase/conf
>> yarn jar /usr/hdp/current/phoenix-client/phoenix-client.jar
>> org.apache.phoenix.mapreduce.CsvBulkLoadTool
>> -Dmapreduce.job.queuename=$QUEUE_NAME -s VAMSI -t TABLE_A -c COL1,COL2,COL3
>> -i test/test_data.csv -d ',' -z $ZOOKEEPER_QUORUM*
>>
>> Could someone please help me if you see anything wrong with what I'm
>> doing?
>>
>> Here is how I'm setting up my table:
>> CREATE TABLE IF NOT EXISTS VAMSI.TABLE_A (COL1 VARCHAR(36) , COL2
>> VARCHAR(36) , COL3 VARCHAR(36) CONSTRAINT TABLE_A_PK PRIMARY KEY (COL1))
>> COMPRESSION='SNAPPY', SALT_BUCKETS=5;
>> CREATE LOCAL INDEX IF NOT EXISTS IDX_TABLE_A_COL2 ON VAMSI.TABLE_A (COL2);
>> upsert into vamsi.table_a values ('abc123','abc','123');
>> upsert into vamsi.table_a values ('def456','def','456');
>>
>> test_data.csv contains 2 records:
>> ghi789,ghi,789
>> jkl012,jkl,012
>>
>> Thanks,
>> Vamsi Attluri
>> --
>> Vamsi Attluri
>>
>
> --
Vamsi Attluri

Re: CsvBulkLoadTool not populating Actual Table & Local Index Table when '-it' option specified

2016-07-05 Thread rajeshb...@apache.org 
Hi Vamsi,

There is a bug with local indexes in 4.4.0 which is fixed in 4.7.0
https://issues.apache.org/jira/browse/PHOENIX-2334

Thanks,
Rajeshbabu.

On Tue, Jul 5, 2016 at 6:21 PM, Vamsi Krishna 
wrote:

> Team,
>
> I'm working on HDP 2.3.2 (Phoenix 4.4.0, HBase 1.1.2).
> When I use '-it' option of CsvBulkLoadTool neither Acutal Table nor Local
> Index Table is loaded.
> *Command:*
> *HADOOP_CLASSPATH=/usr/hdp/current/hbase-master/lib/hbase-protocol.jar:/etc/hbase/conf
> yarn jar /usr/hdp/current/phoenix-client/phoenix-client.jar
> org.apache.phoenix.mapreduce.CsvBulkLoadTool
> -Dmapreduce.job.queuename=$QUEUE_NAME -s VAMSI -t TABLE_A -c COL1,COL2,COL3
> -it IDX_TABLE_A_COL2 -i test/test_data.csv -d ',' -z $ZOOKEEPER_QUORUM*
>
> When I use the same command without specifying the '-it' option it
> populates the Actual Table but not Local Index Table (Which is as expected).
> *Command:*
> *HADOOP_CLASSPATH=/usr/hdp/current/hbase-master/lib/hbase-protocol.jar:/etc/hbase/conf
> yarn jar /usr/hdp/current/phoenix-client/phoenix-client.jar
> org.apache.phoenix.mapreduce.CsvBulkLoadTool
> -Dmapreduce.job.queuename=$QUEUE_NAME -s VAMSI -t TABLE_A -c COL1,COL2,COL3
> -i test/test_data.csv -d ',' -z $ZOOKEEPER_QUORUM*
>
> Could someone please help me if you see anything wrong with what I'm doing?
>
> Here is how I'm setting up my table:
> CREATE TABLE IF NOT EXISTS VAMSI.TABLE_A (COL1 VARCHAR(36) , COL2
> VARCHAR(36) , COL3 VARCHAR(36) CONSTRAINT TABLE_A_PK PRIMARY KEY (COL1))
> COMPRESSION='SNAPPY', SALT_BUCKETS=5;
> CREATE LOCAL INDEX IF NOT EXISTS IDX_TABLE_A_COL2 ON VAMSI.TABLE_A (COL2);
> upsert into vamsi.table_a values ('abc123','abc','123');
> upsert into vamsi.table_a values ('def456','def','456');
>
> test_data.csv contains 2 records:
> ghi789,ghi,789
> jkl012,jkl,012
>
> Thanks,
> Vamsi Attluri
> --
> Vamsi Attluri
>

CsvBulkLoadTool not populating Actual Table & Local Index Table when '-it' option specified

2016-07-05 Thread Vamsi Krishna 
Team,

I'm working on HDP 2.3.2 (Phoenix 4.4.0, HBase 1.1.2).
When I use '-it' option of CsvBulkLoadTool neither Acutal Table nor Local
Index Table is loaded.
*Command:*
*HADOOP_CLASSPATH=/usr/hdp/current/hbase-master/lib/hbase-protocol.jar:/etc/hbase/conf
yarn jar /usr/hdp/current/phoenix-client/phoenix-client.jar
org.apache.phoenix.mapreduce.CsvBulkLoadTool
-Dmapreduce.job.queuename=$QUEUE_NAME -s VAMSI -t TABLE_A -c COL1,COL2,COL3
-it IDX_TABLE_A_COL2 -i test/test_data.csv -d ',' -z $ZOOKEEPER_QUORUM*

When I use the same command without specifying the '-it' option it
populates the Actual Table but not Local Index Table (Which is as expected).
*Command:*
*HADOOP_CLASSPATH=/usr/hdp/current/hbase-master/lib/hbase-protocol.jar:/etc/hbase/conf
yarn jar /usr/hdp/current/phoenix-client/phoenix-client.jar
org.apache.phoenix.mapreduce.CsvBulkLoadTool
-Dmapreduce.job.queuename=$QUEUE_NAME -s VAMSI -t TABLE_A -c COL1,COL2,COL3
-i test/test_data.csv -d ',' -z $ZOOKEEPER_QUORUM*

Could someone please help me if you see anything wrong with what I'm doing?

Here is how I'm setting up my table:
CREATE TABLE IF NOT EXISTS VAMSI.TABLE_A (COL1 VARCHAR(36) , COL2
VARCHAR(36) , COL3 VARCHAR(36) CONSTRAINT TABLE_A_PK PRIMARY KEY (COL1))
COMPRESSION='SNAPPY', SALT_BUCKETS=5;
CREATE LOCAL INDEX IF NOT EXISTS IDX_TABLE_A_COL2 ON VAMSI.TABLE_A (COL2);
upsert into vamsi.table_a values ('abc123','abc','123');
upsert into vamsi.table_a values ('def456','def','456');

test_data.csv contains 2 records:
ghi789,ghi,789
jkl012,jkl,012

Thanks,
Vamsi Attluri
-- 
Vamsi Attluri