MongoDB 3.4 was released February 1st and has several new features that customers of ObjectRocket 3 Node Replica Sets should be excited to see!
Three of the most prominent new features that are available in our 3.4 Replica Sets are faceted search capabilities, graph capabilities, and the introduction of views. As with any release there are some minor new features and improvements as well, so we touch on those most relevant to you on our platform.
Faceted search allows for grouping and ordering multiple data points in a single aggregation stage. This helps drill down into the data you want to see, getting only what’s relevant to your needs. This also adds similar capabilities of enterprise search software to MongoDB but uses its own aggregation framework on your existing dataset.
The new graph lookup and associated operators help track down the proverbial six degrees of separation often seen with social data relationships utilized frequently today. It will allow you to set a recursion level and query across one or more collections to see just how your documents are tied together.
MongoDB 3.4 also introduces read-only views similar to those utilized in relational databases! Using an aggregation pipeline and underlying indexes, a view will allow you to create a projection from one or more collections. In situations where granular access is required this can also allow you to restrict specific data access further using MongoDB’s role based authentication.
Furthermore this release allows for improvements to the aggregation pipeline by adding additional stages and operators for functionalities such as recursive searches, field exclusion via $project, new date operators utilizing $iso related functionalities, control flow via the $switch operator, and other new string operators.
Many customers will find these useful for their applications that work to aggregate many aspects of data for fields such as e-commerce related applications that often require searching on many different facets of data including medical and healthcare data or even IoT related events.
The decimal128 format is now used as a new decimal data type allowing support for up to 34 significant digits and an exponent range of −6143 to +6144. This is a great improvement over using integer or double for any application that needs more numerical precision.
Another huge improvement is around how MongoDB 3.4 handles initial syncs by creating indexes as documents are copied over. And they have improved the logic around retries making it a bit more resilient to network issues. What this means for our customers is that weekly compactions finish faster, making sure your data footprint is as small as possible!
Coming up next for our ObjectRocket MongoDB customers, we are working on adding MongoDB 3.4 for sharded clusters. Improvements to expect there will be related to balancing and improving speed and reliability associated with balancing operations. Parallel chunk migrations drastically reduce the time required to balance a collection and allow for shorter balancing windows. To see more differences between our Replica Sets and Sharded plans, please take a look at our MongoDB overview page. For customers who want even more customization or larger deployments than our standard plans allow, we recommend looking at our hosted MongoDB options.