November 2025 Summaries

Kubernetes has become a standard for managing stateless workloads with agility, elasticity, and resilience, but the challenge remains for stateful databases. CockroachDB addresses this by combining the reliability of traditional SQL databases with cloud-native scalability. It is a distributed SQL database that provides full ACID compliance, horizontal scalability, and automatic data sharding, replication, and rebalancing. This makes it particularly compatible with Kubernetes, aligning with its core principles of resiliency, scalability, and self-healing. CockroachDB can be deployed across various environments, including on-premises and cloud platforms, and supports automation through Kubernetes Operators. It effectively manages the challenges associated with stateful workloads on Kubernetes by ensuring strong consistency and resilience even during node failures or scaling events. Ultimately, CockroachDB makes stateful workloads as flexible and reliable as stateless services, leveraging Kubernetes' capabilities to maintain high availability and performance.

At the Cockroach Connect event in San Francisco, hosted by Cockroach Labs during Microsoft Ignite, a panel discussion led by CEO Spencer Kimball explored the profound impacts of AI on technology infrastructure and business operations. Panelists, including investors Peter Fenton, Satish Dharmaraj, and Lucas Swisher, highlighted the significant architectural shifts driven by AI, particularly the need for resilient and globally distributed systems to handle continuously running, autonomous AI workloads. The conversation underscored the tension between speed to market and the necessity for reliable infrastructure, with AI pushing the boundaries of current tech stacks and creating new demands for computational memory and consistency. The panelists also discussed the investment landscape, predicting both failures and opportunities as AI reshapes industries, emphasizing that future success will depend on adopting AI responsibly while ensuring system reliability and security. Additionally, the event featured insights from partners like ClickHouse and Automation Anywhere on leveraging CockroachDB for real-time AI applications and process automation, reflecting the collaborative efforts to build robust data architectures for the AI era.

Change data capture (CDC) is essential for modern data platforms, supporting event-driven microservices and real-time analytics, and CockroachDB enhances this by incorporating Protocol Buffers (Protobuf) as a serialization option. By adding Protobuf, organizations that standardize on this format, particularly in Kafka-centric architectures, can seamlessly integrate CockroachDB without additional code or converters, reducing latency, cost, and complexity. The Protobuf support includes various format options and schema delivery through an Apache-licensed .proto definition, facilitating a single, versioned contract between producers and consumers. This integration is designed for teams already using Protobuf, aiming to streamline workflows by eliminating the need for transformations from JSON or Avro to Protobuf, and is particularly beneficial for those migrating to CockroachDB from systems where Protobuf is prevalent. While the Protobuf bytes are not canonical and should not be used as stable identifiers, CockroachDB plans to document best practices and explore key-value encoders for improved stability.

As modern applications expand, selecting the right database architecture becomes crucial for handling global user bases, 24/7 availability, and evolving regulations. A new whitepaper titled "CockroachDB vs. Amazon Aurora & DSQL: A Comprehensive Enterprise Distributed SQL Database Analysis" offers an in-depth comparison of Amazon Aurora, Amazon DSQL, and CockroachDB. Amazon Aurora, while efficient for read-heavy workloads and providing a managed PostgreSQL-compatible service, faces limitations in its single-writer architecture, leading to potential performance bottlenecks and disruptions. Aurora DSQL addresses some of these issues with distributed writes but remains limited in scope and maturity, lacking true global availability and multi-cloud support. In contrast, CockroachDB is designed to provide global scale without operational complexity, allowing reads and writes across multiple regions while maintaining strong consistency and transactional guarantees. It also offers flexibility in deployment across various environments and avoids vendor lock-in, making it an appealing choice for enterprises seeking resilience and architectural flexibility. The database landscape is evolving, and this comparison helps organizations make informed decisions about scalable and reliable database solutions.

CockroachDB offers a transformative solution for enterprises by integrating transactional and vector-based AI workloads within a single, globally distributed SQL database, overcoming the traditional separation that leads to data fragmentation and operational complexity. This unified platform supports both structured and unstructured data, enabling real-time, intelligent AI applications that leverage large language models and personalization without the need for multiple specialized systems. CockroachDB's C-SPANN architecture enhances real-time vector indexing, ensuring low-latency retrieval with high accuracy through innovative methods like hierarchical K-means partition trees and RaBitQ quantization. By eliminating data silos and providing a consistent, scalable infrastructure, CockroachDB empowers businesses to create agentic systems that can reason and act autonomously, thus setting a new standard for enterprise AI by unifying performance, cost-efficiency, and scalability.

Enterprises are navigating a complex infrastructure landscape that includes on-prem, hybrid, and multi-cloud environments, necessitating databases that can move seamlessly across these platforms. CockroachDB provides cross-platform portability by abstracting CPU dependencies and encoding data into CPU-agnostic, canonical byte sequences, which ensures consistent performance, security, and compliance across x86-64, ARM64, and IBM LinuxONE architectures. This architecture-agnostic approach prevents issues related to endianness by avoiding the direct writing of in-memory structures to disk, thus maintaining data integrity and allowing true multi-architecture clusters. CockroachDB's design principles include encoding all data types so that they are architecture-neutral, which supports predictable behavior and performance while allowing enterprises to deploy, migrate, or scale across infrastructures with minimal risk. However, limitations exist in mixed-architecture clusters, where certain workloads like DistSQL and geospatial data may produce inconsistent results, although these can be managed by configuration adjustments. Overall, CockroachDB's focus on cross-platform portability provides enterprises with the operational freedom to manage their data infrastructure efficiently and reliably.

Last month's AWS us-east-1 outage highlighted the fragility of cloud-dependent systems, revealing how businesses with single-region reliance suffered significant disruptions while those utilizing resilient architectures, like CockroachDB, maintained operational continuity. This incident served as a wake-up call for companies to reconsider their dependency on cloud monocultures and the hidden risks they pose. Cockroach Labs showcased the robustness of its distributed SQL clusters, which continued to function seamlessly despite the outage, emphasizing the importance of designing systems with fault tolerance and resilience at their core. The event underscored a shift in industry priorities from merely scaling operations to ensuring survivability and business continuity in the face of infrastructure failures. As companies increasingly recognize the cost of downtime, there is a growing emphasis on implementing resilient data architectures to mitigate risks and maintain service availability, positioning resilience as a competitive advantage in the digital economy.

In a recent webinar titled “From Oracle to CockroachDB: Lessons from Real Migrations,” experts Rohan Joshi and Alistair Parry explored how organizations successfully transitioned critical workloads from Oracle to CockroachDB, addressing challenges such as schema conversion and PL/SQL refactoring. The session highlighted how CockroachDB's MOLT (Migrate Off Legacy Technology) toolkit streamlines the migration process by providing a structured approach from planning to validation, ensuring transparency and control. The discussion underscored the importance of resilience over speed, with CockroachDB offering enhanced uptime compared to Oracle, as demonstrated in benchmark tests. Additionally, the webinar emphasized the significance of automation in database modernization, with Cockroach Labs developing AI tools to ease migration complexities, and advised starting with SQL-compliant and OLTP-focused workloads. Ultimately, the migration to CockroachDB is presented as a means to relieve engineering teams from maintenance burdens, enabling them to concentrate on innovation and organizational goals.

RoachFest 2025 brought together a global community of architects, engineers, and innovators across London, Bengaluru, and Las Vegas to explore future-proofing database infrastructure with CockroachDB, focusing on scale, agility, and performance. The event featured technical sessions, customer stories, and product announcements, including the releases of CockroachDB 25.2 and 25.3, which highlighted advancements in performance, reliability, and security. Notable innovations included vector indexing for AI workloads and new tools for Oracle migrations. The event also introduced a new benchmarking methodology, "Performance under Adversity," which tests the resilience and high performance of distributed databases under real-world operational stressors. RoachFest emphasized the importance of reducing operational burdens and infrastructure costs, showcasing customer success stories from various industries and underlining modernization as a competitive advantage. Additionally, the event fostered community collaboration and insights into how distributed architectures are reshaping the enterprise database landscape, reinforcing CockroachDB's position as a foundation for next-generation applications.

Cockroach Labs has enhanced the security features of CockroachDB Cloud on Microsoft Azure by introducing Customer-Managed Encryption Keys (CMEK) and Egress Perimeter Controls, aimed at meeting stringent regulatory requirements for industries such as healthcare and financial services. These new capabilities allow customers to manage their cryptographic keys directly through Azure Key Vault, ensuring complete control over key lifecycle, rotation, and revocation, which is crucial for HIPAA and PCI DSS compliance. Egress Perimeter Controls provide an additional layer of security by allowing organizations to define specific network endpoints for outbound data traffic, thus preventing unauthorized data exfiltration. This integration with Azure's ecosystem leverages modern authentication methods, such as Azure Workload Identity, allowing secure key access without exposing key material. These features support the development of scalable, resilient applications on Azure while maintaining robust security and compliance, making CockroachDB a suitable choice for enterprises in regulated sectors.

CockroachDB Cloud addresses the need for scalability and cost-efficiency in globally distributed applications by offering three service tiers—Basic, Standard, and Advanced—each designed to balance flexibility, control, and cost through a multi-tenant architecture. The Basic and Standard tiers focus on affordability and ease of use, including a "free forever" option for small clusters, while the Advanced tier provides dedicated resources for larger enterprises. This multi-tenant approach allows for sub-second cold starts, automatic scaling to zero, multi-region operations, and strong workload isolation, ensuring consistent performance and resilience akin to distributed SQL systems. These capabilities, detailed in a research paper presented at SIGMOD 2025, have been integral to the CockroachDB Cloud production environment for over four years, enabling efficient and secure management of tenant clusters. Rebecca Taft, Director of SQL Engineering at Cockroach Labs, has been instrumental in developing the cost-based query optimizer and supporting features like geospatial indexing, drawing on her academic background from MIT and Yale.