You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@hbase.apache.org by William Theisinger <wt...@attinteractive.com> on 2011/02/15 18:34:46 UTC

Hbase Hardware needs

Hi

Thinking of implementing Hbase on top of our data processing pipeline (Hadoop) and was curious if there are some guidelines to memory needs, number of region servers recommended based on the size of the grid/volume of data etc.

Any thoughts here would be appreciated.... I would interested in some calculations that would provide us a good baseline to spec our hardware needs.


Thanks!

Bill

Re: Hbase Hardware needs

Posted by Stack <st...@duboce.net>.
Hey William:

Have you checked the mailing list archives.  This topic has come up in
various guises in our past.  Here's one such thread:
http://search-hadoop.com/m/4DQfl2TGBb22/hardware&subj=Hadoop+HBase+hardware+requirement

Hopefully this helps some.
St.Ack



On Tue, Feb 15, 2011 at 9:34 AM, William Theisinger
<wt...@attinteractive.com> wrote:
> Hi
>
> Thinking of implementing Hbase on top of our data processing pipeline (Hadoop) and was curious if there are some guidelines to memory needs, number of region servers recommended based on the size of the grid/volume of data etc.
>
> Any thoughts here would be appreciated.... I would interested in some calculations that would provide us a good baseline to spec our hardware needs.
>
>
> Thanks!
>
> Bill

Re: Hbase Hardware needs

Posted by Jean-Daniel Cryans <jd...@apache.org>.
Start with this:
http://www.cloudera.com/blog/2010/08/hadoophbase-capacity-planning/

Then regarding the number of servers... it's really hard to tell,
you'd have to test with a handful of machines first and see how they
perform under your type of load. Scaling is then as easy as adding the
new machines to the network, copying the bins and starting the
daemons.

J-D

On Tue, Feb 15, 2011 at 9:34 AM, William Theisinger
<wt...@attinteractive.com> wrote:
> Hi
>
> Thinking of implementing Hbase on top of our data processing pipeline (Hadoop) and was curious if there are some guidelines to memory needs, number of region servers recommended based on the size of the grid/volume of data etc.
>
> Any thoughts here would be appreciated.... I would interested in some calculations that would provide us a good baseline to spec our hardware needs.
>
>
> Thanks!
>
> Bill