hadoop-common-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ricky Ho <...@adobe.com>
Subject RE: Q about storage architecture
Date Sun, 07 Dec 2008 18:12:01 GMT
Without familiarity of your classification tree, I think HDFS is a virtualized data store following
a very common design pattern of partitioning and replication.  (e.g. Distributed Hash Table).

The details is in how the partitions are splits, where is the location of replicas, what are
the consistency model, what semantics is exposed to the users ... etc.

So, HDFS is "File system semantics", "master/slave", "read-any-write-all"

-----Original Message-----
From: Sirisha Akkala [mailto:akkala@Colorado.edu]
Sent: Saturday, December 06, 2008 10:55 PM
To: core-user@hadoop.apache.org
Subject: Re: Q about storage architecture

Thanks for the link. I had gone through the material. I'm looking for help to classify this
architecture into broader category, if possible. SAN/NAS/any other distributed terminology


---- Original message ----
Date: Sat, 06 Dec 2008 22:22:49 -0800
From: Sagar Naik <snaik@attributor.com>
Subject: Re: Q about storage architecture
To: core-user@hadoop.apache.org


Sirisha Akkala wrote:
 I would like to know if Hadoop architecture more resembles SAN or NAS? -I'm guessing it is
 Or does it fall under a totally different category? If so, can you please email brief information?


View raw message