I am a current Cassandra user evaluating MongoDB Atlas. I’m curious about others’ experiences with MongoDB or MongoDB Atlas.
Has it been able to meet your write throughput needs? My applications regularly exceed 150,000 write operations per second. Has anyone had any good or bad experiences with Mongo or Atlas at that level?
I feel like basic CRUD operations are going to be great on Mongo (as long as I can meet throughput needs) but I’m wondering if I should rely on Mongo’s more advanced functionality such as aggregations and filters or if I should plan to let my app handle that. Any thoughts?
One of Cassandra’s banner features is Multi-DC replication. I’ve read mixed reviews about Atlas’ cross-region replication so this is a particular area of concern for me. Any experiences with this?
Last, does anyone have any experiences with Atlas’ VPC peering? Does it work well and is it cost-effective?
Hi, did you find any answers for this? Even I am in similar doubts now… would like to how mongodb perform for larger read and write as compare to Cassandra
Hi there
There is no doubt Cassandra offers good scalability for key-value workloads. MongoDB is also highly capable for the most performance and availability-demanding applications, demonstrated by a selection of examples shown on this page: MongoDB At Scale | MongoDB
Cassandra tends to be well suited for workloads that need to insert data quickly, but where that data is rarely updated, and is accessed only by its primary key or by a limited set of secondary indexes. For queries any more complex than simple point look-ups or range scans, the data will generally need to be replicated from Cassandra to dedicated analytics and search nodes.
MongoDB offers a number of capabilities that enable organizations to ship more functional applications faster with lower cost and complexity. These capabilities include its intuitive and flexible document data model, powerful query engine and aggregation pipeline, secondary indexes, transactional ACID guarantees and strong consistency.
With native multi-region, multi-cloud sharding and replication, MongoDB Atlas can be securely scaled out to support global applications. This is further enhanced with MongoDB Cloud (MongoDB Cloud | MongoDB) offering MongoDB Atlas, Search, and Data Lake to serve different workloads through a common API, while Realm Database extends the data foundation to the edge.
I hope this information is useful, and I’d be keen to hear how you progress in your evaluation of MongoDB. Please don’t hesitate to contact us at MongoDB if you have any questions, or can help in any way.