How will you store millions of subscribers list (assume it as email id) ?
8370

How will you store millions of subscribers list (assume it as email id) In distributed systems ? And send email notifications to all of them.

This system should be scalable and fault tolerant and as fast as possible (parallel processing).

EDIT:
Millions of subscribers for each category let's say. There are several thousand categories(I over simplified the question.)

I'm aware about how to handle it in RDBMS for medium to large scale. I'm not sure how to handle at big data scale.

Should we use HDFS ? No SQL ?
How do we partition them ? Or Shard ? And how do we fetch it ?
How to ensure there is no single point of failure in persistence ?
How to retrieve them in and process in parallel ? i.e emails can be sent in parallel with available resources

Comments (5)