While others use their first major release to signal production readiness, the maintainers of cloud-native API gateway Kong feel that they’re already a step ahead as they hit v1.0 this week.
To them putting the full digit label on their project shows an established and backward compatible API, with future updates not changing but adding functionality. The project has also received a spring clean, getting rid of leftover API entities and receiving a plug-in development kit (PDK).
Kong 1.0 comes with a few new features nevertheless. Improvements include name properties for routes and HTTPS health checks. Thanks to mutual Transport Layer Security between Kong instances, and modifications to the plug-in run loop, it can be used as a sort of service mesh, brokering data flow between services.
Another addition is a new database abstraction object, DAO for short. It should help migrating between database schemas when upgrading to a new Kong version, and allow for complete upgrades. Earlier manual intervention would have been necessary to upgrade nodes sequentially.
With the release of version 1.0, the Kong project also renames its editions. The Apache-2-licensed community edition Kong CE from now on only goes by the name of Kong, while its enterprise edition (Kong EE) has become Kong Enterprise. The latter offers an additional Admin GUI, security features such as Oauth 2.0 introspection, a developer portal, analytics and scalability functions, as well as 24×7 support and migration assistance (pricing available on request only).
Kong can be used as a gateway or sidecar for microservice requests, and provides functionality for load balancing, logging, and authentication, amongst other things. Features include circuit breaking for unhealthy upstream services, bot detection, invocation of serverless functions, geo replication, rate limiting, and caching. Kong distributions are available for Docker, various Linux distributions, Vagrant, Homebrew, CloudFormation, AWS AMI, Microsoft Azure, Heroku, and Instaclustr for Cassandra clusters. Nokia, IBM, and Intel are amongst the users of the project.