I think I can speak for all of the HBase devs that in our opinion this vendor "benchmark" was designed by hypertable to demonstrate a specific feature of their system -- autotuning -- in such a way that HBase was, obviously, not tuned. Nobody from the HBase project was consulted on the results or to do such tuning, as is common courtesy when running a competitive benchmark, if the goal is a fair test. Furthermore the "benchmark" code was not a community accepted benchmark such as YCSB. 

I do not think the results are valid beyond being some vendor FUD and do not warrant much comment beyond this.

Best regards,

    - Andy

Problems worthy of attack prove their worth by hitting back. - Piet Hein (via Tom White)


--- On Wed, 5/25/11, edward choi <mp2893@gmail.com> wrote:

> From: edward choi <mp2893@gmail.com>
> Subject: hbase and hypertable comparison
> To: user@hbase.apache.org, common-user@hadoop.apache.org
> Date: Wednesday, May 25, 2011, 12:47 AM
> I'm planning to use a NoSQL
> distributed database.
> I did some searching and came across a lot of database
> systems such as
> MongoDB, CouchDB, Hbase, Cassandra, Hypertable, etc.
> 
> Since what I'll be doing is frequently reading a varying
> amount of data, and
> less frequently writing a massive amount of data,
> I thought Hbase, or Hypertable is the way to go.
> 
> I did some internet and found some performance comparison
> between HBase and
> HyperTable.
> Obviously HT dominated Hbase in every aspect (random
> read/write and a couple
> of more)
> 
> But the comparison was made with Hbase 0.20.4, and Hbase
> had much
> improvements since the current version is 0.90.3.
> 
> I am curious if the performance gap is still large between
> Hbase and HT.
> I am running Hadoop already so I wanted to go with Hbase
> but the performance
> gap was so big that it made me reconsider.
> 
> Any opinions please?
>