June 2025 Summaries

Role-Based Access Control (RBAC) is a critical component for secure and scalable platform design in cloud infrastructure, effectively managing access by assigning roles instead of individual permissions. When applied across multiple cloud providers like AWS, GCP, and Azure, as well as within Kubernetes clusters, RBAC ensures engineers receive only the necessary access, maintaining security without hindering productivity. In Kubernetes, RBAC operates at both the cloud provider level and within the cluster through IAM policies, ClusterRoles, and ClusterRoleBindings. CockroachDB Cloud implements a robust RBAC strategy by granting engineers read-only credentials and using just-in-time (JIT) access for temporary admin needs, leveraging tools like the Lumos autonomous identity platform. This approach maintains operational safety while allowing developers and SREs to efficiently manage workloads across AKS, EKS, and GKE. The key to successful RBAC lies in aligning cloud IAM roles with Kubernetes access and establishing workflows for temporary privilege elevation, ensuring a consistent and secure framework across different cloud environments.

RoachFest is a distinctive tech conference focused on the real-world challenges of building and running distributed applications, emphasizing practical insights over sales pitches. Following its initial event in London, RoachFest expands to Bengaluru and Las Vegas, offering deep dives into the future of distributed SQL databases with a focus on CockroachDB. The conference features keynotes by Spencer Kimball, CEO of Cockroach Labs, who introduces a new benchmarking framework to test database performance under real-world adversities. Attendees can engage in hands-on workshops to learn about distributed SQL and multi-region database architecture, led by experts from Cockroach Labs. The event also showcases success stories from companies like Yubi, Global Payments, FanDuel, and Squarespace, illustrating the benefits of adopting CockroachDB for scalability and reliability. Industry luminary Kelsey Hightower will deliver a keynote on building dependable and adaptable systems, emphasizing the human aspect of technology. RoachFest is designed for professionals tasked with maintaining uptime, ensuring regulatory compliance, and driving innovation in rapidly evolving environments.

A new report by Intellyx, titled "The New Blueprint for AI-Ready Data: Vector Search Meets Distributed SQL," explores the challenges and opportunities of integrating AI features into existing data infrastructures without adding separate vector databases. The report argues against the prevalent approach of using standalone vector databases, as this often results in increased complexity, data inconsistencies, and operational fragility. Instead, it highlights the innovative C-SPANN solution by Cockroach Labs, which integrates high-performance vector search capabilities directly into the distributed SQL database CockroachDB. This approach allows organizations to efficiently manage AI workloads alongside operational data, ensuring consistency, scalability, and real-time updates without the need for separate systems. The report emphasizes the practical benefits of this unified architecture, such as enabling complex queries and maintaining data freshness, and presents real-world applications in sectors like retail, insurance, and banking. Overall, the analysis suggests that incorporating vector indexing into distributed SQL databases can simplify data management and enhance AI capabilities, offering a compelling alternative to the fragmented data strategies currently in use.

The text explores the evolution of technology in handling large-scale user-generated content, particularly in photo-sharing apps, and highlights the advancements in AI-powered semantic search capabilities. As traditional keyword searches are no longer sufficient, the demand for search results based on content meaning has increased. CockroachDB, a distributed SQL database, addresses this by applying innovative vector indexing techniques, specifically through the introduction of C-SPANN. This algorithm efficiently manages and searches vectors by embedding meaning into them, reducing complex problems to simpler tasks of finding nearby vectors. The system emphasizes scalability, fault tolerance, and real-time updates without relying on a central coordinator, and adapts partitioning strategies to ensure even distribution of data and workload across clusters. C-SPANN utilizes quantization to significantly compress vector sizes, maintaining accuracy while reducing storage and computational costs. This approach allows personalized searches by partitioning indexes based on ownership, ensuring efficient and secure data handling across regions. The text concludes by mentioning ongoing improvements and inviting users to try CockroachDB, highlighting its capabilities for managing large vector datasets.

With the recent GCP and AWS outage, resilience is more crucial than ever. RoachFest25 London's virtual replay will bring together key insights from top sessions, product updates, and customer stories that defined this year's database event. The event features a curated experience with Rob Reid and Cassie Zimmerman hosting live discussions on topics such as performance under adversity, modern identity and access management, data migration, and emerging regulations like DORA and GDPR. This virtual experience offers a front-row seat to innovations shaping distributed databases and is not to be missed for those interested in performance, resilience, and AI-driven data workflows.

RAG is a method that allows Large Language Models (LLMs) to access external knowledge sources as context, improving accuracy and reliability. By introducing grounded data, RAG addresses the limitations of LLMs, such as hallucinations, which occur when models generate inaccuracies or factually incorrect statements. RAG is increasingly popular in various industries, including customer support, content generation, knowledge base queries, and real-time information retrieval. CockroachDB is a highly scalable and resilient distributed SQL database that provides an ideal backend for building efficient and robust RAG applications. The unified storage of source data, metadata, and vector data in CockroachDB simplifies data management, lowers costs, and improves performance for RAG applications. Additionally, CockroachDB's scalability, reliability, and simplicity of operations make it well-suited for powering RAG systems that handle intensive workloads and large data volumes. The database also provides advanced access control and security features, such as role-based access controls and row-level security, to ensure that users access only the data they're authorized to view. By leveraging CockroachDB's vector capabilities, developers can scale their RAG applications while maintaining consistent performance under increasing load. A demo of a knowledge base chatbot using RAG on CockroachDB is available for interactive exploration.

Modern applications require always-on identity management to meet the demands of global, connected users. Legacy systems can't handle this level of demand and often suffer from downtime, which can cripple product experiences. Ory redefines IAM for the cloud-native era with modular, open-source, API-first components that scale independently and support complex use cases like fine-grained access control, passwordless login, and social sign-ins. OpenAI, a leading AI company, needed a resilient, multi-region CIAM platform to support its global user base for ChatGPT. They deployed Ory's identity stack on CockroachDB, a distributed SQL database known for enterprise-ready resilience, to achieve global consistency, automated failover, and high availability across regions. This solution ensures secure, scalable, always-available IAM and has become a baseline requirement for customer-facing applications. The integration of Ory and CockroachDB provides a powerful pattern for building composable, scalable IAM backed by a globally distributed, resilient SQL database. By adopting this approach, organizations can build secure, available, and consistent identity systems that support uninterrupted access and engagement.

CockroachDB, a distributed SQL database, has been benchmarked using a new methodology called "Performance under Adversity" which tests the database's resilience in various failure scenarios such as network partitions, regional outages, disk stalls, and node restarts. The benchmark reveals that CockroachDB maintains steady throughput, low latency, and zero downtime even when faced with these complex operational realities of modern enterprise systems. The database's engineered features, including back-pressured range rebalancing, admission control, co-located leaseholders, and multi-disk WAL failover, contribute to its resilience. This new benchmark demonstrates that CockroachDB is not only fast but also reliable and available in the face of adversity, making it an attractive option for modern enterprises.

SQL has been a widely adopted lingua franca for data management for decades due to its rich functionality and strong consistency guarantees. However, as businesses scale, traditional SQL databases face challenges such as maintaining global transactional consistency and resilience in the face of growing transaction volumes. Distributed SQL databases are redefining what's possible by combining the benefits of traditional SQL with the scalability and resilience of NoSQL databases. CockroachDB, a distributed SQL database, is designed to handle modern demands such as 24/7 availability, instant scaling up and down, and regular maintenance without downtime. By leveraging technologies like multi-version concurrency control, admission control, and rolling upgrades, CockroachDB provides inherent resilience, allows for online schema changes, and enables businesses to scale with their needs, making it an attractive solution for global-scale businesses.

Change data capture (CDC) is a crucial infrastructure pattern that enables event-driven microservices, real-time analytics, and incremental ELT pipelines. Debezium, an open-source CDC tool, has become the standard for many organizations, but its adoption comes with the cost and complexity of a distributed streaming platform. CockroachDB's native CDC system takes a different approach by building changefeeds directly into the database engine, removing the need for Kafka Connect and its associated infrastructure. The new enriched changefeeds in v.25.2 of CockroachDB can match Debezium's message format, providing operational simplicity without breaking downstream consumers. This allows teams to migrate off Kafka Connect without giving up format formality, making it an attractive alternative for those looking to simplify their CDC pipelines and reduce infrastructure overhead.

Payment systems are facing a critical juncture, with legacy infrastructure struggling to keep pace with modern demands for speed, scale, and security. To stay competitive, enterprises must modernize their payment systems, and experts will share proven strategies for doing so at an upcoming event hosted by Cockroach Labs and AWS. Industry leaders from Atos Group, AWS, and Red Hat will present advanced modernization approaches that enable payments to handle demanding conditions while keeping pace with AI applications. Additionally, a fireside chat with AuthZed's Jake Moshenko and David Joy will focus on planning ahead for authorization at scale, highlighting the importance of resilience in payment systems. The event also includes a casual round of golf and socializing opportunities, providing a unique blend of technical learning and networking.

The world of Generative AI (GenAI) has revolutionized how we store, query, and analyze data, leveraging vector embeddings to understand, organize, and generate meaningful content. However, traditional databases struggle to handle the high-dimensional and noisy nature of unstructured data, making specialized systems like vector databases essential for efficient storage and retrieval. CockroachDB, a highly resilient distributed database, has emerged as a powerful backend for Retrieval-Augmented Generation (RAG) systems, delivering fresh, scalable, and consistent data to enrich AI-generated content. With its built-in Vector datatype, strong consistency, and 99.999% SLA in its cloud offering, CockroachDB enables the performance and availability required for next-gen workloads, making it an exciting addition to the field of GenAI.

CockroachDB has released vector indexing support into preview, enabling customers to combine transactional data with vector data in a single data store without compromising the underlying fundamental promises of distributed SQL. This innovation is achieved through Cockroach-SPANN (C-SPANN), a new distributed vector indexing algorithm that adapts Microsoft's SPANN and SPFresh algorithms to work well with CockroachDB's unique distributed architecture, enabling efficient answer approximate nearest neighbor queries with high accuracy, low latency, and fresh results. C-SPANN supports accuracy, freshness, scalability, latency, resource usage, and build performance, allowing users to create secondary indexes to accelerate searches, retrieve results in the low 10's of milliseconds, and use a fraction of available CPU and memory during index builds. CockroachDB brings vector similarity searches into developer's hands with the full power of SQL, enabling simple filtering, joins, and aggregations in combination with vector queries, and taking full advantage of its key promises around resilience, scale, performance, geo-data placement, and low cost of ownership.

The CockroachDB team has released version 25.2, a major update that celebrates 10 years of continuous distributed SQL innovation. This release introduces significant performance and scalability improvements, including a 50% increase in throughput compared to the previous version. Other notable features include buffered writes, which improves performance by reducing latency and hardware requirements, database restore enhancements that speed up restoration from backups, vector indexing for AI applications, row-level security, enhanced observability, new CockroachDB Cloud regions, and plan switching without downtime. These updates aim to address scalability concerns of vector applications and provide robust mechanisms for managing roles and privileges to restrict access to sensitive data.

When a product experiences exponential growth, like ChatGPT, its infrastructure must scale at an equally fast pace. OpenAI, the company behind ChatGPT, partnered with Cockroach Labs to build a modern customer identity and access management (CIAM) stack using Ory and CockroachDB. This solution enables unmatched developer agility and infrastructure confidence, allowing OpenAI to support over 400 million weekly active users while maintaining security, scalability, and global availability. The combination of Ory and CockroachDB provides an elastically scalable and resilient infrastructure with native geo-partitioning capabilities, ACID compliance, and operational simplicity, empowering developers to build secure, scalable, and globally distributed applications without sacrificing customer experience or security.