You should use mongodb sharding and pick a shard key that isn't by company
or date. Then add secondary indexes to the collection. You only need one
collection. You can use separate databases per client or separate
collections, but I don't believe that is needed.
On Tuesday, May 10, 2016 at 10:38:15 AM UTC-4, Boas Enkler wrote:
I'm currently migrating a huge amount of data to mongodb.
There are abour 10 Mio Documents. They have are defined by a date and a
assigned company id.
Now I'm thinking about wether it would make any difference when I split
the data . So that I would have foreach company and / or date a collection
Is it to be expected to have performance differences when having all
documents in one collection or in multiple collections / databases?
You received this message because you are subscribed to the Google Groups "mongodb-user"