September 2018 Summaries
2 posts from Kong
Filter
Month:
Year:
Post Summaries
Back to Blog
Kong has reached a significant milestone with the release of Kong 1.0, marking the maturity of this open-source project, which has been widely adopted by large companies and government agencies worldwide. Originally launched nearly four years ago as Kong Community Edition (CE), it has now been renamed simply as Kong, while its counterpart, Kong Enterprise Edition (EE), will be known as Kong Enterprise. This update introduces major features including Service Mesh support, enabling mutual TLS between Kong instances, and a new Database Abstraction Object (DAO) that simplifies database migrations with minimal downtime. The release signifies that Kong's API is stable, backward-compatible, and ready for future advancements, emphasizing its readiness for production environments. Additionally, popular enhancements such as route naming and HTTPS health checks have been added. The release is celebrated as a stepping stone towards future platform innovation, supported by the community and users who continuously test Kong's capabilities.
Sep 18, 2018
529 words in the original blog post.
Kong Cloud has been utilizing StatsD and Prometheus for monitoring and metrics collection but encountered performance issues with high CPU usage from the StatsD Prometheus exporter during peak request rates. The exporter, which translates StatsD metrics to Prometheus formats, was consuming excessive CPU resources, prompting a profiling analysis that revealed significant CPU time spent on regular expression processing. To address this, Kong Cloud developers refactored the exporter to implement a lightweight matcher using a finite state machine, tailored to their specific use cases which did not require the full complexity of regular expressions. This optimization reduced CPU consumption significantly by minimizing reliance on the Go regular expression library, thereby improving performance by concentrating processing power on system calls and the Go runtime instead. These enhancements led to a 40% reduction in processing time for matching functions and were integrated into the Prometheus StatsD Exporter version v0.8.0-rc1, released in October 2018. Kong Cloud, still in beta, offers seamless updates and is compatible with major cloud providers like AWS, Azure, and Google Cloud.
Sep 05, 2018
1,789 words in the original blog post.