November 2018 Summaries

GraphQL has gained popularity among developers for implementing APIs due to its simplicity and discoverability features. It is especially useful when working with Apache Cassandra, where designing the proper data model is crucial. In this text, a detailed guide on how to create an API on top of Apache Cassandra using GraphQL is provided. The main advantages of GraphQL include allowing only valid queries, ensuring performance, and providing a perfect match for Cassandra usage by not allowing users to execute forbidden queries like select * from myHugeTable. The text also covers the implementation process in Java with the graphql-java framework and how to enable the interactive GraphQL IDE, GraphiQL, for testing purposes. Overall, using GraphQL with a Cassandra backend can help enforce good performance on your clusters while allowing queries you want to allow.

Black Friday presents a significant challenge to retailers as they must handle massive data influx without experiencing system outages. To ensure success during the holiday season, retailers should be prepared with databases that can manage high traffic and maintain zero fault tolerance. This can be achieved by using masterless architectures, NoSQL databases, and hybrid-cloud deployment strategies. DataStax Enterprise (DSE) offers continuous availability and infinite scalability to support these requirements, ensuring seamless holiday sales experiences for customers.

Companies today aim to increase productivity and improve user experiences by accelerating their computer networks and ensuring high availability, fault tolerance, and scalability of their apps and data centers. Hybrid cloud environments pose challenges in achieving these goals due to geographical distances and regulatory requirements such as GDPR. Distributed, scale-out databases have emerged as a solution, allowing an Active Everywhere approach where data is replicated across all data centers during read or write operations. This approach provides resilience and additional benefits like high availability, fault tolerance, and easy scaling.

Relational database management systems were dominant in the 1980s and 1990s, but their limitations led to the emergence of NoSQL databases in the 2000s. NoSQL databases can be categorized into key-value stores, wide columnar stores, document databases, and graph databases. Key-value stores are simple and highly performant for equality searches with simple payloads. Wide columnar stores allow storage of more complex structured data and support distribution and scaling. Document databases offer flexibility in structures and can be easily mapped to application objects. Graph databases represent data as vertices and edges, excelling at handling use cases where data relationships are valuable. Choosing the right NoSQL database depends on understanding each tool's capabilities and matching them to specific application requirements. The concept of polyglot persistence involves using a variety of tools for storing data based on an application's needs, especially in systems demanding extreme performance at scale.

Hybrid databases are gaining popularity as companies move to hybrid cloud environments, dealing with resulting data silos. Traditionally, hybrid databases combined in-memory and on-disk storage for efficient data processing. However, today's hybrid databases need to be architecturally versatile, combining public and private clouds. As data storage costs decrease and performance increases, organizations focus more on deploying applications in hybrid cloud environments. Modern hybrid databases should provide consistent experiences across various platforms and locations, ensuring high availability, reliability, and scalability. Examples of such databases include DataStax Enterprise (DSE), built on Apache Cassandra's masterless architecture, which runs across multiple clouds and on-premises in a single second cluster.

The global hybrid cloud market is expected to grow from $44.6 billion in 2018 to $97.64 billion by 2023, driven by benefits such as agility, security, and cost-effectiveness. However, organizations often face challenges when migrating to the hybrid cloud, including vendor lock-in, inconsistent data models leading to data silos, and difficulties in developing transformative applications. To avoid these issues, companies should conduct thorough due diligence before choosing a cloud vendor, ensure flexibility and control over their data, and adopt solutions that support multi-cloud replication for seamless application deployment across multiple environments.

Five out of six DataStax Developer Days have been held this month, with a focus on topics such as Cassandra skills and multi-cloud strategies. The London event highlighted the growing interest in multi-cloud solutions among developers to avoid vendor lock-in and ensure flexibility. DataStax is actively seeking feedback from attendees to improve their offerings for the developer community. Upcoming events include Paris and DataStaxAccelerate, a conference on Apache Cassandra taking place in May 2019 in Washington D.C. Area.

DataStax, a major contributor to the Apache Cassandra community, focuses on three key areas: Code, Coaching, and Conferences. They contribute significantly to the development of Cassandra itself, including features for upcoming releases. Additionally, they maintain popular drivers for seven languages, totaling an estimated 36 engineer-years of effort. DataStax also contributes to Apache TinkerPop™, a graph framework used by various vendors. In terms of education and coaching, DataStax Academy offers free training on Cassandra and DSE, while the Developer Relations team hosts hands-on learning events and participates in meetups. Finally, DataStax organizes conferences like the new DataStax Accelerate for data professionals to learn about cutting-edge content related to Cassandra.

VMware and DataStax have collaborated to power big data environments using VMware's vSAN. This joint effort aims to provide operational simplicity, central management of storage, and cost savings for cloud-native customers. DataStax Enterprise is built on Apache Cassandra™ and uses a masterless architecture with data replication between nodes for availability. The Host-Affinity policy offers additional flexibility in configuring vSAN placement and replication specific to the deployed application while maintaining data locality with DataStax Enterprise compute. This solution provides benefits such as simplified visibility, lifecycle management, and consolidation of computing workloads. The collaboration results in a reference architecture that demonstrates performance consolidation, failure mitigation, and linear scaling using vSAN and DataStax together.

Hybrid cloud architectures are becoming increasingly common among enterprises, with 69% expected to use one by next year. This approach combines on-premises, public cloud, and private cloud options to meet the demands of customer experience. However, hybrid clouds bring more complexity as IT departments face various choices and comparisons. Balancing the right mix of clouds based on company goals, risk tolerance, application development needs, and budget is crucial. Hybrid clouds enable enterprises to take advantage of the pros of each cloud while minimizing risks and meeting compliance needs. An enterprise data layer that sits on top of multi-cloud and hybrid cloud architectures can address three major concerns: inconsistent data models, vendor lock-in, and data silos. DataStax Enterprise Active Everywhere architecture enables enterprises to build a consistent enterprise data layer across on-premises, hybrid cloud, and multi-cloud environments, eliminating data silos and accelerating hybrid/multi-cloud deployments.