Google vs MS (Was Re: [Ilugc] Google vs Yahoo, again!)
- From: venkat83@xxxxxxxxx (Venkatraman S)
- Date: Wed Jul 2 20:10:15 2008
On Wed, Jul 2, 2008 at 1:54 PM, Suraj Kumar <suraj@xxxxxxxxx> wrote:
Extracting and storing an RSS feed into a database is a very similar
question too. It has quite a few elements that can be optimised and
they were just testing your design, data structure and approach, IMO.
Imagine, you have a billion RSS feeds to process, wouldn't you have to
think differently about how you'll do trivial things such as sorting
or defining your database schema? Would you even be able to run a
single database instance? ... and how are you going to be
fault-tolerant?
Am sorry for i havent replied to the bulk of the email - i felt asleep..i
got blinded. Suddenly i woke up and saw some references to 'you' and since
the previous email was mine - then i presume 'you is me' :P
I guess you didnt read the Q properly : storing the fields in a Mysql
table(as in 'desc tablename') is different from creating a format(read
'Data structure') in which lookups are faster and is also feasible on the
space(we learnt it in college as 'Time/Space complexity').
On second thoughts, probably you are right, probably the interviewee should
have anticipated the Qs that 'desc tablename' will lead to 'design a custom
rdbms'.
beers,
Venkat
Other related posts:
