Overall Satisfaction with MongoDB
For a few years now our company has been replacing some very expensive Oracle DB's with much less expensive and lightweight combination of MongoDB with Elastic Search assisted collections. We have some extremely data heavy collections that used to take upwards of 30 seconds to search. With only Mongo collections and not having a normalized database I think we would have seen improvements, however, using Mongo in conjunction with Elastic has allowed us to make similar and more complex queries in fractions of a second.
- Easy to set up in AWS.
- Easy to scale. If you're worried about growth while maintaining consistent performance adding nodes is easy.
- Mongo typically will typically require more storage space for the "same" amount of data stored in a normalized database.
- Many features of other popular databases aren't available in Mongo such as Joins and Transactions.
- Being able to get rid of many of our Oracle licenses has saved the company a lot of money. A big part of that was being able to reduce our cores because the load without some of our search calls lowered our requirements significantly. However now that that problem has been solved we don't have any real dependencies holding us to Oracle and are moving everything which will allow us to get rid of the rest of our licenses.
Looking into PostgreSQL happened post move to Mongo. Had we considered both options at the time we likely would have went with PostgreSQL. We may migrate at some point in the future but currently it doesn't make sense.