Mailing-List: contact user-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hbase.apache.org
From: Alex Loffler <alex@loffler.org>
Content-Type: text/plain;
	charset=utf-8
Content-Transfer-Encoding: quoted-printable
Mime-Version: 1.0 (1.0)
Date: Mon, 18 Dec 2017 20:17:04 -0800
Subject: Re: HBase increment of a specific cell version
Message-Id: <2BECF98B-18CB-4403-BEFD-4692DE6C6864@loffler.org>
References: <F1CDA0FB-3B32-422D-87EC-9F45F2B241F3@loffler.org> <CADcMMgEDM=SXaejXZXf+dwrg-Y170DAaR0BZ52gy5YNW_ThkDQ@mail.gmail.com>
In-Reply-To: <CADcMMgEDM=SXaejXZXf+dwrg-Y170DAaR0BZ52gy5YNW_ThkDQ@mail.gmail.com>
To: user@hbase.apache.org
archived-at: Tue, 19 Dec 2017 04:17:19 -0000

Hi Stack,

Thanks for the response, I am trying to maintain an hourly count of messages=
 between two keys/entities: Sender->recipient E.g. a->b

There are multiple ways of modelling this, but one that seems to fit nicely i=
s:
Row key =3D a
Col =3D b
Timestamp/version=3D e.g hour-of-day or hour-of-epoch
Val =3D count of messages

This approach utilizes the three dimensions of rowkey, col & version nicely.=


I will never need to look messages up by recipient but will be frequently qu=
erying for all recipients contacted by a sender (ie. return the value(count)=
 for each column (recipient) for a specific rowkey (sender) during a particu=
lar timespan - ie. at version x)

Everything is in place for this to work except the ability to increment a sp=
ecific version of a cell per the above.

If I don=E2=80=99t keep count (increment) and just write a flag to represent=
 a message between the two, this scheme/approach scales really nicely with t=
he put version of addColumn

If there=E2=80=99s a better pattern/approach, I=E2=80=99d really appreciate a=
 pointer in the right direction

-Alex.

> On Dec 18, 2017, at 8:49 AM, Stack <stack@duboce.net> wrote:
>=20
> Hello Alex. We don't have such an ability. Can you say what the use case i=
s
> because I at least am having trouble understanding why you would want to d=
o
> such a thing.
>=20
> Thank you,
> S
>=20
>> On Wed, Dec 13, 2017 at 2:07 PM, Alex Loffler <alex@loffler.org> wrote:
>>=20
>> Hi Folks,
>>=20
>> I am using the HBase=E2=80=99s timestamp/version concept to track
>> aggregates/counts for time periods/spans.
>>=20
>> The put function allows me to update a specific version, ie.
>> put(rk).addColumn(cf, column, version, value)
>>=20
>> But I can=E2=80=99t find a way of incrementing a specific version ie.
>> increment(rk).addColumn(cf, column, version, value) doesn=E2=80=99t exist=
.
>>=20
>> I can only find increment(rk).addColumn(cf, column, value) which exhibits=

>> the default behaviour of taking the latest version of the cell,
>> incrementing it=E2=80=99s value and updating the timestamp/version with
>> current-timestamp-millis.
>>=20
>> What I=E2=80=99d really like is an increment to the value in the specifie=
d
>> cell/version without the version update.
>>=20
>> Am I missing something, is this not possible for some reason in not
>> getting, or would it be a good feature request?
>>=20
>> Thanks again for a fantastic platform!
>> -Alex.
>>=20
>>=20
>>=20
>>=20
>>=20