hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stack <st...@duboce.net>
Subject Re: efficient storage of timeseries
Date Thu, 04 Mar 2010 06:15:00 GMT
On Wed, Mar 3, 2010 at 6:39 PM, Saptarshi Guha <saptarshi.guha@gmail.com> wrote:
> But if I specify which column families I want, will it still load the
> entire row?


No.  It will just load the content of the specified column families.


Suppose row key R has col. families A,B, C, D and A itself
> has 100 column labels, B has 150 and C 5, D 5
> I request row key R, but specify column families C and D.
> Will it still load A,B?

No.  Just C and D.

St.Ack


> Regards
> Saptarshi
>
>
> th
>> taller tables and scans than going with thick rows - this is because
>> the current way RPCs work we must hold an entire row in memory when
>> streaming it back to the client.  If your data eventually grows
>> extremely large, you could run into OOME.  Thus taller tables with
>

Mime
View raw message