What you are describing is MapReduce application scenario, where the DB is
handled from your MR code, nothing special from Hadoop side.
On Wed, Aug 10, 2011 at 5:34 AM, Segel, Mike <[EMAIL PROTECTED]> wrote:
> It's been far too many years since I was handed my diploma and kicked off
> campus. :-)
> IMHO it's a bit esoteric for a class room homework assignment. Maybe an
> interview question used to stump most of the candidates?
> The funny thing is that on the walk to work, I started to think of if it
> made sense for a certain subset of m/r problems to use derby as an in memory
> db/local lightweight DB
> Ok and before you say WTF, I'm talking about a subset of problems where
> depending on the input to the Mapper.map() method, you may want to do a
> quick look up against a database which contains lookup data that is static
> and is indexed.
> Ok, so maybe I shouldn't read my e-mails before heading off to work... :-)
> -----Original Message-----
> From: Ted Dunning [mailto:[EMAIL PROTECTED]]
> Sent: Wednesday, August 10, 2011 12:55 AM
> To: [EMAIL PROTECTED]
> Subject: Re: Derby with Hadoop --Why?
> No. He meant nothing of the kind. The other explanations expanded on
> This sounds like homework. If so, I would recommend a bit of reading
> before asking.
> On Tue, Aug 9, 2011 at 10:37 PM, Saravana Kumar
> <[EMAIL PROTECTED]>wrote:
> The information contained in this communication may be CONFIDENTIAL and is
> intended only for the use of the recipient(s) named above. If you are not
> the intended recipient, you are hereby notified that any dissemination,
> distribution, or copying of this communication, or any of its contents, is
> strictly prohibited. If you have received this communication in error,
> please notify the sender and delete/destroy the original message and any
> copy of it from your computer or paper files.