If you store tweets and users in separate databases and you use 2.2.x then you will get better concurrency
On 2.0.x you can use db or collection to shard it doesn't make any difference
On Friday, November 2, 2012 10:07:02 AM UTC, 侯德军 wrote:
I have a data set which contain two collections users and tweets, this data come from different source ,current I stored different data by source into different database. there are two method I can do to shard the mongodb. shard
1. Combine all database into single database and shard on collection tweets and users?
2. Keep the all different database, and make sharding on each database?
ps, I will only setup a single sharding cluster, and the data is very huge. which one is better for me?