What about if I spread these columns across 20 rows ? Then I have to query each of these 20 rows for 500 columns. but still this seems a better solution than one row for all cols or separate row for each email id approaches !?
On Fri, Jul 27, 2012 at 11:36 AM, Aklin_81 <asdk...@gmail.com> wrote: > Sorry for the confusion created. I need to store emails registered > just for a single application. So although my data model would fit > into just a single row. But is storing a hundred million columns(col > name size= 8 byte; col value size=4 byte ) in a single row a good idea > ? I am very much tempted to store it in single row but I also heard it > is recommended to keep a row size within 10s of MBs for optimal > performance.