Stack,
Yes, I would think so. For every application with a large data
set, I have 50 smaller OLTP oriented ones. More often we want to scale
concurrent users, throughput, etc. rather than the data set.
Bottom line, HBase will be more relevant and applicable for these
scenarios.
Thanks again,
Guy
-----Original Message-----
From: saint....@gmail.com [mailto:saint....@gmail.com] On Behalf Of
stack
Sent: Thursday, September 24, 2009 4:11 PM
To: hbase-user@hadoop.apache.org
Subject: Re: Hbase and linear scaling with small write intensive
clusters
On Thu, Sep 24, 2009 at 3:58 PM, Molinari, Guy
<guy.molin...@disney.com>wrote:
>
> There are still significant pauses when splits occur. I suspect that
if
> this could be improved then Hbase would be more relevant for
applications
> managing smaller data sets where higher scalability requirements
exist.
>
>
We're working on it... http://issues.apache.org/jira/browse/HBASE-1506.
Should we up the priority on this issue?
Thanks for reporting back.
St.Ack
