Will mongodb a good choice as a data warehouse considering following situation?
- We are building analytics, report generation platform for our client and a typical table growth in terms of rows will be 2 million rows every year (We need 5 yr history at max, total tables ~ 50)
- We have proof of concept ready using mysql and it works but my worry is as we hit production the growth in data might create issues with mysql. mysql doesn't support horizontal scaling whereas sharding in mongodb makes it scale well
- On average once or twice a year schema of a table changes where either new column is being added or existing one is deleted (deletion is not a concern as we can always keep the column but how difficult it would be to add new column in mysql table containing few million records?). This is other reason we are thinking about mongodb as it is flexible in terms of schema changes
- Writes: Average 5000 records per table per day. Reads will be less then this and it will be performed for generating analytics.
I don't have much experience dealing with this kind of data volume both with mysql and mongodb hence pardon my ignorance if some of the things mentioned above are naive.