cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Masood Mortazavi <masoodmortaz...@gmail.com>
Subject Re: The Difference Between Cassandra and HBase
Date Mon, 26 Apr 2010 19:52:28 GMT
On Sat, Apr 24, 2010 at 10:20 AM, dir dir <sikerasakti@gmail.com> wrote:

> In general what is the difference between Cassandra and HBase??
>
> Thanks.
>


Others have already said it ...

Cassandra has a peer architecture, with all peers being essentially
equivalent (minus the concept of a "seed," as far as I can tell).

This is a great architectural advantage of Cassandra and Cassandra-like
systems. It wasn't really possible to make practical systems like this in
earlier ages because of computing (memory, CPU, disk) limitations which made
characteristic times (including expected characteristic response, recovery,
replication, etc. times) and system dynamics almost impossible to deal with.
This problem persists but has become far more manageable because expected
response times haven't evolved or narrowed any faster than computational
capabilities.

HBase on the other hand is a layered system already. It relies on the
underlying HDFS, beyond and above the OS. As a more layered systems, it has
better service architecture, in a sense, but it relies and is limited to the
capabilities of those "services" ... say the distributed file service.

Cassandra rolls its own partitioning and replication mechanisms at the level
of its peers. It does not rely on some underlying system service for these
capabilities. Cassandra is definitely easier to provision and use, from an
operational point of view, and this is a great advantage -- although
installations that afford scanning (through ordered partitioning) would
become more involved.

(As suggested by others, reading the BigTable and Dynamo paper will help you
to establish the difference between HBase and Cassandra in more clear,
architectural terms.)

- m.

Mime
View raw message