Today marks the opening of KubeCon + CloudNativeCon, with a number of events spotlighting several of the projects being advanced under the auspices of the Cloud Native Computing Foundation (CNCF). The co-located events, being held in Atlanta, include CiliumCon, ArgoCon, BackstageCon, Open Source SecurityCon and more.
Many of the sponsors showcasing products at KubeCon have announcements. Among then are:
Yugabyte releases Distributed Database Trends Report
The distributed database company’s report found that innovation and modernization aimed at creating next-generation, AI-based applications is a top organizational priority among 76% of respondents, while 63% are prioritizing application modernization.
The report noted that legacy databases are a “significant roadblock” to modernization efforts. The biggest challenges created by legacy databases were inefficient operations and slow innovation, reported by 50% of respondents, followed by limited scalability (43%), high costs (36%), and resilience issues (30%).
Re-architecting systems for cloud-native or distributed databases was a solution for overcoming those challenges reported by (48%) of respondents, and 45% said they are migrating to fully managed cloud databases. When looking for modern database solutions, 71% reported that high availability and resilience were the most critical attributes they sought, followed by scalability (69%) and enterprise-grade security (60%).
“These findings signal that tech leaders are making a decisive move toward modernization, underscoring the need for informed architectural choices that not only support AI strategy but also ultra-resilience,” said Karthik Ranganathan, co-founder and co-CEO of Yugabyte. “To support the move from legacy databases to highly resilient and seamlessly scalable distributed databases, we are offering AI agents to power migrations and AI-powered Performance Advisor for ongoing monitoring and tuning. These findings validate our investments to create cutting-edge tools to support enterprises in their modernization journey.”
For the full research, download the full Distributed Database Trends report here.
Gremlin, Dynatrace partner on maintaining application states in Kubernetes
Chaos engineering platform pioneer Gremlin and observability and application performance monitoring provider Dynatrace announced they have created an integration to improve reliability testing for Kubernetes environments. The integration will improve the ability to perform fault injection tests in Kubernetes applications, which helps organizaitons maintain those applications in the desired state.
“Many teams have faced challenges operationalizing reliability testing across complex cloud-native architectures, often requiring multiple manual steps to identify and target the right resources. By combining advanced AI observability and topology insights with Gremlin’s fault injection and reliability capabilities, customers can more easily identify, test, optimize, and strengthen critical services at scale,” Samuel Rossoff, CTO of Gremlin, said in the company announcement.
With the joint solution, Kubernetes services can now be automatically discovered within Gremlin, leveraging Dynatrace’s AI-driven observability and topology mapping. The statement said that “health checks are then applied to Kubernetes’ objects, allowing organizations to efficiently implement standardized reliability testing and gain deeper insights into their environments.”
Wayne Segar, Global Field CTO at Dynatrace, said in the statement: “As AI-driven innovation accelerates, the reliability of Kubernetes becomes mission-critical. Our partnership with Gremlin simplifies chaos engineering, helping teams ensure resilience and performance across complex, distributed systems.”
Mezmo Launches AI SRE
AI agent telemetry platform provider Mezmo released what it is calling the world’s fastest and most accurate AI SRE (Site Reliability Engineering) agent for root cause analysis ahead of KubeCon. The company uses context engineering to help AI agents work with speed and precision in finding the root cause of issues.
The company cited recent LLM benchmarks that it says reveal struggles among models such as Claude Sonnet 4, OpenAI GPT-4.1, Gemini 2,5 and GPT-5 with basic observability tasks. saying they lack the right context to perform better. The announcement said that when Mezmo’s platform was benchmarked against these other models, the results found a 90% cost reduction per incident, root cause analysis with much less prompting, and more efficiency with tokens, using 27,000 instead of more than 500,000.
