What is MongoDB Atlas Good For?

[WorldImpex]

1. Are we currently using any of Mongodb Atlas's services (backups, provisioning, etc.)?
2. What do we gain by hosting the DB ourselves?
3. Should we be using / continue to use Mongodb Atlas?
4. How much time and effort would it take to switch all instances to GCP?

[WorldImpex]

@jj105

I checked services for no-sql database, and our current status of mongodb atlas to fix 186. at the moment we are billing price for atlas-instance, atlas-data-transfer service and cloud backup service on mongodb atlas. daily price for altas-instances won't changed unless we update our instance. (we are consistently using $1.87 per day for dev and prod database at the moment.) price for atlas-data-transfer service depends on number of users we have. (at the moment, we are spending $0.06-0.07 per day for each of dev and prod database at the moment because we are using 6-7 GB of data each day.) and we backup database every 6 hours and I think this service makes us to spend about $0.005 for each day. I checked gcp services for no-sql database as well. on gcp, there's firestore service which is for no-sql that we are using at the moment. but I think moving our backend from mongodb to firestore requires some time because their queries and supported functions are different so we need to update all of our database functions and queries to make our backend work with firestore. pricing for gcp firestore is different from mongodb altas - firestore charges by number of document and we read/write in the database. Document reads - $0.033 per 100,000 document Document writes - $0.099 per 100,000 documents Document deletes - $0.011 per 100,000 documents mongodb atlas vs gcp firestore when we are having 100000 users, I think price for mongodb atlas is less than gcp firestore because most of our documents (except applet) don't have much size, so billing by GB usage (mongodb atlas) is better than billing by number of documents (gcp firestore). GB usage on mongodb atlas can be increased dramatically when we are having 100000 users. (if we cache applets efficiently on mobile app we can prevent it from increasing dramatically I think.) in my opinion, mongodb atlas service is good for our case ... but if we are going to move to gcp service, I think moving backend at this time is better than moving it later, because number of backend functionalities are being increased, so if we move backend later it'll require more work than we move at this time.