Can someone please explain me why taller tables are better than wider tables in MapR-DB in terms of performance.
I think I can answer the question...
Here's the short answer: Rows cannot exceed the width of a region. So performing queries and accessing rows can become problematic and kill performance. Internally you'll end up with fewer rows in a region.
Going with thin rows, you can fetch/process each row faster and you'll have more rows to a region.
Note: There are some exceptions... but in general you'll want a thin row.
Obviously there's more to it and it could take up an hour in an intro course just talking about this issue. If you need a better answer, maybe this should become a discussion?
Hi Avnish Gupta,
I am not clear what you meant by "taller" and "wider' tables. Could you please let us know which version of MapR-DB you are on and your objective? Thanks.
Retrieving data ...