Beginning in 2015, all our official drivers strive to implement the same algorithm to discover and stay connected to your standalone server, replica set, or sharded cluster.

To learn about the Server Discovery And Monitoring Spec:

• "MongoDB Drivers And High Availability: Deep Dive", my presentation at MongoDB World about PyMongo 3's implementation of server discovery and monitoring, along with a reasonable strategy for operation retries in your application.
• My article "Server Discovery And Monitoring In PyMongo, Perl, And C".
• My colleague David Golden's article on the related Server Selection Spec.
• A summary of the spec.
• The Server Discovery And Monitoring Spec.
• The spec source, including the YAML test files.
• Slides from my MongoDB World 2015 talk, "MongoDB Drivers And High Availability: Deep Dive"