You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hbase.apache.org by Stefano Manca <st...@gmail.com> on 2018/04/10 13:33:09 UTC
HBase Developer - Questions
Dear,
My name is Stefano, I am a young software developer.
In these months I am using HBase as database in a Hadoop cluster, to write
data from Apache Spark with scala language.
May I ask you any question for specific information?
*Q1*: there is a way to insert/update a specific cell as latest field into
the same record?
Example:
ROW COLUMN+CELL
row1 column=CF1:QUAL1, timestamp=1523350296746, value=0.0
row1 column=CF1:QUAL2, timestamp=1523350296746, value=test
row1 column=CF1:QUAL3, timestamp=1523350296746, value=2700
row1 *column=CF2:QUAL4, timestamp=1523350296746, value=01218*
row1 column=CF2:QUAL5, timestamp=1523350296746, value=example
I would like to be sure that the value of the key QUAL4 will be insert as
last field, respect to the other values of the same rowkey (row1).
*Q2*: What is the best way to write in bulk with Spark 2.2? Is the
HBaseContext class available to use the bulkload with Hfiles?
Thank you very much in advance.
Best regards,
*__________________________________________________Stefano Manca*
*Phone: +39 349216059*
*E-mail: ste.manca23@gmail.com <st...@gmail.com>*
Re: HBase Developer - Questions
Posted by Josh Elser <el...@apache.org>.
(-to dev, +bcc dev, +to user)
Hi Stefano,
Moving your question over to the user@ mailing list as it's not so much
about development of HBase, instead development when using HBase.
Q1: what do you mean by the "latest field"? Are you talking about the
latest version of a Cell for a column (which is row + cf/cq)?
When you write a new cell without explicitly setting a timestamp, it
will use the current time from that RegionServer which is effectively
making it the "latest".
Q2: Not sure here :)
On 4/10/18 9:33 AM, Stefano Manca wrote:
> Dear,
>
> My name is Stefano, I am a young software developer.
> In these months I am using HBase as database in a Hadoop cluster, to write
> data from Apache Spark with scala language.
> May I ask you any question for specific information?
>
> *Q1*: there is a way to insert/update a specific cell as latest field into
> the same record?
>
> Example:
> ROW COLUMN+CELL
> row1 column=CF1:QUAL1, timestamp=1523350296746, value=0.0
> row1 column=CF1:QUAL2, timestamp=1523350296746, value=test
> row1 column=CF1:QUAL3, timestamp=1523350296746, value=2700
> row1 *column=CF2:QUAL4, timestamp=1523350296746, value=01218*
> row1 column=CF2:QUAL5, timestamp=1523350296746, value=example
>
> I would like to be sure that the value of the key QUAL4 will be insert as
> last field, respect to the other values of the same rowkey (row1).
>
> *Q2*: What is the best way to write in bulk with Spark 2.2? Is the
> HBaseContext class available to use the bulkload with Hfiles?
>
>
> Thank you very much in advance.
>
> Best regards,
>
>
>
> *__________________________________________________Stefano Manca*
>
> *Phone: +39 349216059*
> *E-mail: ste.manca23@gmail.com <st...@gmail.com>*
>
Re: HBase Developer - Questions
Posted by Josh Elser <el...@apache.org>.
(-to dev, +bcc dev, +to user)
Hi Stefano,
Moving your question over to the user@ mailing list as it's not so much
about development of HBase, instead development when using HBase.
Q1: what do you mean by the "latest field"? Are you talking about the
latest version of a Cell for a column (which is row + cf/cq)?
When you write a new cell without explicitly setting a timestamp, it
will use the current time from that RegionServer which is effectively
making it the "latest".
Q2: Not sure here :)
On 4/10/18 9:33 AM, Stefano Manca wrote:
> Dear,
>
> My name is Stefano, I am a young software developer.
> In these months I am using HBase as database in a Hadoop cluster, to write
> data from Apache Spark with scala language.
> May I ask you any question for specific information?
>
> *Q1*: there is a way to insert/update a specific cell as latest field into
> the same record?
>
> Example:
> ROW COLUMN+CELL
> row1 column=CF1:QUAL1, timestamp=1523350296746, value=0.0
> row1 column=CF1:QUAL2, timestamp=1523350296746, value=test
> row1 column=CF1:QUAL3, timestamp=1523350296746, value=2700
> row1 *column=CF2:QUAL4, timestamp=1523350296746, value=01218*
> row1 column=CF2:QUAL5, timestamp=1523350296746, value=example
>
> I would like to be sure that the value of the key QUAL4 will be insert as
> last field, respect to the other values of the same rowkey (row1).
>
> *Q2*: What is the best way to write in bulk with Spark 2.2? Is the
> HBaseContext class available to use the bulkload with Hfiles?
>
>
> Thank you very much in advance.
>
> Best regards,
>
>
>
> *__________________________________________________Stefano Manca*
>
> *Phone: +39 349216059*
> *E-mail: ste.manca23@gmail.com <st...@gmail.com>*
>