Cloud infrastructure has changed the way we architect, develop and deploy services. Starting from Infrastructure as a Service, we have Platform as a Service and a variety of SaaS tools available. Database as a Service stands among as a Service offerings, as a database is typically one of the most operational heavy elements of many software systems.
MongoDB, as one of the most popular NoSQL databases and a non-relational database used in many cases in Web based systems, has grown along side PaaS and IaaS providers growth. DBaaS with MongoDB as the backing database is available since early versions on Heroku, with MongoHQ, MongoLab and ObjectRocket being the main offerings. MongoHQ has since been renamed to Compose and is part of IBM, ObjectRocket has been acquired by RackSpace and MongoLab has been renamed to mLab. Most of these services have since expanded to offer a whole range of DBaaS offerings from Redis to Cassandra or ElasticSearch.
MongoDB Inc., the company behind MongoDB, launched its own offering in the field of DBaaS last June, MongoDB Atlas. MongoDB Atlas, like most DBaaS, aims to take away the complexity of setting up, operating and scaling a MongoDB database and help developers focus on their core application development efforts. That being said, architects should still follow best practices when developing against a cloud database.
MongoDB Atlas has just announced availability in 14 AWS regions (with nine being new) globally with each replica set being deployed in a separate AWS availability zone (except for cases like London that only has two AZ). Local storage is now available in Americas, EMEA and APAC areas, allowing for data to be closer to the customer and enabling applications to comply with various data protection acts and directives. MongoDB Atlas also announced a new tool to help migrating existing datasets, reducing operational overhead and downtime named Live Migrations.
MongoDB Atlas has a free tier as of last month and usage is billed by server, usage and a set of features selected as shown here.
