For those unfamiliar with MongoDB, the official definition is that it is an open source database management system (DBMS) that uses a document-oriented database model which supports various forms of data. It is one of numerous nonrelational database technologies which arose in the mid-2000s under the NoSQL banner for use in big data applications and other processing jobs involving data that doesn't fit well in a rigid relational model. Instead of using tables and rows as in relational databases, the MongoDB architecture is made up of collections and documents.
The relational data model has been the mainstay database for developers for decades.
In his most recent book, ‘Next Generation Databases: NoSQL, NewSQL, and Big Data’, Guy Harrison writes about the database options available to developers when making what have become complex decisions that can have significant business implications. Harrison writes extensively on emerging database technologies and in his book, states that the relational model remains relevant and, in his opinion, will continue to be the database technology most used for the majority of database solutions, certainly in the short term. But today, other database types are required for new and advanced circumstances – especially in the realm of horizontal scaling and document focused databases. As these circumstances develop, Harrison sees a convergence happening between different database types. For example, there is now (limited) support for JSON in Oracle; with SQL and (limited) transaction support in MongoDB.
This is the context in which MongoDB has grown. As a database architect, a crucial part of my role is to constantly consider the future database strategies for Datalex. At Datalex, our reputation as a leading provider of digital commerce has been built on our expertise, and the agile and innovative digital technologies that we provide to travel retailers. It is precisely this need for agility and innovation that has made MongoDB such a good fit for our business for certain applications.
I spoke with MongoDB CTO & Co-Founder Eliot Horowitz about this at the developer advocacy day and asked him honestly where he felt MongoDB was best placed to support modern database demands. Eliot accepted the difficulty of porting existing databases to MongoDB (porting between different relational databases is difficult enough) but there is a huge appetite for new applications to have native JSON support, replica sets and sharding. Eliot also highlighted the power of the aggregation framework over traditional SQL. While traditional RDBMSs SQL optimisers have benefited from being fine-tuned over the last 40 years, he remarked that large SQL code can be a nightmare to understand and support compared to an aggregation pipeline. It was made clear at the advocacy day that the goal of MongoDB is to be a general purpose database and this will be the focus for all future enhancements.
In terms of the latest features of the MongoDB platform, I also spoke with a number of senior project managers, who provided their insights:
The travel retailers that Datalex works with demand speed and agility from their technology partners to lead, compete and grow. MongoDB is based on agile methodologies, offering dynamic schema that easily evolve with developers’ applications. This gives Datalex the flexibility to create and develop the transformative technologies and solutions our customers need to continue to grow their businesses, and to innovate at pace.
In terms of the latest features of the MongoDB platform, I also spoke with a number of senior project managers, who provided their insights:
- This year saw the introduction of transactions in version 4.0. Project Manager for Transactions Aly Cabral, said that the introduction of transactions on replica sets for version 4.0 is a major enhancement. It is a limited form of transactions compared to RDBMSs with limits on time and volume of data, but will be applicable to most simple use cases of transactions.
- Transactions for shards is being introduced in version 4.2. As expected it will be implemented using a two-phase commit approach (classic distributed transaction implementation). Aly stated that that most transactions will still be expected to operate on a single shard but this new feature will allow some specific key use cases.
- Project Manager for Business Intelligence, Seth Payne, talked about the BI connector for MongoDB. One of the benefits of this connector is that it allows the developer to run SQL statements against the MongoDB database. The SQL statements are transformed to aggregation commands in order to obtain the data. This is working well for MongoDB for single collections, even with multiple levels of embedded data.
- Finally, a demonstration of Atlas showcased MongoDB’s automated cloud service, built with agile teams in mind.
The travel retailers that Datalex works with demand speed and agility from their technology partners to lead, compete and grow. MongoDB is based on agile methodologies, offering dynamic schema that easily evolve with developers’ applications. This gives Datalex the flexibility to create and develop the transformative technologies and solutions our customers need to continue to grow their businesses, and to innovate at pace.
Conor Brophy
Database Architect , Datalex
